Compare commits

...
64 Commits
Author SHA1 Message Date
yiling f7ed3bafdf Release v0.5.3 2026-10-16 22:50:47 +08:00
Yiling-J 9f76e61e23 Add documentation for LiteRT-LM Embedding Server setup
Updated documentation for LiteRT-LM Embedding Server setup and other features.
2026-10-16 22:42:07 +08:00
Yiling-J 7c297895e5 feat(embedding): add compose tiers, docs, and regenerate-embeddings CLI (#480)
* feat(embedding): update compose tiers, docs, and add regenerate-embeddings CLI

- Add Lite, Embedding, and Full Docker Compose configurations (and Aliyun mirrors)
- Set CONCURRENCY_AGENT to 1 for compose files containing embedding-server
- Update README and README_CN with compose options, recommended RAM, and semantic search links
- Add LiteRT-LM Embedding Server setup guide in docs/using-shumai/embedding-server.mdx
- Add shumai regenerate-embeddings CLI command and EmbeddingService to regenerate embeddings across teams
- Add comprehensive test coverage for EmbeddingService

* fix(docker): remove basic auth for turntable-renderer in full compose

* fix(agent): lock and disable embedding agent when ENABLE_EMBEDDING_AGENT is false

Ensure that when ENABLE_EMBEDDING_AGENT is set to 'false', existing embedding agents are automatically disabled on startup, marked as read-only, and blocked from manual API modifications or creations. Additionally guard background task creation and semantic search when the environment variable disables embedding agents.

* test(upload): remove flaky status assertion on workflow tasks
2026-10-16 21:20:25 +08:00
Yiling-J 0f318b4050 feat(search): render results in responsive card grid with sprite preview and new-tab navigation (#479)
* feat(search): render results in responsive card grid with sprite preview and new-tab navigation

* fix(search): align semantic matched duration badge style with standard duration badge

* fix(webui): use pointer cursor instead of crosshair cursor for sprite scrubber on hover
2026-10-16 10:42:32 +08:00
Yiling-J 1a46f63160 fix(agent): adaptively set video embedding frame count based on aspect ratio (#477)
* fix(agent): adaptively set video embedding frame count based on aspect ratio

Adjust sampled video frame counts between 15 and 14 based on the video's aspect ratio. Standard aspect ratios (1:1, 4:3, 3:4, 16:9, 9:16, 21:9) stay within the 2048 token signature bucket when using 15 frames, while non-standard or custom aspect ratios use 14 frames to guarantee that sequence length never exceeds 2048 tokens, preventing out-of-memory errors on the embedding server.

* feat(agent): persist video chunk embeddings incrementally and resume on retry (#478)
2026-10-16 08:36:51 +08:00
Yiling-J 6cc90bf973 feat(search): generate query embeddings directly for semantic search (#476)
* feat(search): generate query embeddings directly for semantic search

* fix(search): resolve embedding provider from env, redact query logs, and preserve case in cache
2026-10-15 20:49:45 +08:00
Yiling-J b7271777e0 feat(agent): support gemma-embedding-2-openai provider (#475)
* feat(agent): support gemma-embedding-2-openai provider

Add support for local LiteRT-LM Gemma Embedding 2 provider with 768-dimensional vector embeddings, environment-configured read-only embedding agents, 16-frame uniform sampling for video chunks, and automatic WebP-to-JPEG conversion.

* refactor(agent): remove unused legacy generateEmbeddingActivity

* chore(agent): align sharp version to 0.35.4

* refactor(agent): rename embedding server env to GEMMA_EMBEDDING_OPENAI_BASE_URL

* refactor(agent,transcode): use native Bun.Archive for video chunk frames

* fix(transcode,agent): populate buffer entries in Bun.Archive and validate frame sizes
2026-10-15 15:49:24 +08:00
yiling c302c162a0 Release v0.5.2 2026-10-13 21:03:51 +08:00
Yiling-J 432be04a96 Update CHANGELOG to remove fixed and changed entries
Removed fixed and changed entries related to 3D visual markup and settings adjustments.
2026-10-13 20:18:03 +08:00
yiling 6c95d14fb3 fix(webui): reverse 3D sprite scrubbing and remove container volumes from turntable doc 2026-10-13 16:49:42 +08:00
Yiling-J 3263580e89 fix(webui): clear markup annotations when rotating 3d model (#474) 2026-10-13 15:21:12 +08:00
yiling f3da916905 fix(asset): remove unbundled GLTF format from supported 3D models 2026-10-13 11:38:25 +08:00
Yiling-J fe8d404a44 feat(agent): support 3D degree comments, markup annotations, and config3d in read_asset (#473)
* feat(agent): support AI embedding, autofill, and read_asset for 3D assets

* feat(agent): support 3D degree comments, markup annotations, and config3d in read_asset

* fix(agent): resolve 3D session annotations to turntable degree frame time
2026-10-13 00:03:21 +08:00
Yiling-J 0f9d213b1d docs: update 3D asset and Gotenberg integration guides (#472) 2026-10-12 16:55:20 +08:00
Yiling-J f952318b34 feat(team): add integrations settings with turntable and gotenberg (#471)
* feat(team): add integrations settings with turntable and gotenberg

* fix(upload): pass teamId to gotenberg availability check during upload preflight
2026-10-12 15:38:15 +08:00
Yiling-J 43a2eced75 feat(transcode): add 3D asset support and turntable viewer (#454)
* feat(transcode): add 3D asset support and turntable viewer

* perf(turntable): stream file via Bun.file and increase upload timeout

* feat(turntable): increase upload timeout to 15m to support 5GB uploads

* fix(webui): align 3D model viewer layout and full-width control bar with video viewer

* feat(webui): unify 3D viewer bottom bar and redesign cyclic ruler slider

- Place RulerSlider in single row with playback and zoom controls, centered with max 320px width
- Flank RulerSlider with minus and plus buttons to step angle by 15°
- Redesign RulerSlider with fixed center indicator and moving cyclic track with continuous 360° wrapping
- Render major ticks longer at 60° intervals without numeric text labels on track
- Display floating degree tooltip above center indicator only during dragging

* fix(webui): prevent RulerSlider from collapsing to 0px width in control bar

- Use flex layout with flex-1 on left and right clusters to preserve center cluster width
- Enforce min-w-[140px] and w-full on RulerSlider
- Fix tooltip positioning and remove outer overflow-hidden

* feat(webui): snap RulerSlider track directly to discrete 15° ticks during drag

* fix(webui): prevent ruler snap-back caused by floating point second truncation

- Use Math.round instead of Math.floor in secondToDegree to avoid 0.999996 dropping to previous frame
- Ensure final dragDegree is committed on pointer up

* fix(webui): fix arrow key frame rounding and duplicate event bubbling

- Use Math.round instead of Math.floor for keyboard arrow frame navigation in ModelViewer
- Add e.stopPropagation() to RulerSlider keyboard events to prevent duplicate bubbling to window listener

* feat(webui): redesign RulerSlider with encapsulated step buttons and capsule pill track

* feat(webui): make RulerSlider responsive on mobile and remove bottom bar reset button

* fix(webui): load 3d video directly without poster and invert drag rotation to counterclockwise

* fix(webui): fix 3d asset viewer frame seeking and drag jitter

* feat(transcode): show 3D asset poster preview immediately after render start

* fix(viewer): synchronize 3D turntable playback via frame callback to prevent skipped degrees

* fix(viewer): handle touchpad zoom gestures and prevent page zoom in 3D viewer

* fix(viewer): eliminate 3D model rotation jitter at 345 degrees

* fix(share): watermark 3D asset turntable video on share links

* fix(share): preserve video transcode resolution on watermarked share links

* fix(viewer): download original 3D model file instead of rendered turntable MP4
2026-10-11 21:32:28 +08:00
Christopher BuzickyandBristopher e894457275 fix(transcode): apply EXIF orientation to non-RAW image transcodes and reported size (#461)
* fix(transcode): apply EXIF orientation to image transcodes and reported size

Cameras and phones store portrait shots as landscape pixels plus an EXIF
orientation tag. sharp ignores the tag and the webp output drops it, so
thumbnails, previews, watermarked share-link images and annotated snapshots
came out sideways, and getImageInfo reported width and height swapped
(a 5152x7728 portrait recorded as 7728x5152).

RAW previews keep their existing handling: transcodeImage still rotates them
from the RAW container orientation and only auto-orients inputs without one,
so RAW and non-RAW images are never rotated twice.

- transcodeImage: autoOrient non-RAW input and size previews from the
  oriented dimensions.
- getImageInfo: report metadata.autoOrient dimensions for non-RAW images.
- compositeOverlayToWebpFile and overlayAnnotationsOnBuffer: autoOrient
  the original, and draw annotations in the displayed coordinate space.
- Tests: service tests that reproduce the bug and guard against double
  rotation of RAW, and a workflow e2e case with a real 600x400 JPEG tagged
  orientation 6 that must come out 400x600.

* refactor(transcode): share orientation helper and test all 8 EXIF orientations

Move the autoOrient decision and the displayed-size lookup into orientedSharpOptions and displayedDimensions so getImageInfo, transcodeImage, compositeOverlayToWebpFile and overlayAnnotationsOnBuffer share one rule: auto-orient unless a RAW container orientation was already applied. Behavior is unchanged.

Add transcode.orientation.test.ts, which runs real sharp over a quadrant-colored JPEG tagged with each of the 8 EXIF orientations and checks the reported and transcoded size plus the top-left and bottom-right corner colors.

* fix(transcode): never auto-orient dcraw output

The dcraw_emu fallback reports orientation undefined because LibRaw already rotates upright, so transcodeImage applied autoOrient to the TIFF and could rotate twice if it kept an Orientation tag. extractAndValidateRawPreview now flags the fallback result as orientationApplied and transcodeImage treats it as already oriented (autoOrient false). Tests cover both the dcraw path and an embedded preview with no container orientation.

---------

Co-authored-by: Bristopher <Bristopher@users.noreply.github.com>
2026-10-10 15:41:49 +08:00
4b1b64fe2b feat(kanban): link related assets to their file or folder (#453)
Related assets in the task edit dialog were plain rows with no way to
open them. Add an opt-in `linkable` prop to TaskRelatedAssets that turns
each row's thumbnail and name into a router Link to the asset in its
project (files -> /projects/$projectId/files/$fileId, folders ->
/projects/$projectId/folders/$folderId), falling back to the task's
project when the asset has none. The edit dialog opts in; the create
dialog does not, so a click cannot discard an unsaved draft.

Co-authored-by: Frederic De Decker <hello@trinity.video>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 20:14:36 +08:00
yiling f8beb195dd Release v0.5.1 2026-10-06 22:29:50 +08:00
yiling a97996cb59 docs(changelog): note clickable notification cards 2026-10-06 22:18:18 +08:00
jorgeandYiling-J f870ca9b49 feat(asset): link to assets on notification cards. (#451)
* feat(asset): link to assets on notification cards.

* Update packages/webui/components/notification-card.tsx

Co-authored-by: Yiling-J <njjyl723@gmail.com>

---------

Co-authored-by: Yiling-J <njjyl723@gmail.com>
2026-10-06 22:14:11 +08:00
Yiling-J 2e5ff281a8 feat(transcode): add hardware decoding support for nvenc, qsv, and rkmpp (#452)
* feat(transcode): add hardware decoding support for nvenc, qsv, and rkmpp

* refactor(transcode): eliminate redundant hardware decode gating and dead condition
2026-10-06 17:35:14 +08:00
yiling de8decf113 fix(ci): prevent tagging main on PR workflow_dispatch docker builds 2026-10-06 11:49:49 +08:00
Cameron HarrisandClaude Opus 5.5 dd57fb9ed2 feat(transcode): decode and scale on the GPU for hardware transcodes (VA-API) (#449)
* feat(transcode): decode and scale on the GPU for VA-API transcodes

When the selected encoder is h264_vaapi, decode with -hwaccel vaapi and
scale with scale_vaapi so frames stay in GPU memory, instead of decoding
and scaling on the CPU and hwupload-ing before the encoder. On a Haswell
iGPU this makes 4K proxy/HLS renditions 3-5x faster.

Fallback order: GPU decode + VA-API encode -> software decode + VA-API
encode (previous behavior) -> libx264. Watermark overlays, HDR sources
and rotated sources skip GPU decode: ffmpeg cannot auto-insert its
rotation filter on hardware frames, so rotated input would otherwise come
out sideways instead of failing. SHUMAI_VAAPI_HW_DECODE=false disables it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* docs(transcode): document VA-API GPU decoding and SHUMAI_VAAPI_HW_DECODE

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* refactor(transcode): generic hardware decode config, stall watchdog, reuse probed rotation

Follow-ups from comparing with Immich and Jellyfin:

- Replace the VA-API-specific helpers with a per-encoder HW_DECODE_CONFIGS
  table (input args, output args, GPU scale filter) and build the MP4 and
  HLS input/filter graph in one shared method so the paths cannot drift.
  Only h264_vaapi has an entry; other backends can be added once tested on
  real hardware. The opt-out is now SHUMAI_HW_DECODE=false.
- Kill a GPU-decode attempt with SIGKILL when its output stops growing for
  SHUMAI_HW_DECODE_STALL_TIMEOUT (default 120s) and fall back, since some
  VA-API driver failures hang ffmpeg instead of exiting (immich#31933).
- Add -noautoscale on the GPU-decode path so a mid-stream decoder reinit
  (e.g. 1080 -> 1088) cannot auto-insert a software scaler into the GPU
  graph (jellyfin#17133).
- Pass the rotation already probed into MediaMetadata through the workflow
  and activities as sourceRotation; ffprobe only runs when it is missing.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-05 12:00:01 +08:00
Yiling-J 9a640632aa ci(release): allow manual Docker image build from external PR (#450) 2026-10-05 09:48:48 +08:00
yiling e61641d7ec fix(workflow): increase activity startToCloseTimeout to 120 minutes 2026-10-04 10:22:31 +08:00
Yiling-J 7ca8ba8b2a fix(transcode): resolve RAW transcoding hang in Docker deployments (#447) 2026-10-03 12:45:12 +08:00
Yiling-J 8805d5799a feat(transcode): add dcraw_emu fallback for RAW image preview generation (#446) 2026-10-03 11:03:31 +08:00
yiling 013c4a8156 Release v0.5.0 2026-10-02 20:28:37 +08:00
Yiling-J 20b5ebeefd fix(asset): mark asset as failed and display failure state in webui when transcoding fails (#445)
* fix(asset): mark asset as failed and display failure state in webui when transcoding fails

* fix(transcode): do not mark asset as failed if media processing already succeeded

* test(workflow): add e2e test for video transcode failure status transition

* fix(webui): update failed file card badge to top-right Transcoding Failed label with error tooltip
2026-10-02 20:09:32 +08:00
Yiling-J daa20aed20 fix(webui): add video loading state and poster placeholder (#444)
* fix(webui): add video loading state and poster placeholder

* fix(webui): prevent buffer progress events from canceling stall indicator
2026-10-02 16:19:49 +08:00
Yiling-J 3564d4d164 fix(transcode): avoid selecting attached picture streams for video transcoding (#443)
* fix(transcode): avoid selecting attached picture streams for video transcoding

Filter out attached picture streams (disposition.attached_pic) during video probing and stream selection, matching Immich's default and bitrate selection strategy. Forward explicit primary video and audio stream indices to FFmpeg pipelines to prevent picking embedded cover art or thumbnail streams.

* fix(transcode): skip poster and sprite generation for audio-only assets and add multi-stream e2e tests

* test(workflow): scope pixel format to h264 stream and use yuvj420p for mjpeg cover art
2026-10-02 13:51:17 +08:00
Yiling-J 206763e966 feat(auth): add demo user login button and update demo description (#442)
* feat(auth): add demo user login button and update description

* feat(auth): move demo hint below button and add discord link
2026-10-01 22:04:15 +08:00
Yiling-J 4e4da595a6 feat(settings): redesign video transcoding settings into dedicated cards and rename team AI settings (#441)
* feat(settings): redesign video transcoding settings into dedicated cards and rename team AI settings

* feat(settings): rename Video Strategy to MP4 Transcoding with sub-description
2026-10-01 21:19:25 +08:00
Yiling-J a21a0c9655 fix(settings): prevent deselecting last ladder and coalesce rapid toggles (#440) 2026-10-01 16:35:47 +08:00
Yiling-J fc7282581f fix(share): prevent unwatermarked HLS playback on public shares before transcode completion (#439) 2026-10-01 15:41:45 +08:00
Yiling-J e007054f47 feat(transcode): add multi-resolution strategy with configurable ladders and fallback (#438)
* feat(transcode): add multi-resolution strategy with configurable ladders and fallback

* feat(settings): show toast on settings update and add manual verification plan to PR template

* fix(webui): preserve video resolution and fix 480p label in download dropdowns

* fix(webui): display standard short-side resolution for vertical videos in seekbar

* fix(webui): fix portrait transcode resolution labeling and serialize settings toggles
2026-10-01 13:27:50 +08:00
Yiling-J a84da8678e feat(webui): switch from video.js to hls.js and native video player (#437)
* feat(webui): switch from video.js to hls.js and native video player

* fix(webui): prevent stale seek hijacking and guard native HLS compare mode
2026-09-30 20:23:44 +08:00
Yiling-J 123a8cb5f4 feat(transcode): add HLS live streaming support (#436)
* feat(transcode): add HLS live streaming support

* fix(hls): resolve version stacks for public HLS and add workflow E2E test
2026-09-30 16:31:40 +08:00
Yiling-J 5fe22ac52c feat(transcode): add initial camera RAW image support (#435)
- Add centralized RAW file extension set and detection helper for 23 common camera RAW formats
- Update getProxyType and EXTENSION_MIME_TYPES to route RAW formats into the image pipeline
- Add exiftool-vendored runtime dependency and mark as external in build-npm script
- Implement embedded JPEG extraction to temporary file with priority JpgFromRaw2 -> JpgFromRaw -> PreviewImage
- Read and apply EXIF orientation from RAW container to prevent rotation/mirroring issues
- Integrate into TranscodeService.getImageInfo and transcodeImage with proper temporary file cleanup
- Add hasUsablePreview guard in transcodeImageWorkflow to gracefully handle RAW files without embedded previews
- Add unit, service, workflow, and E2E tests for RAW processing
2026-09-29 21:09:27 +08:00
Yiling-J 11d21444cf fix(trash): prevent empty trash dialog from closing before API call completes (#434)
AlertDialogAction from Radix UI auto-closes the dialog on click.
Added e.preventDefault() so the dialog stays open and shows the
'Deleting...' loading state until the operation finishes.

Fixes both desktop and mobile file browsers.
2026-09-29 10:40:13 +08:00
yiling f2b0415580 Release v0.4.9 2026-09-28 20:17:50 +08:00
Yiling-J c9d0b476cd feat(transcode): fallback to software encoding when hardware transcode fails (#433) 2026-09-28 16:45:28 +08:00
yiling bf5ac566b6 refactor(notification): simplify email wording and remove redundant team references 2026-09-28 13:37:16 +08:00
Yiling-J af8f24aa95 docs(settings): add email notifications documentation to team settings (#432)
* docs(settings): add email notification documentation to team settings

* fix(notification): use correct webui route URLs for project and asset links in emails
2026-09-28 11:27:59 +08:00
Yiling-J d98f486c9d feat(notification): add email notification infrastructure, smtp configuration, and periodic background delivery job (#431)
* feat(notification): add email notification support and dedicated workflow

* feat(webui): add team email notification settings and test email dialog

* refactor(webui): remove redundant card header from email notification settings

* fix(notification): prevent sending notification emails to AI agents

* feat(notification): display project, avatar, upload time and resolve mention names in notification emails

* refactor(notification): remove uploadDebounceSeconds and naive workflow sleep

* refactor(notification): remove temporal notification workflow and keep email infrastructure

* fix(team): prevent leaking email notification settings and credentials in team settings

* feat(notification): add periodic background job for batched email notification delivery

* fix(notification): embed avatars and thumbnails as inline cid attachments
2026-09-28 09:58:59 +08:00
Yiling-J 94ea4c917a feat(transcode): add VA-API hardware acceleration support (#430)
* feat(transcode): add VA-API hardware acceleration support

* fix(docker): switch to jellyfin-ffmpeg 8.1.2-5 for VA-API support

* fix(transcode): probe hardware encoder usability to skip uninitialized drivers

* ci(release): add workflow_dispatch trigger with custom tag input

* fix(transcode): use 256x256 frame dimension for encoder usability probe

* feat(docker): add hwaccel transcoding compose presets matching immich

* docs: add hardware transcoding documentation and quickstart references

* feat(transcode): prioritize h264_vaapi before h264_qsv on Linux

* fix(docker): extend device permission loop to rkmpp and wsl accelerator nodes
2026-09-26 20:55:57 +08:00
Yiling-J 941bf0fd1b feat(transcode): smart poster frame selection and decoupled generation (#428)
* feat(transcode): smart poster frame selection and decoupled generation

* fix(transcode): detect MPEG-TS extension from URL pathname for signed remote sources
2026-09-25 16:37:16 +08:00
Yiling-J 59742dde78 feat(transcode): optimize video sprite generation with seek pool and compositing (#427)
* feat(transcode): optimize video sprite generation with seek pool and compositing

* fix(transcode): handle abort signal during and after sharp compositing
2026-09-25 15:06:15 +08:00
yiling 29cd5d807c test(scripts): find release by version in changelog integration test 2026-09-24 20:34:53 +08:00
yiling f6e27095a5 Release v0.4.8 2026-09-24 20:26:23 +08:00
Yiling-J 55637bc9a8 feat(transcode): optimize video proxy quality and fix videotoolbox first-frame blocking (#426)
- Fix first-frame macroblocking on VideoToolbox hardware acceleration by avoiding -maxrate/-bufsize which triggers Apple's DataRateLimits initial rate controller starvation, applying target -b:v instead.
- Optimize software encoding (libx264) quality by adjusting CRF from 26 to 23.
- Disable B-frames (-bf 0) for libx264 proxies to ensure zero-latency frame stepping (JKL) and scrubbing, matching Frame.io review proxy standards.
- Update unit tests to verify the new encoder configurations and bitrate calculation logic.
2026-09-24 17:16:02 +08:00
Yiling-J 0d6671c3e1 fix(webui): show play button and darken tint only before initial playback (#425)
* fix(webui): remove paused play button overlay and darken cover from video players

* fix(webui): show play button and darken tint only before initial playback
2026-09-24 15:19:46 +08:00
Yiling-J c3375ee935 feat(transcode): add HDR video proxy support and playback (#424)
* feat(transcode): add HDR output proxy support

* fix(transcode): auto-detect MoltenVK Vulkan ICD on macOS for libplacebo filters

* fix(transcode): split filter in generateSprite to ensure poster uses tone-mapped stream

* docs(changelog): update docs/changelog.mdx and ignore in prettier

* fix(transcode): respect video rotation metadata to preserve display aspect ratio

* fix(webui): render native video with hardware EDR acceleration instead of 2D canvas

* fix(webui): show HDR label on mobile video control bar

* fix(share): preserve hdr flag on watermarked video transcodes

* fix(webui): differentiate HDR download options in video player

* fix(transcode): handle Dolby Vision Profile 5 via standard HDR pipeline without libplacebo

* fix(transcode): prepend setparams filter to prevent zscale colorspace path error for DV P5

* refactor(transcode): simplify HDR proxies to always produce HDR without SDR setting
2026-09-24 14:32:47 +08:00
Yiling-J f579947770 feat(docs): add changelog section to mintlify documentation (#422) 2026-09-22 21:02:27 +08:00
Yiling-J 53c9485f10 fix(core): only manage PID files in daemon mode (#421) 2026-09-22 20:31:38 +08:00
Yiling-J 75bf9b3b88 fix(docker): stop publishing postgres port in compose files (#420)
The bundled PostgreSQL service published port 5432 on the host, which caused
docker compose up to fail when that port was already in use. A failed start
leaves the container without a network endpoint, so subsequent runs reuse the
broken container and the app cannot reach the database (Prisma P1001).

The app only needs the database inside the compose network, so the host port
mapping is unnecessary. Remove it from all local and temporal compose files.
2026-09-20 11:46:13 +08:00
yiling 657cd761bc Release v0.4.7 2026-09-19 16:01:20 +08:00
Yiling-J de634a5dc5 feat(asset): show poster and hover preview while video is transcoding (#419)
* feat(asset): show poster and hover preview while video is transcoding

Generate and persist the sprite (and its poster) before the proxy
transcodes run, so the poster/sprite are available in the file list
while the asset is still processing.

Render the preview in the file card behind a translucent processing
overlay instead of a blank skeleton, and replace hardcoded status text
with existing i18n keys.

* feat(asset): pulse the cover and show Preparing/Uploading labels while processing

Drop the processing text overlay on the file card preview in favour of a
subtle pulsing opacity on the cover, and show "Uploading..."/"Preparing..."
in the creator row so the real author and date only appear once the asset
is ready.

* feat(asset): use a hollow ring for the preparing placeholder

Match the geometry of the upload progress ring (empty inside, no
percentage) and hide the duration badge while processing, so the
upload-to-processing transition reads as one continuous animation.

* style(asset): shrink-only preparing ring in muted grey at fixed opacity

Animate only the ring size (1 -> 0.8 -> 1) and hold a constant 0.6
opacity, using a muted grey stroke instead of primary.
2026-09-19 15:25:41 +08:00
Yiling-J 09b65a1eb6 feat(asset): Recently Deleted previews, retention countdown, and version stack delete/restore (#418)
* feat(asset): improve Recently Deleted previews and retention indicators

- Include soft-deleted children when building folder previews for the
  Recently Deleted view, and resolve deleted version stacks to their
  latest version so a trashed folder shows what was inside it.
- Add a "days left" badge (clock + Nd) to file and folder cards in the
  Recently Deleted view, placed to the left of the comments badge.
- Fix an issue where deleting or restoring a version stack from the file
  list did nothing because the WebUI dropped version_stack items before
  calling the API.
- Share the 30-day trash retention window through TRASH_RETENTION_DAYS.

* test(asset): use typed MediaInfo in recently-deleted preview tests

Replace the explicit any cast with a fully typed MediaInfo object so the new tests comply with the no-explicit-any policy.
2026-09-19 11:36:14 +08:00
Yiling-J b52b2ce7d8 fix(transcode): preserve trashed status during background jobs and handle permanent purge (#416)
* fix(transcode): preserve trashed status during background jobs and handle permanent purge

* fix(workflow): clear heartbeat and add info logs on task cancellation

* fix(transcode): terminate media processes and guard task activities on cancellation

- Pass AbortSignal to ffmpeg and pdftoppm child processes in transcodeService to terminate processes on task cancellation
- Add local task AbortController context and cancellation guards in executeActivity to prevent subsequent workflow activities
- Forward taskId and cancellation signals in transcode, screenshot, and PDF activities
- Use updateMany instead of update in task activities to eliminate Prisma P2025 errors on purged task records

* fix(workflow): allow cleanup activities in finally blocks during task cancellation

Remove blanket cancellation check in executeActivity so that workflows can execute cleanup and failure handling activities in catch and finally blocks upon cancellation, while relying on AbortSignal in active activities to abort execution.

* docs(changelog): improve user friendliness of changelog entries and formatting example

* docs(agents): update changelog guidelines to emphasize end-user friendly descriptions

* fix(transcode): handle queued cancellation and delete races in media workflow
2026-09-19 10:40:49 +08:00
Yiling-J 145aa6bfb4 fix(webui,core): center breadcrumb menu on chevron and fix proxy download filename (#417) 2026-09-18 23:14:54 +08:00
Yiling-J a91d8c4521 feat(settings): add FFmpeg threads setting for media processing (#415)
* feat(settings): add FFmpeg threads setting for media processing

* style(settings): add bottom padding to media processing page
2026-09-18 13:30:33 +08:00
Yiling-J 3a509537fd feat(changelog): add CHANGELOG.md, release extraction script, and developer guidelines (#414)
* feat(changelog): add CHANGELOG.md, release extraction script, and developer guidelines

* docs(agents): simplify changelog guidelines for brevity
2026-09-18 11:57:25 +08:00
273 changed files with 38063 additions and 3058 deletions
+26 -2
View File
@@ -68,18 +68,42 @@ jobs:
- name: Build NPM packages
run: bun run build-npm
- name: Generate release notes
env:
GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
run: |
CHANGELOG_NOTES=$(bun run scripts/extract-changelog.ts "${{ env.RELEASE_TAG }}" 2>/dev/null || true)
PR_NOTES=$(gh api "repos/${{ github.repository }}/releases/generate-notes" -f tag_name="${{ env.RELEASE_TAG }}" --jq .body 2>/dev/null || true)
{
if [ -n "$CHANGELOG_NOTES" ]; then
echo "$CHANGELOG_NOTES"
echo ""
echo "---"
echo ""
fi
if [ -n "$PR_NOTES" ]; then
echo "$PR_NOTES"
fi
} > /tmp/release-notes.md
- name: Create GitHub Release
env:
GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
run: |
if [ ! -s /tmp/release-notes.md ]; then
NOTES_FLAG="--generate-notes"
else
NOTES_FLAG="--notes-file /tmp/release-notes.md"
fi
if gh release view "${{ env.RELEASE_TAG }}" >/dev/null 2>&1; then
gh release edit "${{ env.RELEASE_TAG }}" \
--title "${{ env.RELEASE_TAG }}" \
--generate-notes
$NOTES_FLAG
else
gh release create "${{ env.RELEASE_TAG }}" \
--title "${{ env.RELEASE_TAG }}" \
--generate-notes
$NOTES_FLAG
fi
- name: Upgrade npm for trusted publishing
+18 -1
View File
@@ -4,6 +4,16 @@ on:
push:
branches: ['main']
tags: ['v*']
workflow_dispatch:
inputs:
tag:
description: 'Docker image tag (e.g. feature-vaapi)'
required: false
type: string
pr:
description: 'Pull request number (e.g. 449)'
required: false
type: string
jobs:
build-and-push:
@@ -14,7 +24,10 @@ jobs:
steps:
- name: Checkout repository
id: checkout
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr && format('refs/pull/{0}/head', inputs.pr) || github.ref }}
# Set up Docker Buildx
- name: Set up Docker Buildx
@@ -45,7 +58,11 @@ jobs:
crpi-000nkafzu6xvpwsc.cn-hongkong.personal.cr.aliyuncs.com/shumai-one/shumai
tags: |
type=semver,pattern={{version}}
type=ref,event=branch
type=ref,event=branch,enable=${{ github.event_name == 'push' || (github.event_name == 'workflow_dispatch' && !inputs.pr && !inputs.tag) }}
type=raw,value=pr-${{ inputs.pr }},enable=${{ github.event_name == 'workflow_dispatch' && inputs.pr != '' && inputs.tag == '' }}
type=raw,value=${{ inputs.tag }},enable=${{ github.event_name == 'workflow_dispatch' && inputs.tag != '' }}
labels: |
org.opencontainers.image.revision=${{ steps.checkout.outputs.commit }}
# Build and push Docker image
- name: Build and push Docker image
+2 -1
View File
@@ -7,4 +7,5 @@ packages/core/src/generated
src/ui/components/ui
packages/webui/paraglide
*.md
**/*.md
**/*.md
docs/changelog.mdx
+24 -3
View File
@@ -14,9 +14,10 @@ Complete the task on that branch.
After finishing the work and verifying that all checks pass:
1. Stage and commit your changes.
2. Push the branch to `origin`.
3. Open a pull request using the GitHub CLI.
1. Update `CHANGELOG.md` under `## [Unreleased]` for user-facing changes (see [Changelog Guidelines](#changelog-guidelines)).
2. Stage and commit your changes.
3. Push the branch to `origin`.
4. Open a pull request using the GitHub CLI.
```bash
git add .
@@ -61,6 +62,19 @@ https://www.conventionalcommits.org/en/v1.0.0/#summary
---
## Changelog Guidelines
We maintain `CHANGELOG.md` following [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
Before opening a pull request, add an entry under `## [Unreleased]` for any user-facing changes:
- **End-User Friendly**: Make entries easy to understand for end users. For new features, explain what they actually do in plain language without heavy technical jargon. For bug fixes, describe what the bug was in a friendly, relatable way rather than focusing on internal code mechanics, file paths, or private symbols.
- **Formatting**: `- **<scope>**: <Description>` (e.g. `- **asset**: Fix an issue where moving a video to trash while it was still transcoding could cause it to lose its trashed status`).
- **No PR Numbers**: Do not include PR numbers or links; the release pipeline automatically appends the detailed PR list.
- **Internal Changes**: Routine refactors, test additions, or internal chores without external behavior changes may omit changelog updates.
---
## Pull Request Template
```md
@@ -78,6 +92,13 @@ Briefly describe what this PR changes and why.
- Describe the checks or tests you ran.
- Include relevant outputs if applicable.
## Manual Verification Plan
- Provide concise instructions for manually verifying this change.
- Assume familiarity with the project and existing development workflow.
- Skip boilerplate steps (e.g. starting the app, logging in, opening common pages) unless they are part of the change.
- Focus on what needs to be verified, including scenarios, edge cases, and expected results.
## Notes
Add any additional context, caveats, or follow-up work.
+815
View File
@@ -0,0 +1,815 @@
# Changelog
All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
## [Unreleased]
### Added
### Fixed
### Changed
## [0.5.3] - 2026-10-16
### Added
* **cli**: Add administrative `shumai regenerate-embeddings` command to regenerate multimodal vector embeddings for all existing media files across active teams
* **docker**: Provide three Docker Compose deployment tiers (Lite, with Embedding Server, and Full) with recommended memory specifications
* **agent**: Save video chunk embeddings incrementally to the database as each chunk finishes, preserving completed progress on failure and resuming from where it left off on retry
* **agent**: Add support for local LiteRT-LM Gemma Embedding 2 (`gemma-embedding-2-openai`) provider and environment-enforced read-only embedding agents with automatic 768-dimensional vector search indexing
### Fixed
### Changed
* **search**: Display search results in a responsive card grid with interactive hover video sprite scrubbing, semantic match chunk time range previews, and new-tab opening on click
* **search**: Significantly improve semantic search response times by generating query text embeddings directly with in-memory caching instead of dispatching background tasks
## [0.5.2] - 2026-10-13
### Added
* **3d asset**: Add support for 3D model files (.glb, .gltf, .obj, .fbx, .stl, .dae, .usd, .usda, .usdc, .usdz) with automated turntable video rendering, hover sprite scrubbing, and an interactive 3D turntable viewer featuring 360° ruler scrubbing and degree-based markup comments
* **kanban**: Related assets on a task are now clickable, taking you directly to that file or folder in its project
* **settings**: Add an Integrations section to Team Settings with dedicated configuration for Gotenberg Document Converter, supporting connection testing and environment variable override protection
### Fixed
* **transcode**: Fix an issue where portrait photos from cameras and phones appeared sideways in thumbnails, previews, watermarked share links and annotated snapshots, and were listed with their width and height swapped
### Changed
## [0.5.1] - 2026-10-06
### Added
* **notification**: Notification cards are now clickable, taking you directly to the relevant file
* **transcode**: Add hardware decoding and GPU scaling support for NVIDIA NVENC, Intel QuickSync Video (QSV), Rockchip (RKMPP), and VA-API (Intel and AMD GPUs), significantly speeding up video proxy and HLS rendition generation across supported platforms. Set `SHUMAI_HW_DECODE=false` to turn hardware decoding off
* **transcode**: Add `dcraw_emu` fallback to generate WebP previews for camera RAW files without embedded JPEG previews
### Fixed
### Changed
## [0.5.0] - 2026-10-02
### Added
* **transcode**: Add configurable target resolution ladders (480p, 720p, 1080p, 1440p, 2160p) for MP4 Multi Resolutions transcoding in team settings, with unified resolution matching and lowest-resolution fallback for both MP4 and HLS
* **transcode**: Add HLS live streaming with fMP4/CMAF adaptive bitrate ladders (480p, 720p, 1080p, 1440p, 2160p), configurable in Transcode Settings, with automatic quality switching, manual resolution selection in the video viewer, seamless rendition switching, and full watermarking support for shared links
* **transcode**: Add basic camera RAW image support for popular camera formats (including Sony ARW, Canon CR2/CR3, Nikon NEF, Adobe DNG, Fujifilm RAF, and Panasonic RW2), extracting high-quality embedded JPEG previews and automatically applying camera orientation for fast thumbnail and proxy generation
### Fixed
- **asset**: Fix an issue where media files remained stuck in a perpetual "Preparing..." state when transcoding failed, by displaying a clear failed status and error details in the file list and file viewer
* **transcode**: Fix an issue where videos or audio-only files containing attached cover art or picture streams could cause transcoding to fail during poster generation or select a single static image instead of the video
* **webui**: Fix inconsistent video resolution labels across download menus, mobile file headers, and the video player by correctly deriving resolutions from the video's short side, including portrait videos and 480p videos
* **trash**: Fix an issue where the "Empty Trash" confirmation dialog would close immediately instead of showing a loading state while files were being permanently deleted
### Changed
* **settings**: Redesign team Video Transcoding settings with dedicated cards for Transcode Policy, Hardware Acceleration, and FFmpeg Threads, clearly explaining MP4 and HLS playback vs download behavior and storage implications
* **settings**: Rename AI Settings group to Team AI Settings in settings navigation
## [0.4.9] - 2026-09-28
### Added
- **notification**: You can now receive notifications via email, with batched digest emails and configurable team-level SMTP settings
- **transcode**: Add VA-API hardware acceleration support for Linux hosts and Docker environments with Intel and AMD GPUs, featuring automatic DRI render node discovery, variable bitrate control, and structured transcoding observability logs
- **transcode**: Add automatic fallback to software transcoding (libx264) when hardware-accelerated video encoding encounters GPU resolution constraints or runtime driver errors
- **docker**: Add modular hardware-accelerated transcoding presets matching Immich's compose configuration via `hwaccel.transcoding.yaml` with support for NVENC, QuickSync, RKMPP, and VA-API
### Fixed
### Changed
- **transcode**: Improve video poster selection using intelligent scene detection and color variance filtering to automatically skip black frames, fade-ins, and blank intros in favor of representative video content
- **transcode**: Optimize video sprite generation for large and long videos using a concurrent fast-seek pool and image compositing, avoiding full-video decoding while preserving instant single-pass generation for short clips
## [0.4.8] - 2026-09-24
### Added
- **transcode**: Add HDR video proxy support: HDR videos automatically generate HDR video proxies and preview videos preserving HDR color characteristics, tone-mapped SDR thumbnails and posters, and automatic HDR player playback on supported displays with hardware EDR acceleration
- **docs**: Add a Changelog section to the documentation site so users can explore release notes and updates directly in Mintlify
### Fixed
- **webui**: Only show the center play button and darkened tint on initial video open before playback starts; once played, pausing or scrubbing back to start keeps the video canvas clean and unobstructed for review
- **transcode**: Fix an issue where videos with orientation or rotation metadata (such as smartphone vertical videos) were incorrectly displayed with stretched aspect ratios in the video player
- **cli**: Only manage and check PID files in daemon mode, preventing stale PID files, crash-loop conflicts, or spurious process checks during normal foreground and Docker execution
### Changed
- **transcode**: Improve video proxy visual quality and disable B-frames for smoother frame-by-frame seeking and scrubbing
## [0.4.7] - 2026-09-19
### Added
- **changelog**: Introduce Keep a Changelog support, historical release summaries, and automated release notes extraction
- **release**: Automatically populate GitHub release bodies from CHANGELOG.md and promote unreleased notes during version bumps
- **settings**: Add FFmpeg threads slider to team media processing settings
- **asset**: Show previews of a folder's contents and a "days left" countdown for each item in Recently Deleted
- **asset**: Show the poster and hover preview for a video in the file list while it is still transcoding, with a gently pulsing cover (or an animated waiting circle before the poster is ready), "Uploading..."/"Preparing..." labels, and the duration hidden until it is ready
### Fixed
- **asset**: Fix an issue where moving a video to trash while it was still transcoding could cause it to lose its trashed status
- **transcode**: Immediately stop background video processing and transcoding when a file is permanently deleted from the trash
- **asset**: Fix proxy download filename using original file extension instead of proxy extension
- **asset**: Fix an issue where deleting or restoring a version stack from the file list did nothing
- **webui**: Center breadcrumb file action popup menu on the chevron icon
## [0.4.6] - 2026-09-17
### Added
- **comment**: Add emoji reaction support for comments ([#405](https://github.com/shumaiOne/shumai/pull/405))
- **premiere-uxp**: Introduce Adobe Premiere Pro UXP plugin for Shumai ([#407](https://github.com/shumaiOne/shumai/pull/407))
- **settings**: Add team appearance setting to hide agent features ([#411](https://github.com/shumaiOne/shumai/pull/411))
### Fixed
- **premiere-uxp**: Preserve remote cloud storage presigned URLs in asset resolution ([#408](https://github.com/shumaiOne/shumai/pull/408))
- **premiere-uxp**: Scope sequence links by server endpoint and surface sync errors ([#409](https://github.com/shumaiOne/shumai/pull/409))
- **api**: Enforce SHUMAI_DEMO_MODE for API token and CLI requests ([#413](https://github.com/shumaiOne/shumai/pull/413))
### Documentation
- Add NLE export and Premiere Pro UXP plugin documentation ([#412](https://github.com/shumaiOne/shumai/pull/412))
## [0.4.5] - 2026-09-13
### Added
- **core**: Export comments to FCP, Premiere, Media Composer, and Resolve ([#400](https://github.com/shumaiOne/shumai/pull/400))
- **webui**: Add NLE comments export and migrate export API to POST ([#401](https://github.com/shumaiOne/shumai/pull/401))
- **webui**: Add space key quick preview for files and folders ([#402](https://github.com/shumaiOne/shumai/pull/402))
- **webui**: Show comments count badge in file card preview area ([#403](https://github.com/shumaiOne/shumai/pull/403))
- **agent**: Use deterministic storage keys for preset avatars ([#404](https://github.com/shumaiOne/shumai/pull/404))
## [0.4.4] - 2026-09-09
### Added
- **upload**: S3 multipart upload with offline resume and wake lock ([#391](https://github.com/shumaiOne/shumai/pull/391))
- **transcode**: Optimize video operations using ffmpeg http range requests ([#393](https://github.com/shumaiOne/shumai/pull/393))
- **agent**: Enable read_asset tool for autofill agent ([#396](https://github.com/shumaiOne/shumai/pull/396))
- **sandbox**: Abstract SandboxService and hot-reload allowed domains ([#398](https://github.com/shumaiOne/shumai/pull/398))
### Fixed
- **s3**: Stream downloadToFile and putObject to prevent OOM on large files ([#392](https://github.com/shumaiOne/shumai/pull/392))
- **agent**: Pass agent thinkingLevel to session name generation ([#395](https://github.com/shumaiOne/shumai/pull/395))
- **agent**: Restrict pageCount and duration recommendations to matching media types ([#397](https://github.com/shumaiOne/shumai/pull/397))
- **transcode**: Co-locate screenshots and derived artifacts in asset storage directory ([#399](https://github.com/shumaiOne/shumai/pull/399))
## [0.4.3] - 2026-09-06
### Added
- **asset**: Record and display agent provenance on assets and audit logs ([#380](https://github.com/shumaiOne/shumai/pull/380))
- **agent**: Store preset avatars as real files in s3 storage ([#381](https://github.com/shumaiOne/shumai/pull/381))
- **agent**: Add built-in image and video generation tools and settings ([#382](https://github.com/shumaiOne/shumai/pull/382))
- **media-generation**: Add gemini-omni-1.1-flash video model and redesign settings UI ([#383](https://github.com/shumaiOne/shumai/pull/383))
- **settings**: Redesign image/video settings to provider dialog with model toggles ([#385](https://github.com/shumaiOne/shumai/pull/385))
- **agent**: Add asset management tools and CLI commands ([#387](https://github.com/shumaiOne/shumai/pull/387))
- **cli**: Control asset deletion via SHUMAI_ALLOW_DELETE env var ([#388](https://github.com/shumaiOne/shumai/pull/388))
### Fixed
- **webui**: Clamp file card creation info to 2 lines and show tooltip on hover ([#390](https://github.com/shumaiOne/shumai/pull/390))
### Documentation
- **agents**: Add documentation for image and video generation settings ([#389](https://github.com/shumaiOne/shumai/pull/389))
### Maintenance
- **deps**: Upgrade bun to 1.4.2 and update packages ([#384](https://github.com/shumaiOne/shumai/pull/384))
- **i18n**: Update paraglide-js to 2.25.0 and stop checking in generated code ([#386](https://github.com/shumaiOne/shumai/pull/386))
## [0.4.2] - 2026-09-03
### Added
- **chat**: Support markups, attachments, and timestamps in agent 1-to-1 chat ([#368](https://github.com/shumaiOne/shumai/pull/368))
- **agent**: Move thread metadata to structured context XML ([#373](https://github.com/shumaiOne/shumai/pull/373))
- **agent**: Improve timestamp precision and proxy selection for video screenshots ([#374](https://github.com/shumaiOne/shumai/pull/374))
- **agent**: Support dynamic annotationId resolution across chat and comments ([#375](https://github.com/shumaiOne/shumai/pull/375))
- **agent**: Consolidate media tools into read_asset and add S3 key support to download_asset ([#376](https://github.com/shumaiOne/shumai/pull/376))
- **provider**: Optimize provider and model configuration management ([#377](https://github.com/shumaiOne/shumai/pull/377))
- **provider**: Add provider and model sync feature with vendored generator ([#378](https://github.com/shumaiOne/shumai/pull/378))
### Fixed
- **agent**: Set empty details in asset tool results to avoid data duplication ([#370](https://github.com/shumaiOne/shumai/pull/370))
- **webui**: Strip styles and preserve multiline text on paste in chat input ([#371](https://github.com/shumaiOne/shumai/pull/371))
- **agent**: Hide main comment sessions in dashboard sessions list ([#372](https://github.com/shumaiOne/shumai/pull/372))
### Changed
- **webui**: Use semantic system colors in notification settings ([#379](https://github.com/shumaiOne/shumai/pull/379))
## [0.4.1] - 2026-08-30
### Added
- **webui**: Add manage versions option to file card context menu ([#364](https://github.com/shumaiOne/shumai/pull/364))
- **agent**: Refactor message context into unified shumai_message ([#365](https://github.com/shumaiOne/shumai/pull/365))
- **webui**: Redesign comment attachment display into compact rows ([#366](https://github.com/shumaiOne/shumai/pull/366))
### Fixed
- **asset,webui**: Disallow dropping version stacks onto files and version stacks ([#363](https://github.com/shumaiOne/shumai/pull/363))
- **webui**: Prevent empty folder flash during initial loading ([#367](https://github.com/shumaiOne/shumai/pull/367))
### Changed
- **agent**: Remove executeAgentToolWorkflow and invoke core services directly ([#362](https://github.com/shumaiOne/shumai/pull/362))
## [0.4.0] - 2026-08-28
### Added
- **transcode**: Add hardware-accelerated video transcoding support ([#352](https://github.com/shumaiOne/shumai/pull/352))
- **webui**: Truncate long file names with 100ms hover tooltip ([#355](https://github.com/shumaiOne/shumai/pull/355))
- **webui**: Add simplified mobile-friendly layout for file and share lists ([#356](https://github.com/shumaiOne/shumai/pull/356))
- **webui**: Add mobile friendly file detail and public share view ([#359](https://github.com/shumaiOne/shumai/pull/359))
### Fixed
- **transcode**: Apply team hardware acceleration and source bitrate capping to watermark transcoding ([#353](https://github.com/shumaiOne/shumai/pull/353))
- **share**: Hide and soft-delete symlinks when target assets are trashed ([#354](https://github.com/shumaiOne/shumai/pull/354))
- **webui**: Fix breadcrumb file dropdown trigger stretching and alignment ([#360](https://github.com/shumaiOne/shumai/pull/360))
- **webui**: Align mobile comment cards and thread vertical indicator ([#361](https://github.com/shumaiOne/shumai/pull/361))
### Documentation
- Add SECURITY.md with a private disclosure path ([#357](https://github.com/shumaiOne/shumai/pull/357))
## [0.3.11] - 2026-08-25
### Added
- **auth**: Add admin password reset CLI command and reset password web UI ([#347](https://github.com/shumaiOne/shumai/pull/347))
- **project**: Add recent viewed files section to project folder tree ([#348](https://github.com/shumaiOne/shumai/pull/348))
### Fixed
- **webui**: Prevent left carousel remounting and scroll jump on file selection ([#351](https://github.com/shumaiOne/shumai/pull/351))
## [0.3.10] - 2026-08-23
### Added
- **agentic kanban**: Phase 1 not-agentic kanban board ([#330](https://github.com/shumaiOne/shumai/pull/330))
- **kanban**: Resolve latest version for linked version stack assets and update demo login text ([#342](https://github.com/shumaiOne/shumai/pull/342))
- **kanban**: Update task status permissions, reporter defaults, and participant access ([#345](https://github.com/shumaiOne/shumai/pull/345))
### Fixed
- **kanban**: Use true thumbnail and poster for task asset previews ([#343](https://github.com/shumaiOne/shumai/pull/343))
- **webui**: Sidebar selection styling, kanban active state, pointer cursors, and owner dashboard ([#344](https://github.com/shumaiOne/shumai/pull/344))
- **kanban**: Improve drag-drop to column bottom, remove empty drop line, and fix timeline alignment ([#346](https://github.com/shumaiOne/shumai/pull/346))
## [0.3.9] - 2026-08-18
### Added
- **agent**: Add permission controls for chat agents ([#313](https://github.com/shumaiOne/shumai/pull/313))
- **authz**: Enforce project-aware roles for agent/skill/mcp permissions ([#314](https://github.com/shumaiOne/shumai/pull/314))
- **webui**: Use project members for asset comment mentions ([#315](https://github.com/shumaiOne/shumai/pull/315))
- **quota**: Add 2-level database-backed resource quota system ([#318](https://github.com/shumaiOne/shumai/pull/318))
- **settings**: Reorganize settings into personal, team, and AI sections ([#324](https://github.com/shumaiOne/shumai/pull/324))
- **quota**: Move usage monitoring to team dashboard ([#327](https://github.com/shumaiOne/shumai/pull/327))
### Fixed
- **webui**: Preserve selected chatbot agent preference on reload ([#323](https://github.com/shumaiOne/shumai/pull/323))
- **quota**: Exclude agents from quota rules and usage records ([#325](https://github.com/shumaiOne/shumai/pull/325))
### Documentation
- **team-settings**: Add resource quotas documentation and update readme features ([#326](https://github.com/shumaiOne/shumai/pull/326))
## [0.3.8] - 2026-08-15
### Added
- **webui**: Replace legacy markdown editor with new lexical wysiwyg editor ([#308](https://github.com/shumaiOne/shumai/pull/308))
- **agent**: Add nested agents.md support across backend and webui ([#309](https://github.com/shumaiOne/shumai/pull/309))
- **webui**: Improve AGENTS.md editor placeholder and read-only notice ([#310](https://github.com/shumaiOne/shumai/pull/310))
- **version-stack**: Add manage versions dialog and version removal ([#311](https://github.com/shumaiOne/shumai/pull/311))
### Fixed
- **core**: Rename latest version child record on version stack rename ([#312](https://github.com/shumaiOne/shumai/pull/312))
## [0.3.7] - 2026-08-14
### Added
- **metadata**: Replace aiAutofill boolean with autofillSource enum ([#302](https://github.com/shumaiOne/shumai/pull/302))
- **agent**: Type-safe CREATION_CONTEXT metadata on create_file/create_version ([#303](https://github.com/shumaiOne/shumai/pull/303))
- **sandbox**: Default Network Sandbox to off and add settings toggle switch ([#304](https://github.com/shumaiOne/shumai/pull/304))
- **webui**: Redesign select metadata fields and improve long text field widget ([#305](https://github.com/shumaiOne/shumai/pull/305))
- **agent**: Support dynamic newOption in select and selectMulti metadata fields ([#306](https://github.com/shumaiOne/shumai/pull/306))
### Fixed
- **webui**: Fix select field height and multi-select overflow/interaction issues ([#307](https://github.com/shumaiOne/shumai/pull/307))
## [0.3.6] - 2026-08-12
### Added
- **mcp**: Add MCP server support with proxy tool and OAuth flow ([#289](https://github.com/shumaiOne/shumai/pull/289))
- **mcp**: Support tool-level directTools array and 3-state UI configuration ([#297](https://github.com/shumaiOne/shumai/pull/297))
- **webui**: Direct canvas PDF rendering with selectable text ([#299](https://github.com/shumaiOne/shumai/pull/299))
### Fixed
- **agent**: Prevent agents from accessing disabled skills ([#292](https://github.com/shumaiOne/shumai/pull/292))
- **agent**: Resolve version stack name and size in list_assets ([#296](https://github.com/shumaiOne/shumai/pull/296))
- **webui**: Content-only pinch-zoom and two-finger pan in media viewers ([#298](https://github.com/shumaiOne/shumai/pull/298))
- **i18n**: Update Chinese translation for dashboard to 仪表板 ([#301](https://github.com/shumaiOne/shumai/pull/301))
### Documentation
- **mcp**: Add MCP support documentation to README and docs ([#300](https://github.com/shumaiOne/shumai/pull/300))
## [0.3.5] - 2026-08-09
### Added
- **transcode**: Add Gotenberg support for Office, HTML, CSV, and Markdown PDF proxies ([#283](https://github.com/shumaiOne/shumai/pull/283))
### Fixed
- **webui**: Resolve upload card freeze for unsupported files ([#284](https://github.com/shumaiOne/shumai/pull/284))
- **webui**: Exclude file extension when selecting file name during inline rename ([#285](https://github.com/shumaiOne/shumai/pull/285))
- Use renamed asset name in download Content-Disposition ([#286](https://github.com/shumaiOne/shumai/pull/286))
- **webui**: Batch download no longer flashes a tab per file ([#287](https://github.com/shumaiOne/shumai/pull/287))
- **webui**: Enable rename and delete actions in file detail top bar breadcrumb ([#288](https://github.com/shumaiOne/shumai/pull/288))
## [0.3.4] - 2026-08-08
### Added
- **sharelink**: Add sharelink watermark support ([#270](https://github.com/shumaiOne/shumai/pull/270))
- **webui**: Remove original-file display fallbacks, use proxies only ([#273](https://github.com/shumaiOne/shumai/pull/273))
- **share**: Add allow download config to share links ([#275](https://github.com/shumaiOne/shumai/pull/275))
- **transcode**: Update preview thumbnail resolution to 300p with max length constraint ([#281](https://github.com/shumaiOne/shumai/pull/281))
### Fixed
- **share**: Resolve symlink asset downloads in public share pages ([#274](https://github.com/shumaiOne/shumai/pull/274))
- **webui**: Persist project sorting and prevent full page flash ([#276](https://github.com/shumaiOne/shumai/pull/276))
### Documentation
- **share-links**: Document allow download and lazy watermark settings ([#278](https://github.com/shumaiOne/shumai/pull/278))
- **agent**: Clarify create_file/create_version context param guidance ([#282](https://github.com/shumaiOne/shumai/pull/282))
### Maintenance
- **e2e**: Watermarked download-disabled share link coverage (image + video) ([#277](https://github.com/shumaiOne/shumai/pull/277))
- **e2e**: Stabilize watermark share spec on Firefox ([#279](https://github.com/shumaiOne/shumai/pull/279))
- Split e2e scripts into app and webui targets ([#280](https://github.com/shumaiOne/shumai/pull/280))
## [0.3.3] - 2026-08-05
### Added
- **agent**: Enable LLM API call retries in AgentHarness ([#250](https://github.com/shumaiOne/shumai/pull/250))
- **audit**: Add audit log support for APIs and WebUI dashboard ([#251](https://github.com/shumaiOne/shumai/pull/251))
- **webui**: Persist agent preference in localStorage ([#254](https://github.com/shumaiOne/shumai/pull/254))
- **e2e**: Add file domain comment and draw tests ([#260](https://github.com/shumaiOne/shumai/pull/260))
- **agent**: Restrict bash access for non-owner users ([#262](https://github.com/shumaiOne/shumai/pull/262))
- **agent**: Allow create_file tool to create files directly from content ([#263](https://github.com/shumaiOne/shumai/pull/263))
- **agent**: Propagate autofill context from chat to autofill workflow ([#264](https://github.com/shumaiOne/shumai/pull/264))
- **webui**: Render select options in field creation mode and pick random option color ([#265](https://github.com/shumaiOne/shumai/pull/265))
- **project**: Include team-scoped members in project member list and add 3-section dialog layout ([#267](https://github.com/shumaiOne/shumai/pull/267))
- **webui**: Disable role dropdown for project-scoped members in team view ([#269](https://github.com/shumaiOne/shumai/pull/269))
### Fixed
- **webui**: Mention popup avatars, keyboard auto-scroll, and non-sticky section headers ([#256](https://github.com/shumaiOne/shumai/pull/256))
- **webui**: Persist field order on drag and drop ([#266](https://github.com/shumaiOne/shumai/pull/266))
### Changed
- **agent**: Load models directly from database without pi fallback ([#253](https://github.com/shumaiOne/shumai/pull/253))
- **core**: Replace generated provider files with pi-ai dynamic methods ([#255](https://github.com/shumaiOne/shumai/pull/255))
### Maintenance
- **core**: Add missing tests for methods introduced in 97069042 ([#252](https://github.com/shumaiOne/shumai/pull/252))
- **e2e**: Add domain-based auth and project web app e2e tests ([#257](https://github.com/shumaiOne/shumai/pull/257))
- **e2e**: Add file, share, and collection web app e2e tests ([#258](https://github.com/shumaiOne/shumai/pull/258))
- **e2e**: Add project-level invite and file-move e2e tests ([#259](https://github.com/shumaiOne/shumai/pull/259))
- **e2e**: Add search in file list and save-search-as-collection cases ([#261](https://github.com/shumaiOne/shumai/pull/261))
- **e2e**: Add member permission change tests ([#268](https://github.com/shumaiOne/shumai/pull/268))
## [0.3.2] - 2026-07-30
### Added
- **agent**: Add output truncation and streaming limits to sandboxed bash tool ([#248](https://github.com/shumaiOne/shumai/pull/248))
### Fixed
- **agent**: Download original document asset instead of pdf proxy ([#247](https://github.com/shumaiOne/shumai/pull/247))
- **webui**: Persist chatbot sidebar input text across route navigation ([#249](https://github.com/shumaiOne/shumai/pull/249))
## [0.3.1] - 2026-07-30
### Added
- **transcode**: Trigger ai metadata autofill post transcode ([#234](https://github.com/shumaiOne/shumai/pull/234))
- **dashboard**: Add agent sessions view for team owners ([#237](https://github.com/shumaiOne/shumai/pull/237))
- **db**: Preserve agent sessions on asset deletion by setting onDelete to SetNull ([#240](https://github.com/shumaiOne/shumai/pull/240))
- **agent**: Add download_asset tool and prompt instructions ([#242](https://github.com/shumaiOne/shumai/pull/242))
- **db**: Update Asset.sizeByte and AiUsage token fields to BigInt ([#244](https://github.com/shumaiOne/shumai/pull/244))
### Fixed
- **agent**: Include select option labels and field name in autofill schema ([#235](https://github.com/shumaiOne/shumai/pull/235))
- **agent**: Scope agent comment sessions to thread sessions anchored at root comments ([#239](https://github.com/shumaiOne/shumai/pull/239))
- **agent**: Disable tools for session naming and make naming non-blocking ([#241](https://github.com/shumaiOne/shumai/pull/241))
- **webui**: Update file size formatting to SI decimal standard ([#245](https://github.com/shumaiOne/shumai/pull/245))
- **s3**: Enforce property syntax in S3Service interface and fix ReadableStream handling in S3StorageService ([#246](https://github.com/shumaiOne/shumai/pull/246))
## [0.3.0] - 2026-07-28
### Added
- **file-browser**: Add copy name and path option to context menu ([#225](https://github.com/shumaiOne/shumai/pull/225))
- **agent**: Inject user name and role into chat context ([#226](https://github.com/shumaiOne/shumai/pull/226))
- **agent**: Refactor database session storage to normalized DAG entries and lazy thread sync ([#229](https://github.com/shumaiOne/shumai/pull/229))
- **agent**: Support ENABLE_WEAKER_NESTED_SANDBOX environment variable ([#233](https://github.com/shumaiOne/shumai/pull/233))
### Fixed
- **agent**: Ensure atomic appendEntry and persist leaf entries in DatabaseSessionStorage ([#224](https://github.com/shumaiOne/shumai/pull/224))
- Sharp version mismatch bug ([#230](https://github.com/shumaiOne/shumai/pull/230))
- **agent**: Resolve S3 MIME type detection and prompt parameter in image analysis ([#232](https://github.com/shumaiOne/shumai/pull/232))
### Changed
- **agent**: Remove implicit agent mention trigger in comment threads ([#227](https://github.com/shumaiOne/shumai/pull/227))
### Maintenance
- **workflow**: Add docker compose deployment verification job ([#231](https://github.com/shumaiOne/shumai/pull/231))
## [0.2.3] - 2026-07-25
### Fixed
- **webui**: Show right sidebar by default on file detail page ([#219](https://github.com/shumaiOne/shumai/pull/219))
- **webui**: Refresh file list and detail when agent executes asset creation tools ([#220](https://github.com/shumaiOne/shumai/pull/220))
- **webui**: Fix space key drag overlay and shift range selection for version stacks ([#221](https://github.com/shumaiOne/shumai/pull/221))
- **transcode**: Handle timestamp rounding tolerance for screenshot annotations ([#223](https://github.com/shumaiOne/shumai/pull/223))
### Changed
- **e2e**: Split transcode workflow E2E tests into per-workflow test files ([#222](https://github.com/shumaiOne/shumai/pull/222))
## [0.2.2] - 2026-07-23
### Added
- **webui**: Integrate AI usage API into dashboard page ([#216](https://github.com/shumaiOne/shumai/pull/216))
- **ai-usage**: Add team and member token usage tracking and APIs ([#215](https://github.com/shumaiOne/shumai/pull/215))
- **transcode**: Add PSD support via ImageMagick and enforce sRGB conversion ([#217](https://github.com/shumaiOne/shumai/pull/217))
### Fixed
- **search**: Fix buildSqlCondition date range, numeric metadata, and string parsing bugs ([#213](https://github.com/shumaiOne/shumai/pull/213))
- **webui**: Prevent horizontal overflow in search dialog result rows ([#214](https://github.com/shumaiOne/shumai/pull/214))
- **webui**: Invalidate upload tasks query when upload task starts and completes ([#218](https://github.com/shumaiOne/shumai/pull/218))
## [0.2.1] - 2026-07-22
### Added
- **skill**: Add team owner skill permission control ([#209](https://github.com/shumaiOne/shumai/pull/209))
### Fixed
- **agent**: Handle non-builtin models in getModelFromDb and upgrade pi-ai ([#211](https://github.com/shumaiOne/shumai/pull/211))
- **provider**: Preserve model ids on provider update ([#212](https://github.com/shumaiOne/shumai/pull/212))
### Changed
- **core**: Split generated providers into per-provider files ([#210](https://github.com/shumaiOne/shumai/pull/210))
## [0.2.0] - 2026-07-21
### Added
- **transcode**: Add PDF preview, sprite generation, and annotation support ([#193](https://github.com/shumaiOne/shumai/pull/193))
- **web**: Register pdf.worker route in production ([#195](https://github.com/shumaiOne/shumai/pull/195))
- **agent**: Add read_pdf_pages tool and PDF page context support ([#196](https://github.com/shumaiOne/shumai/pull/196))
- **transcode**: Split monolithic transcodeMedia into focused workflows ([#197](https://github.com/shumaiOne/shumai/pull/197))
- **transcode**: Add PDF proxy support for TXT and CSV files ([#199](https://github.com/shumaiOne/shumai/pull/199))
- **asset**: Replace mediaType/mimeType logic with proxyType ([#200](https://github.com/shumaiOne/shumai/pull/200))
- **docker**: Install fonts-noto-cjk in runner stage ([#202](https://github.com/shumaiOne/shumai/pull/202))
- **transcode**: Add markdown file to pdf proxy support ([#203](https://github.com/shumaiOne/shumai/pull/203))
- **chat**: Track context asset changes and update agent context ([#204](https://github.com/shumaiOne/shumai/pull/204))
- **agent**: Refactor media tools to static injection with mandatory assetId and authz ([#208](https://github.com/shumaiOne/shumai/pull/208))
### Fixed
- **transcode**: Mark pdfkit as external and extract for docker runtime ([#201](https://github.com/shumaiOne/shumai/pull/201))
- **agent**: Include skill id in formatSkillsForPrompt ([#205](https://github.com/shumaiOne/shumai/pull/205))
- **asset**: Exclude soft-deleted children from latestChildren preview ([#206](https://github.com/shumaiOne/shumai/pull/206))
- **agent**: Pass latest version asset from version stack to prompt builder ([#207](https://github.com/shumaiOne/shumai/pull/207))
## [0.1.10] - 2026-07-16
### Added
- **workflow-e2e**: Add E2E tests for agentAutofillMedia workflow ([#188](https://github.com/shumaiOne/shumai/pull/188))
- **comments**: Support comment deletion and video player shortcuts ([#191](https://github.com/shumaiOne/shumai/pull/191))
- **agent**: Support disabling specific tools for agents ([#192](https://github.com/shumaiOne/shumai/pull/192))
### Maintenance
- **e2e**: Add E2E workflow tests for remaining agent and transcode services ([#189](https://github.com/shumaiOne/shumai/pull/189))
## [0.1.9] - 2026-07-14
### Added
- **e2e**: Set up fullstack app-level e2e test framework and signup test case ([#184](https://github.com/shumaiOne/shumai/pull/184))
- **chat**: Chatbot UI ([#185](https://github.com/shumaiOne/shumai/pull/185))
- **agent**: Support aborting agent execution ([#186](https://github.com/shumaiOne/shumai/pull/186))
### Changed
- **api/chat**: Remove direct Prisma query from API layer ([#187](https://github.com/shumaiOne/shumai/pull/187))
## [0.1.8] - 2026-07-10
### Added
- **chatbot**: Implement type-safe Chat APIs with compaction support ([#180](https://github.com/shumaiOne/shumai/pull/180))
### Fixed
- **webui**: Enable build splitting and code-split all page routes ([#183](https://github.com/shumaiOne/shumai/pull/183))
### Documentation
- **readme**: Clarify displayed file size represents original file size ([#182](https://github.com/shumaiOne/shumai/pull/182))
## [0.1.7] - 2026-07-06
### Added
- **audio**: Support frame-accurate audio playback, comments, and AI safeguards ([#175](https://github.com/shumaiOne/shumai/pull/175))
- **i18n**: Localize members dialog, notifications, uploads, user menus, dates, and players ([#178](https://github.com/shumaiOne/shumai/pull/178))
- **comments**: Allow marking comments as complete or incomplete ([#179](https://github.com/shumaiOne/shumai/pull/179))
### Fixed
- **upload**: Resolve empty and generic MIME types in confirmFileUpload ([#176](https://github.com/shumaiOne/shumai/pull/176))
### Changed
- **webui**: Modularize file viewer components into registry-driven architecture ([#174](https://github.com/shumaiOne/shumai/pull/174))
### Documentation
- Improve documentation and examples for S3 compatible storage ([#177](https://github.com/shumaiOne/shumai/pull/177))
## [0.1.6] - 2026-07-04
### Added
- **file-browser**: Show video duration on file card preview ([#170](https://github.com/shumaiOne/shumai/pull/170))
- **webui**: Compare versions view for version stacks ([#171](https://github.com/shumaiOne/shumai/pull/171))
### Fixed
- **webui**: Resolve conditional hooks and setState-in-render errors ([#167](https://github.com/shumaiOne/shumai/pull/167))
- **webui**: Resolve video player total frames mismatch ([#172](https://github.com/shumaiOne/shumai/pull/172))
- **version-stack**: Resolve correct latest version ID of version stack ([#173](https://github.com/shumaiOne/shumai/pull/173))
### Maintenance
- **webui**: Video player e2e framework + Safari playhead fix ([#166](https://github.com/shumaiOne/shumai/pull/166))
- **webui**: Add frame-accurate comment seek roundtrip e2e ([#168](https://github.com/shumaiOne/shumai/pull/168))
- **webui**: Verify video player restarts playback after end ([#169](https://github.com/shumaiOne/shumai/pull/169))
## [0.1.5] - 2026-07-02
### Changed
- **storage**: Use original filenames for stored files ([#165](https://github.com/shumaiOne/shumai/pull/165))
### Maintenance
- **release**: Add Aliyun ACR push to release workflow ([#164](https://github.com/shumaiOne/shumai/pull/164))
## [0.1.4] - 2026-07-01
### Added
- **transcode**: Change video poster resolution to 480p ([#161](https://github.com/shumaiOne/shumai/pull/161))
- **auth**: Allow setting user locale on signup ([#163](https://github.com/shumaiOne/shumai/pull/163))
### Fixed
- **webui**: Preserve video ended state by skipping playhead snap on ended ([#162](https://github.com/shumaiOne/shumai/pull/162))
## [0.1.3] - 2026-06-30
### Added
- **webui**: Add i18n support using paraglide-js ([#156](https://github.com/shumaiOne/shumai/pull/156))
- **video**: Add frame-accurate video player, SMPTE timecodes, and format switcher ([#157](https://github.com/shumaiOne/shumai/pull/157))
### Fixed
- **webui**: Align video timeline duration and formatting with frameio ([#158](https://github.com/shumaiOne/shumai/pull/158))
- **webui**: Fix missing left sidebar for nested folders ([#159](https://github.com/shumaiOne/shumai/pull/159))
- **api**: Fix update team settings and make request schema type-safe ([#160](https://github.com/shumaiOne/shumai/pull/160))
## [0.1.2] - 2026-06-25
### Changed
- **s3**: Remove automatic port appending to AWS_ENDPOINT_URL_S3 ([#155](https://github.com/shumaiOne/shumai/pull/155))
## [0.1.1] - 2026-06-23
### Added
- **docker**: Add root entrypoint with privilege dropping for data volumes ([#151](https://github.com/shumaiOne/shumai/pull/151))
## [0.1.0] - 2026-06-22
### Added
- **share**: Secure share info API and fix wrong password feedback ([#150](https://github.com/shumaiOne/shumai/pull/150))
### Fixed
- **webui**: Truncate long folder, collection, and share link names in sidebar ([#149](https://github.com/shumaiOne/shumai/pull/149))
## [0.0.11] - 2026-06-22
### Added
- **cli**: Add shumai-cli and token authentication ([#144](https://github.com/shumaiOne/shumai/pull/144))
- **settings**: Use singular headers and remove collapse/counters for single agent types ([#145](https://github.com/shumaiOne/shumai/pull/145))
- **folders**: Auto expand folder tree in left sidebar on navigation ([#146](https://github.com/shumaiOne/shumai/pull/146))
- **metadata**: Add single user and multiple user custom fields ([#147](https://github.com/shumaiOne/shumai/pull/147))
### Maintenance
- **cli**: Rename npm package to @shumai-one/shumai-cli ([#148](https://github.com/shumaiOne/shumai/pull/148))
## [0.0.10] - 2026-06-21
### Changed
- **web**: Dynamically register production HTML bundle routes ([#142](https://github.com/shumaiOne/shumai/pull/142))
### Documentation
- **readme**: Polish structure, correct grammar, and add dependencies table ([#143](https://github.com/shumaiOne/shumai/pull/143))
## [0.0.9] - 2026-06-20
### Added
- **webui**: Show app version in avatar menu ([#139](https://github.com/shumaiOne/shumai/pull/139))
- **cli**: Add background daemonization and subcommands ([#141](https://github.com/shumaiOne/shumai/pull/141))
### Maintenance
- **release**: Add smoke tests to local release stage ([#140](https://github.com/shumaiOne/shumai/pull/140))
## [0.0.8] - 2026-06-19
### Added
- **api**: Implement read-only mode for reviewers ([#122](https://github.com/shumaiOne/shumai/pull/122))
- **server**: Allow customizing server port via SHUMAI_SERVER_PORT ([#124](https://github.com/shumaiOne/shumai/pull/124))
- **embedding**: Defer embedding workflow, use transcoded assets, and execute chunks in separate activities ([#125](https://github.com/shumaiOne/shumai/pull/125))
- **webui**: Add external drag-and-drop file and folder upload support ([#126](https://github.com/shumaiOne/shumai/pull/126))
- **agent**: Implement overlapping video chunks for semantic search ([#127](https://github.com/shumaiOne/shumai/pull/127))
- **webui**: Support jumping to video timestamp from search results ([#128](https://github.com/shumaiOne/shumai/pull/128))
- **auth**: Add demo mode support ([#131](https://github.com/shumaiOne/shumai/pull/131))
- **auth**: Refine demo mode support ([#132](https://github.com/shumaiOne/shumai/pull/132))
- **webui**: Default right sidebar to collapsed for new users ([#133](https://github.com/shumaiOne/shumai/pull/133))
- **webui**: Use Shadcn Switch in toggle field and Rating component in search filter panel ([#137](https://github.com/shumaiOne/shumai/pull/137))
- **upload**: Track byte-level progress with XMLHttpRequest ([#138](https://github.com/shumaiOne/shumai/pull/138))
### Fixed
- **transcode**: Handle high-precision FPS using rational fractions ([#123](https://github.com/shumaiOne/shumai/pull/123))
- **db**: Automatically update project updatedAt on asset modification ([#129](https://github.com/shumaiOne/shumai/pull/129))
- **webui**: Use system colors for fields and improve search filter select options ([#130](https://github.com/shumaiOne/shumai/pull/130))
- **search**: Fix hasAny and hasNone operators on metadata filter ([#134](https://github.com/shumaiOne/shumai/pull/134))
- **webui**: Resolve folder tree sync, search dialog reset, and sidebar actions issues ([#136](https://github.com/shumaiOne/shumai/pull/136))
### Changed
- **webui**: Fix search dialog layout height, scroll areas, and restrict filter toggle click ([#135](https://github.com/shumaiOne/shumai/pull/135))
## [0.0.7] - 2026-06-17
### Added
- **workflow**: Configure transcode and agent concurrency via environment variables ([#116](https://github.com/shumaiOne/shumai/pull/116))
- **webui**: Replace hardcoded colors with system tokens ([#119](https://github.com/shumaiOne/shumai/pull/119))
- **webui**: Replace native overflow with shadcn ScrollArea in settings ([#120](https://github.com/shumaiOne/shumai/pull/120))
- **api**: Use enum for API protocol in provider config ([#121](https://github.com/shumaiOne/shumai/pull/121))
### Fixed
- **webui**: Reload video player and reset state when switching assets ([#114](https://github.com/shumaiOne/shumai/pull/114))
### Changed
- **transcode**: Relocate TranscodeService to core and flatten API package ([#115](https://github.com/shumaiOne/shumai/pull/115))
### Documentation
- Improve and restructure shumai documentation ([#118](https://github.com/shumaiOne/shumai/pull/118))
## [0.0.6] - 2026-06-15
### Added
- **sandbox**: Add pending domains support for network sandbox ([#109](https://github.com/shumaiOne/shumai/pull/109))
- **webui**: Always show right sidebar and implement left vertical carousel ([#111](https://github.com/shumaiOne/shumai/pull/111))
### Fixed
- **build**: Propagate termination signals to active child process in wrappers ([#102](https://github.com/shumaiOne/shumai/pull/102))
- **transcode**: Make deterministic transcode activity failures non-retryable in Temporal ([#104](https://github.com/shumaiOne/shumai/pull/104))
- **player**: Implement seekbar dragging and match hover highlight color ([#108](https://github.com/shumaiOne/shumai/pull/108))
- **webui**: Avoid refetching all pages on file/folder reorder ([#112](https://github.com/shumaiOne/shumai/pull/112))
- **webui**: Force video-player mime type to video/mp4 ([#113](https://github.com/shumaiOne/shumai/pull/113))
### Changed
- **webui**: Use system color design tokens in video player components ([#106](https://github.com/shumaiOne/shumai/pull/106))
- **webui**: Use requestAnimationFrame for smooth video progress bar updates ([#107](https://github.com/shumaiOne/shumai/pull/107))
- **webui**: Fetch agents in mention input list dynamically ([#110](https://github.com/shumaiOne/shumai/pull/110))
### Documentation
- **readme**: Add comprehensive README.md ([#103](https://github.com/shumaiOne/shumai/pull/103))
- **readme**: Improve remote deployment and temporal execution instructions ([#105](https://github.com/shumaiOne/shumai/pull/105))
## [0.0.5] - 2026-06-13
### Added
- **npm**: Auto deploy database migrations on package start ([#101](https://github.com/shumaiOne/shumai/pull/101))
## [0.0.4] - 2026-06-12
### Fixed
- **cli/docker**: Resolve global CLI external dependencies and fix Docker 404 error ([#100](https://github.com/shumaiOne/shumai/pull/100))
## [0.0.3] - 2026-06-12
### Fixed
- **cli**: Resolve platform binary path resolution under nested node_modules ([#99](https://github.com/shumaiOne/shumai/pull/99))
## [0.0.2] - 2026-06-12
### Fixed
- **npm**: Add repository metadata to package.json for provenance verification ([#98](https://github.com/shumaiOne/shumai/pull/98))
### Maintenance
- Add release-npm.yaml workflow ([#97](https://github.com/shumaiOne/shumai/pull/97))
## [0.0.1] - 2026-06-12
### Added
- Initial release of Shumai
+18 -7
View File
@@ -30,11 +30,7 @@ RUN bun run scripts/extract-runtime-deps.ts
# Stage 2: Runner
FROM oven/bun:1.4.2 AS runner
# Install ffmpeg/ffprobe
COPY --from=mwader/static-ffmpeg:8.1 /ffmpeg /usr/local/bin/
COPY --from=mwader/static-ffmpeg:8.1 /ffprobe /usr/local/bin/
# Install system dependencies for sandbox runtime and agent
# Install system dependencies and Jellyfin-FFmpeg (with VA-API support)
RUN apt-get update && apt-get install -y --no-install-recommends \
python3 \
make \
@@ -56,11 +52,26 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
poppler-utils \
imagemagick \
fonts-noto-cjk \
libraw-bin \
libimage-exiftool-perl \
&& . /etc/os-release \
&& TARGETARCH=$(dpkg --print-architecture) \
&& FFMPEG_VERSION="8.1.2-5" \
&& curl -fsSL -o /tmp/jellyfin-ffmpeg.deb "https://github.com/jellyfin/jellyfin-ffmpeg/releases/download/v${FFMPEG_VERSION}/jellyfin-ffmpeg8_${FFMPEG_VERSION}-${VERSION_CODENAME}_${TARGETARCH}.deb" \
&& apt-get install -y --no-install-recommends /tmp/jellyfin-ffmpeg.deb \
&& rm -f /tmp/jellyfin-ffmpeg.deb \
&& ldconfig /usr/lib/jellyfin-ffmpeg/lib \
&& ln -sf /usr/lib/jellyfin-ffmpeg/ffmpeg /usr/local/bin/ffmpeg \
&& ln -sf /usr/lib/jellyfin-ffmpeg/ffprobe /usr/local/bin/ffprobe \
&& ln -sf /usr/lib/jellyfin-ffmpeg/ffmpeg /usr/bin/ffmpeg \
&& ln -sf /usr/lib/jellyfin-ffmpeg/ffprobe /usr/bin/ffprobe \
&& apt-get clean \
&& rm -rf /var/lib/apt/lists/*
# Set up working directory and make it owned by the bun user
RUN mkdir -p /app/data && chown -R bun:bun /app
# Set up working directory and make it owned by the bun user, ensure video/render group access
RUN usermod -a -G video,render bun 2>/dev/null || true \
&& mkdir -p /app/data \
&& chown -R bun:bun /app
# Switch to the non-root bun user
USER bun
+42 -189
View File
@@ -33,7 +33,9 @@ Shumai is an open-source Frame.io alternative built for creative teams to collab
* **Custom Asset Metadata**: Define dynamic metadata fields tailored to your production pipeline.
* **Share Link Watermarking**: Protect shared media with customizable text and image watermarks, and save watermark as reusable templates.
* **Wide File Format Support**: Preview common creative file formats, including images (with PSD support), videos, audio, and documents, while allowing any file type to be uploaded and stored.
* **3D Asset Turntable Previews**: Generate interactive 3D model turntable previews and videos for formats like GLB, FBX, OBJ, STL, USD, and DAE using [turntable-renderer](https://docs.shumai.one/using-shumai/3d-asset-support).
* **Office & Document PDF Proxying via Gotenberg**: Convert Word, PowerPoint, Excel, Markdown, HTML, and CSV documents to high-quality PDF proxies for in-browser previewing using [Gotenberg](https://docs.shumai.one/using-shumai/gotenberg).
* **Semantic Search**: Find assets by visual or conceptual meaning using vector embeddings. See [documentation](https://docs.shumai.one/using-shumai/semantic-search).
* **NLE Export & Premiere Pro Plugin (Experimental)**: Export timestamped review comments to your favorite non-linear editor (Premiere Pro XML, DaVinci Resolve EDL, Final Cut Pro, Avid Media Composer) or use the experimental Premiere Pro UXP plugin to import media assets and auto-sync comments directly to sequence timeline markers. See [documentation](https://docs.shumai.one/using-shumai/nle-and-premiere).
But Shumai goes beyond being a Frame.io alternative. With modern models like GPT-Image-2 and Seedance 2.5, AI can already handle a surprising amount of creative work. That means you can create or update videos and images from a simple prompt without having to jump back into Photoshop, Premiere or Houdini. That is why Shumai includes a powerful agent system:
@@ -46,7 +48,6 @@ But Shumai goes beyond being a Frame.io alternative. With modern models like GPT
* **Model Context Protocol (MCP)**: Connect agents to external tools and APIs through remote MCP servers, with OAuth authentication and per-agent MCP access.
* **Isolated Sandbox Execution**: Run agent-submitted scripts safely in a sandboxed environment.
* **AI Metadata Autofill**: Automatically fill in custom metadata for new assets.
* **Semantic Search**: Find assets by visual or conceptual meaning using vector embeddings (Gemini Embedding 2).
* **Resource Quotas**: Set usage limits on AI tokens, costs, and tool execution per member, role, or team-wide, with real-time tracking.
Here is a short demo video showing what the Shumai agent can do:
@@ -82,10 +83,25 @@ Docker Compose is the fastest way to get Shumai running. You do not need to clon
```bash
mkdir shumai && cd shumai
```
2. Download the `docker-compose.yaml` file:
```bash
curl -o docker-compose.yaml https://raw.githubusercontent.com/shumaiOne/shumai/main/docker-compose/local/docker-compose.yaml
```
2. Download your preferred `docker-compose.yaml` configuration based on your needs and available RAM:
* **Lite (Suggested RAM: 2 GB+)**: Minimal deployment containing Shumai and PostgreSQL database.
```bash
curl -o docker-compose.yaml https://raw.githubusercontent.com/shumaiOne/shumai/main/docker-compose/local/docker-compose.yaml
```
* **With Embedding Server (Suggested RAM: 4 GB+)**: Adds local multimodal semantic search powered by `shumaiOne/embedding-server` (`EmbeddingGemma 2 740M`).
```bash
curl -o docker-compose.yaml https://raw.githubusercontent.com/shumaiOne/shumai/main/docker-compose/local/docker-compose.embedding.yaml
```
* **Full (Suggested RAM: 8 GB+)**: All-in-one setup including `embedding-server`, `turntable-renderer` (3D previews), and `Gotenberg` (document PDF proxying).
```bash
curl -o docker-compose.yaml https://raw.githubusercontent.com/shumaiOne/shumai/main/docker-compose/local/docker-compose.full.yaml
```
> [!NOTE]
> By default, the Docker Compose setups use CPU software transcoding. If your server has a GPU (NVIDIA NVENC, Intel QuickSync, AMD/Intel VA-API, or Rockchip RKMPP) and you want to enable GPU hardware acceleration, check the [Hardware Transcoding Guide](https://docs.shumai.one/configuration/hardware-transcoding).
3. Configure environment variables (optional):
* `SHUMAI_SERVER_PORT` controls the port the Shumai server listens on. The default is `3000`.
@@ -125,188 +141,7 @@ Docker Compose is the fastest way to get Shumai running. You do not need to clon
```
6. Open your browser and access Shumai at `http://localhost:3000` (or `http://<your-server-ip>:3000` for remote deployments).
### Option 2: Install via NPM / Package Manager
Shumai is published as `@shumai-one/shumai` on NPM. This option allows you to run Shumai globally or locally.
#### Step 1: Start PostgreSQL with pgvector
Shumai requires PostgreSQL with the `pgvector` extension. Start a pre-configured database container using Docker:
```bash
docker run --name shumai_postgres \
-e POSTGRES_USER=shumai \
-e POSTGRES_PASSWORD=shumai_password \
-e POSTGRES_DB=shumai_db \
-p 5432:5432 \
-d pgvector/pgvector:pg18
```
#### Step 2: Create a workspace folder
Create a dedicated directory to store your environment configuration and media files (which are saved in a `./data` directory by default):
```bash
mkdir shumai && cd shumai
```
#### Step 3: Install Platform-Specific Dependencies
#### Linux
Shumai requires the following system packages on Linux:
- **`ffmpeg`** – Used for media transcoding and metadata extraction.
- **`poppler`** (`poppler-utils`) – Used for PDF page image extraction and PDF sprite preview generation (`pdftoppm`).
- **`imagemagick`** – Used for PSD format transcoding and color profile conversion to sRGB.
- **`bubblewrap`**, **`socat`**, and **`ripgrep`** – Required by the AI agent sandbox (`anthropic-experimental/sandbox-runtime`) for process isolation, networking, and workspace search.
Install all required packages with one command:
> Ubuntu/Debian
```bash
sudo apt install -y ffmpeg poppler-utils imagemagick bubblewrap socat ripgrep
```
> Fedora
```bash
sudo dnf install -y ffmpeg poppler-utils ImageMagick bubblewrap socat ripgrep
```
> Arch Linux
```bash
sudo pacman -S --noconfirm ffmpeg poppler imagemagick bubblewrap socat ripgrep
```
> [!NOTE]
> **Ubuntu 24.04+** restricts unprivileged user namespaces by default. This prevents `anthropic-experimental/sandbox-runtime` from using `bubblewrap` for sandbox isolation.
>
> To temporarily allow user namespaces:
>
> ```bash
> sudo sysctl -w kernel.apparmor_restrict_unprivileged_userns=0
> ```
>
> Alternatively, configure an AppArmor profile that grants the required `userns` permission to the relevant binaries.
---
#### macOS
Shumai requires:
- **`ffmpeg`** – Used for media transcoding and metadata extraction.
- **`poppler`** – Used for PDF page image extraction and PDF sprite preview generation (`pdftoppm`).
- **`imagemagick`** – Used for PSD format transcoding and color profile conversion to sRGB.
- **`ripgrep`** – Required by the AI agent sandbox (`anthropic-experimental/sandbox-runtime`).
Install the required packages with Homebrew:
```bash
brew install ffmpeg poppler imagemagick ripgrep
```
---
#### Windows (alpha)
> [!WARNING]
> Windows support in anthropic-experimental/sandbox-runtime is **alpha and the design is in flux**.
>
> The current implementation provides:
> - Network egress filtering via the **Windows Filtering Platform (WFP)**.
> - File read/write denial via **ACL stamping**.
>
> However, it is **not a security boundary against a deliberately adversarial sandboxed process**.
>
> The sandbox implementation will be substantially revised in a future release, where sandboxed processes will run under a dedicated sandbox user account.
Shumai requires:
- **`ffmpeg`** – Used for media transcoding and metadata extraction.
- **`imagemagick`** – Used for PSD format transcoding and color profile conversion to sRGB.
Install required packages using your preferred package manager:
**winget**
```powershell
winget install Gyan.FFmpeg ImageMagick.ImageMagick
```
**Chocolatey**
```powershell
choco install ffmpeg imagemagick
```
#### Step 4: Install Shumai globally
Install Shumai globally using your preferred package manager:
```bash
# NPM
npm install -g @shumai-one/shumai
# PNPM
pnpm add -g @shumai-one/shumai
# Bun
bun add -g @shumai-one/shumai
```
#### Step 5: Configure Environment Variables
Create a `.env` file in your workspace folder (`shumai/`) and add the following configuration:
```env
DATABASE_URL=postgresql://shumai:shumai_password@localhost:5432/shumai_db?schema=public
BETTER_AUTH_SECRET=ySxs7DxzHDZBbeeHNPEwBuspYwipBqz5Gk5XdBjNhWw=
STORAGE_BACKEND=local
SHUMAI_SERVER_PORT=3000
AWS_ENDPOINT_URL_S3=http://localhost:3000
# Optional: base URL for MCP OAuth redirects (defaults to BETTER_AUTH_URL)
MCP_OAUTH_REDIRECT_BASE_URL=http://localhost:3000
```
> [!NOTE]
> `SHUMAI_SERVER_PORT` sets the port the server starts on, while `AWS_ENDPOINT_URL_S3` is used by the browser to build file upload URLs.
> If deploying on a remote server (e.g. `http://123.456.7.8`) with a mapped port (e.g. `12345:3000` in docker-compose), set `AWS_ENDPOINT_URL_S3` to `http://123.456.7.8:12345`.
> `MCP_OAUTH_REDIRECT_BASE_URL` is only needed when MCP servers use OAuth: the authorization redirect URI defaults to `{BETTER_AUTH_URL}/api/mcp/oauth/callback`; set it to an externally reachable URL if the app is served behind a proxy.
#### Step 6: Run Shumai
> [!IMPORTANT]
> **For Bun Users:** Since the published package binaries contain a `#!/usr/bin/env node` shebang, running the global command directly (e.g., `shumai`) will execute under Node.js. If you installed with Bun and want to run under the Bun runtime, you must prefix all commands with `bun run --bun` (e.g., `bun run --bun shumai`, `bun run --bun shumai -d`, `bun run --bun shumai stop`).
Start the application from your workspace folder:
```bash
shumai
```
Alternatively, you can run and manage Shumai in daemon mode:
- **Start in daemon mode**:
```bash
shumai -d
```
- **Stop Shumai**:
```bash
shumai stop
```
- **Restart Shumai**:
```bash
shumai restart
```
- **Show/tail logs**:
```bash
shumai logs
```
On startup, Shumai will automatically run database migrations and start the web server at `http://localhost:3000`.
---
### Option 3: Run from Source (Development)
### Option 2: Run from Source (Development)
To set up Shumai locally for development:
@@ -316,8 +151,23 @@ To set up Shumai locally for development:
cd shumai
bun install
```
2. Start the `pgvector` database container (as described in **Option 2, Step 1**).
3. Create a `.env` file at the root of the workspace using the configuration from **Option 2, Step 5**.
2. Start the `pgvector` database container:
```bash
docker run --name shumai_postgres \
-e POSTGRES_USER=shumai \
-e POSTGRES_PASSWORD=shumai_password \
-e POSTGRES_DB=shumai_db \
-p 5432:5432 \
-d pgvector/pgvector:pg18
```
3. Create a `.env` file at the root of the workspace:
```env
DATABASE_URL=postgresql://shumai:shumai_password@localhost:5432/shumai_db?schema=public
BETTER_AUTH_SECRET=ySxs7DxzHDZBbeeHNPEwBuspYwipBqz5Gk5XdBjNhWw=
STORAGE_BACKEND=local
SHUMAI_SERVER_PORT=3000
AWS_ENDPOINT_URL_S3=http://localhost:3000
```
4. Generate i18n translation files:
```bash
bun run i18n:compile
@@ -333,6 +183,9 @@ To set up Shumai locally for development:
bun run dev
```
> [!NOTE]
> Looking to install Shumai globally via NPM or your package manager? See our [NPM Installation Guide](https://docs.shumai.one/getting-started/quickstart).
---
## Command Line Interface (CLI)
+58 -219
View File
@@ -33,7 +33,9 @@ Shumai 是一个开源的 Frame.io 替代方案,专为创意团队打造,让
* **自定义素材元数据**:可以按你的制作流程,自由定义动态元数据字段。
* **分享链接水印**:通过可自定义的文字和图片水印保护共享媒体,并将水印配置保存为模板以便重复使用。
* **丰富的文件格式支持**:支持预览常见创意文件格式,包括图片(支持 PSD)、视频、音频和文档,同时可上传和存储任意类型的文件。
* **基于转盘渲染服务的 3D 资产预览**:通过集成 [turntable-renderer](https://docs.shumai.one/using-shumai/3d-asset-support) 支持为 GLB、FBX、OBJ、STL、USD 和 DAE 等 3D 模型生成高质量的交互式转盘预览与视频。
* **基于 Gotenberg 的文档 PDF 转码代理**:通过集成 [Gotenberg](https://docs.shumai.one/using-shumai/gotenberg) 支持将 Word、PowerPoint、Excel、Markdown、HTML 和 CSV 文档高质量转换为 PDF 代理文件进行在线预览。
* **语义搜索**:支持按画面内容或概念含义搜索素材,底层基于向量嵌入技术。详见[相关文档](https://docs.shumai.one/using-shumai/semantic-search)。
* **非编剪辑导出与 Premiere 插件(实验性)**:支持将时间戳审阅评论导出至常用非编软件(Premiere Pro XML、DaVinci Resolve EDL、Final Cut Pro、Avid Media Composer),或使用实验性的 Premiere Pro 原生 UXP 插件导入媒体素材并将评论自动同步为时间轴标记。详见[相关文档](https://docs.shumai.one/using-shumai/nle-and-premiere)。
但 Shumai 不仅仅只满足于 Frame.io 的开源替代。随着 GPT-Image-2、Seedance 2.5 这类强大模型的问世,AI已经生成产品级的图片/视频。很多时候,你只要给一句提示词,就能直接生成或修改图片和视频,而不用再打开 Photoshop、Premiere 或者 Houdini。也正因为这样,Shumai 才把 AI Agent 做成了核心能力:
@@ -44,7 +46,6 @@ Shumai 是一个开源的 Frame.io 替代方案,专为创意团队打造,让
* **Model Context Protocol (MCP) 支持**:通过远程 MCP 服务端为 Agent 接入外部工具与 API,支持 OAuth 鉴权与按 Agent 独立分配访问权限。
* **沙箱隔离执行**:Agent 生成的脚本会在隔离环境里运行,更安全。
* **AI 自动补全元数据**:新素材上传后,系统可以自动帮你补一些元数据。
* **语义搜索**:支持按画面内容或语义找素材,底层用的是向量 embedding(Gemini Embedding 2)。
* **资源配额**:支持按成员、角色或全团队对 AI Token、成本和工具调用设置使用上限与实时用量监控。
下面是一个简短的 demo 视频,展示了 Shumai Agent 能做什么:
@@ -98,11 +99,25 @@ https://gh-proxy.com/https://raw.githubusercontent.com/shumaiOne/shumai/main/REA
```
2. 下载 `docker-compose.yaml` 配置文件:
```bash
curl -o docker-compose.yaml https://raw.githubusercontent.com/shumaiOne/shumai/main/docker-compose/local/docker-compose-aliyun.yaml
2. 根据您的需求和机器可用内存,下载对应的 `docker-compose.yaml` 配置文件:
```
* **精简版(建议内存:2 GB+)**:包含 Shumai 核心应用与 PostgreSQL 数据库。
```bash
curl -o docker-compose.yaml https://raw.githubusercontent.com/shumaiOne/shumai/main/docker-compose/local/docker-compose-aliyun.yaml
```
* **含嵌入向量服务版(建议内存:4 GB+)**:新增基于 `shumaiOne/embedding-server` (`EmbeddingGemma 2 740M`) 的本地多模态语义搜索能力。
```bash
curl -o docker-compose.yaml https://raw.githubusercontent.com/shumaiOne/shumai/main/docker-compose/local/docker-compose-aliyun.embedding.yaml
```
* **完整版(建议内存:8 GB+)**:一站式全功能套件,包含 `embedding-server`、`turntable-renderer`(3D 模型交互转盘预览)与 `Gotenberg`(文档与表格 PDF 代理转码)。
```bash
curl -o docker-compose.yaml https://raw.githubusercontent.com/shumaiOne/shumai/main/docker-compose/local/docker-compose-aliyun.full.yaml
```
> [!NOTE]
> 默认的 Docker Compose 配置使用基于 CPU 的软件转码。如果您的服务器配备了支持的显卡(NVIDIA NVENC、Intel QuickSync、AMD/Intel VA-API 或 Rockchip RKMPP)并需要开启 GPU 硬件加速转码,请参考 [硬件加速转码配置文档](https://docs.shumai.one/configuration/hardware-transcoding)。
3. 配置环境变量(可选):
@@ -149,231 +164,55 @@ https://gh-proxy.com/https://raw.githubusercontent.com/shumaiOne/shumai/main/REA
5. 打开浏览器,访问 `http://localhost:3000` 即可开始使用(若是远程部署,请访问 `http://<您的服务器IP>:3000`)。
### 方案二:通过 NPM / 包管理器安装
通过 NPM 上的 `@shumai-one/shumai` 进行安装。
#### 第一步:启动带 pgvector 支持的 PostgreSQL
Shumai 依赖支持 `pgvector` 扩展的 PostgreSQL 数据库。您可以使用 Docker 快速启动一个预配置好的数据库容器:
```bash
docker run --name shumai_postgres \
-e POSTGRES_USER=shumai \
-e POSTGRES_PASSWORD=shumai_password \
-e POSTGRES_DB=shumai_db \
-p 5432:5432 \
-d docker.m.daocloud.io/pgvector/pgvector:pg18
```
#### 第二步:创建工作区目录
创建一个专属目录用来存放环境配置和媒体文件(媒体文件默认会保存在 `./data` 目录下):
```bash
mkdir shumai && cd shumai
```
#### 步骤 3:安装平台依赖
#### Linux
在 Linux 上运行 Shumai 需要安装以下系统依赖:
- **`ffmpeg`** —— 用于媒体转码和元数据提取。
- **`poppler`** (`poppler-utils`) —— 用于 PDF 页面图像提取和 PDF 雪碧图预览生成 (`pdftoppm`)。
- **`imagemagick`** —— 用于 PSD 格式图像转码和 sRGB 色彩空间转换。
- **`bubblewrap`**、**`socat`** 和 **`ripgrep`** —— AI Agent 沙箱(`anthropic-experimental/sandbox-runtime`)所需,用于进程隔离、网络通信和工作区搜索。
可以使用以下命令一次性安装所有依赖:
> Ubuntu/Debian
```bash
sudo apt install -y ffmpeg poppler-utils imagemagick bubblewrap socat ripgrep
```
> Fedora
```bash
sudo dnf install -y ffmpeg poppler-utils ImageMagick bubblewrap socat ripgrep
```
> Arch Linux
```bash
sudo pacman -S --noconfirm ffmpeg poppler imagemagick bubblewrap socat ripgrep
```
> [!NOTE]
> **Ubuntu 24.04+** 默认限制非特权用户创建 User Namespace,这会导致 `anthropic-experimental/sandbox-runtime` 无法使用 `bubblewrap` 提供沙箱隔离。
>
> 如需临时启用 User Namespace,请执行:
>
> ```bash
> sudo sysctl -w kernel.apparmor_restrict_unprivileged_userns=0
> ```
>
> 或者,配置 AppArmor Profile,为相关可执行文件授予所需的 `userns` 权限。
---
#### macOS
在 macOS 上运行 Shumai 需要:
- **`ffmpeg`** —— 用于媒体转码和元数据提取。
- **`poppler`** —— 用于 PDF 页面图像提取和 PDF 雪碧图预览生成 (`pdftoppm`)。
- **`imagemagick`** —— 用于 PSD 格式图像转码和 sRGB 色彩空间转换。
- **`ripgrep`** —— AI Agent 沙箱(`anthropic-experimental/sandbox-runtime`)所需。
使用 Homebrew 安装所需依赖:
```bash
brew install ffmpeg poppler imagemagick ripgrep
```
---
#### Windows(Alpha)
> [!WARNING]
> anthropic-experimental/sandbox-runtime 的 Windows 支持目前仍处于 **Alpha** 阶段,整体设计仍在持续调整中。
>
> 当前实现提供了以下能力:
> - 通过 **Windows Filtering Platform (WFP)** 实现网络出口(egress)过滤。
> - 通过 **ACL Stamping** 实现文件读写权限限制。
>
> **但这并不能作为针对恶意沙箱进程的安全边界(security boundary)**。
>
> 未来版本将对沙箱机制进行重大调整,届时沙箱进程将运行在独立的沙箱用户账户下。
在 Windows 上运行 Shumai 需要:
- **`ffmpeg`** —— 用于媒体转码和元数据提取。
- **`imagemagick`** —— 用于 PSD 格式图像转码和 sRGB 色彩空间转换。
使用你偏好的包管理器安装所需依赖:
**winget**
```powershell
winget install Gyan.FFmpeg ImageMagick.ImageMagick
```
**Chocolatey**
```powershell
choco install ffmpeg imagemagick
```
#### 第四步:全局安装 Shumai
使用您常用的包管理器进行全局安装:
```bash
# 使用 NPM
npm install -g @shumai-one/shumai
# 使用 PNPM
pnpm add -g @shumai-one/shumai
# 使用 Bun
bun add -g @shumai-one/shumai
```
#### 第五步:配置环境配置
在刚才创建的工作区目录(`shumai/`)下新建一个 `.env` 文件,并填入以下配置:
```env
DATABASE_URL=postgresql://shumai:shumai_password@localhost:5432/shumai_db?schema=public
BETTER_AUTH_SECRET=ySxs7DxzHDZBbeeHNPEwBuspYwipBqz5Gk5XdBjNhWw=
STORAGE_BACKEND=local
SHUMAI_SERVER_PORT=3000
AWS_ENDPOINT_URL_S3=http://localhost:3000
```
> [!NOTE]
> `SHUMAI_SERVER_PORT` 用于设置 Web 服务的启动端口,而 `AWS_ENDPOINT_URL_S3` 供浏览器端构建文件上传链接。
> 如果您部署在远程服务器(如 `http://123.456.7.8`)并且做了端口映射(如 Docker 中的 `12345:3000`),请将 `AWS_ENDPOINT_URL_S3` 设置为 `http://123.456.7.8:12345`。
#### 第六步:启动 Shumai
> [!IMPORTANT]
> **Bun 用户特别提示:** 由于发布的包二进制文件包含 `#!/usr/bin/env node` 头部声明(Shebang),直接在终端执行全局命令(例如 `shumai`)默认会通过 Node.js 运行。如果您是用 Bun 安装的并且希望调用 Bun 运行时,请在所有命令前加上 `bun run --bun` 前缀(例如:`bun run --bun shumai`、`bun run --bun shumai -d`、`bun run --bun shumai stop`)。
在您的工作区目录下执行以下命令启动应用:
```bash
shumai
```
您也可以让 Shumai 在后台以守护进程(Daemon)模式运行和管理:
* **后台模式启动**:
```bash
shumai -d
```
* **停止服务**:
```bash
shumai stop
```
* **重启服务**:
```bash
shumai restart
```
* **查看并跟踪实时日志**:
```bash
shumai logs
```
启动时,Shumai 会自动执行数据库迁移(Migration)并在 `http://localhost:3000` 拉起 Web 服务。
---
### 方案三:源码编译运行(开发环境)
### 方案二:源码编译运行(开发环境)
如果您希望在本地参与 Shumai 的开发,请按以下步骤配置开发环境:
1. 克隆代码仓库并安装依赖项:
```bash
git clone [https://github.com/shumaiOne/shumai.git](https://github.com/shumaiOne/shumai.git)
cd shumai
bun install
```bash
git clone https://github.com/shumaiOne/shumai.git
cd shumai
bun install
```
```
2. 启动 `pgvector` 数据库容器:
```bash
docker run --name shumai_postgres \
-e POSTGRES_USER=shumai \
-e POSTGRES_PASSWORD=shumai_password \
-e POSTGRES_DB=shumai_db \
-p 5432:5432 \
-d docker.m.daocloud.io/pgvector/pgvector:pg18
```
3. 在项目根目录下创建一个 `.env` 文件,填入以下基础配置:
```env
DATABASE_URL=postgresql://shumai:shumai_password@localhost:5432/shumai_db?schema=public
BETTER_AUTH_SECRET=ySxs7DxzHDZBbeeHNPEwBuspYwipBqz5Gk5XdBjNhWw=
STORAGE_BACKEND=local
SHUMAI_SERVER_PORT=3000
AWS_ENDPOINT_URL_S3=http://localhost:3000
```
2. 启动 `pgvector` 数据库容器(参考**方案二中的第一步**)。
3. 参照**方案二中的第五步**,在项目根目录下创建一个 `.env` 文件。
4. 生成多语言国际化(i18n)代码:
```bash
bun run i18n:compile
```
> [!TIP]
> 在前端开发过程中,可以在单独的终端窗口中运行 `bun run i18n:watch`,在修改 `packages/webui/messages/` 下的翻译文案时自动重新编译。
```bash
bun run i18n:compile
```
> [!TIP]
> 在前端开发过程中,可以在单独的终端窗口中运行 `bun run i18n:watch`,在修改 `packages/webui/messages/` 下的翻译文案时自动重新编译。
5. 执行数据库结构迁移:
```bash
bun run db:migrate
```
```bash
bun run db:migrate
```
6. 启动本地开发服务器:
```bash
bun run dev
```
```bash
bun run dev
```
> [!NOTE]
> 如需通过 NPM 或包管理器全局安装运行 Shumai,请参考[官方 NPM 安装指南](https://docs.shumai.one/getting-started/quickstart)。
+1 -1
View File
@@ -1,6 +1,6 @@
{
"name": "@shumai/agent-app",
"version": "0.4.6",
"version": "0.5.3",
"private": true,
"type": "module",
"bin": {
-1
View File
@@ -1,4 +1,3 @@
export * from '@shumai/agent/src/workflows/agent-autofill'
export * from '@shumai/agent/src/workflows/agent-chat'
export * from '@shumai/agent/src/workflows/agent-embedding'
export * from '@shumai/agent/src/workflows/query-embedding-for-search'
+1 -1
View File
@@ -1,6 +1,6 @@
{
"name": "@shumai/cli",
"version": "0.4.6",
"version": "0.5.3",
"private": true,
"type": "module",
"bin": {
+1 -1
View File
@@ -1,6 +1,6 @@
{
"name": "@shumai/premiere-uxp",
"version": "0.4.6",
"version": "0.5.3",
"private": true,
"type": "commonjs",
"scripts": {
+1 -1
View File
@@ -1,6 +1,6 @@
{
"name": "@shumai/transcode-app",
"version": "0.4.6",
"version": "0.5.3",
"private": true,
"type": "module",
"bin": {
+4
View File
@@ -73,6 +73,8 @@ apps/web/e2e/
| Owner creates a project with a cover image | `tests/project/create-project.spec.ts` |
| Owner uploads a file or an unsupported file via the "Upload File" context menu action | `tests/project/upload-file.spec.ts` |
| Owner deletes a file, finds it in recently deleted, and restores it | `tests/project/delete-restore.spec.ts` |
| Owner deletes a folder and still sees its contents preview and countdown in recently deleted | `tests/project/delete-restore.spec.ts` |
| Owner deletes a version stack and restores it from recently deleted | `tests/project/delete-restore.spec.ts` |
| Owner creates a share link from a file context menu and lands on the share page | `tests/project/create-share-link.spec.ts` |
| Owner protects a share with a password; guests must enter it to view | `tests/project/share-password.spec.ts` |
| Owner creates a collection from the sidebar; the collection shows project files | `tests/project/create-collection.spec.ts` |
@@ -103,6 +105,8 @@ apps/web/e2e/
| Upload an image file, wait for transcode, and create a comment with draw | `tests/file/image-draw-comment.spec.ts` |
| Upload a video file, wait for transcode, and create a comment with draw at a non-zero timestamp | `tests/file/video-draw-comment.spec.ts` |
| Upload an audio file, wait for transcode, and create a comment at a non-zero timestamp | `tests/file/audio-timestamp-comment.spec.ts` |
| Show a breathing poster preview and "Preparing..." while a video is still transcoding | `tests/file/processing-preview.spec.ts` |
| Show a preparing circle before the poster is generated | `tests/file/processing-preview.spec.ts` |
## Conventions
+1
View File
@@ -19,6 +19,7 @@ async function start() {
process.env.BETTER_AUTH_URL = 'http://localhost:5200'
process.env.NODE_ENV = 'development'
process.env.SHUMAI_E2E = 'true'
process.env.ENABLE_EMBEDDING_AGENT = 'false'
// Write variables to .env.e2e file so the teardown can load them
const envContent = [`DATABASE_URL=${process.env.DATABASE_URL}`, `S3_BUCKET=${bucketName}`].join(
@@ -0,0 +1,82 @@
import { expect, test } from '../../fixtures'
import { fileCard } from '../../helpers/files'
// Uploaded as a binary file so no real transcode task races with the simulated mid-transcode
// state we set up directly in the database.
test.use({ fileOptions: { mediaType: 'binary' } })
test('shows the poster preview and "Preparing..." while a video is still transcoding', async ({
file,
prisma,
}) => {
const { page, projectId, fileId, fileName } = file
// Simulate a video that is mid-transcode: the poster/sprite preview already exists, but the
// asset is still marked as processing.
const assetDir = `files/e2e/${fileId}`
await prisma.asset.update({
where: { id: fileId },
data: {
status: 'processing',
mediaType: 'video/mp4',
media: {
duration: 10,
filesize: 0,
frames: 300,
proxyType: 'video',
imageTranscodes: [],
videoTranscodes: [],
sprite: { key: `${assetDir}/sprite.webp`, frames: 100, tileX: 10, tileY: 10 },
poster: { key: `${assetDir}/poster.webp` },
finishedAt: new Date().toISOString(),
metadata: {
originalWidth: 1920,
originalHeight: 1080,
hasAudio: false,
duration: 10,
bitRate: 0,
frameRate: 30,
totalFrames: 300,
startTimecode: '00:00:00:00',
format: {},
},
original: { key: `${assetDir}/original.mp4`, filesizeInBytes: 0, codec: '' },
},
},
})
await page.goto(`/projects/${projectId}`)
const card = fileCard(page, fileName)
await expect(card).toBeVisible()
// The preview (poster thumbnail used as the sprite scrubber base) is rendered instead of a
// blank placeholder, and its opacity breathes while transcoding is in progress.
await expect(card.locator('img[alt="Thumbnail"]')).toBeVisible()
await expect(card.getByTestId('file-card-preview-media')).toHaveClass(/animate-pulse/)
// The creator row is replaced by a "Preparing..." label while transcoding.
await expect(card.getByText(/Preparing|准备中/i)).toBeVisible()
// The duration badge is hidden until the asset is ready.
await expect(card.getByText('00:10')).toHaveCount(0)
})
test('shows a preparing circle before the poster is generated', async ({ file, prisma }) => {
const { page, projectId, fileId, fileName } = file
// Simulate an asset whose upload finished but whose poster does not exist yet.
await prisma.asset.update({
where: { id: fileId },
data: { status: 'processing', mediaType: 'video/mp4' },
})
await page.goto(`/projects/${projectId}`)
const card = fileCard(page, fileName)
await expect(card).toBeVisible()
// A size-breathing circle stands in until the poster is available, with a "Preparing..." label.
await expect(card.getByTestId('file-card-preparing-circle')).toBeVisible()
await expect(card.getByText(/Preparing|准备中/i)).toBeVisible()
})
@@ -1,4 +1,5 @@
import { expect, test } from '../../fixtures'
import { seedFile, seedFolder } from '../../helpers/assets'
import { fileCard } from '../../helpers/files'
test('owner deletes a file, finds it in recently deleted, and restores it', async ({
@@ -43,3 +44,121 @@ test('owner deletes a file, finds it in recently deleted, and restores it', asyn
await page.goto(`/projects/${projectId}`)
await expect(fileCard(page, fileName)).toBeVisible()
})
test('owner deletes a folder and still sees its contents preview and countdown in recently deleted', async ({
project,
prisma,
}) => {
const { page, projectId, rootFolderId } = project
const folderName = `e2e-deleted-folder-${Date.now()}`
const folder = await seedFolder(prisma, projectId, rootFolderId, folderName)
await seedFile(prisma, projectId, folder.id, 'inside-1.png')
await seedFile(prisma, projectId, folder.id, 'inside-2.png')
await page.goto(`/projects/${projectId}`)
await expect(fileCard(page, folderName)).toBeVisible()
// Delete the folder through the context menu + confirmation dialog
await fileCard(page, folderName).click({ button: 'right' })
await page.getByRole('menuitem', { name: 'Delete' }).click()
const deleteDialog = page.getByRole('alertdialog')
await expect(deleteDialog.getByText('Delete Asset?')).toBeVisible()
await deleteDialog.getByRole('button', { name: 'Delete' }).click()
await expect(fileCard(page, folderName)).not.toBeVisible()
// Navigate to the Recently Deleted view
await page.getByText('Recently Deleted').click()
await expect(page).toHaveURL(/\/recently-deleted/)
const deletedFolderCard = fileCard(page, folderName)
await expect(deletedFolderCard).toBeVisible()
// The folder still shows previews of the files that were deleted with it
await expect(deletedFolderCard.getByTestId('folder-preview-grid')).toBeVisible()
// And each item shows how many days are left before permanent deletion
await expect(deletedFolderCard.getByTestId('folder-card-days-left')).toHaveText(/30\s*(d|天)/)
})
test('owner deletes a version stack and restores it from recently deleted', async ({
project,
prisma,
}) => {
const { page, projectId, rootFolderId } = project
const latestVersionName = `e2e-stack-latest-${Date.now()}.png`
const olderVersionName = `e2e-stack-older-${Date.now()}.png`
const stack = await prisma.asset.create({
data: {
name: `e2e-stack-${Date.now()}`,
type: 'version_stack',
status: 'processed',
projectId,
parentId: rootFolderId,
fileCount: 2,
sizeByte: 20,
},
})
await prisma.asset.create({
data: {
name: latestVersionName,
type: 'file',
status: 'processed',
sizeByte: 10,
sortIndex: 'a0',
projectId,
parentId: stack.id,
},
})
await prisma.asset.create({
data: {
name: olderVersionName,
type: 'file',
status: 'processed',
sizeByte: 10,
sortIndex: 'a1',
projectId,
parentId: stack.id,
},
})
// The stack card is labelled with its latest version's name
await page.goto(`/projects/${projectId}`)
await expect(fileCard(page, latestVersionName)).toBeVisible()
await fileCard(page, latestVersionName).click({ button: 'right' })
await page.getByRole('menuitem', { name: 'Delete' }).click()
const deleteDialog = page.getByRole('alertdialog')
await expect(deleteDialog.getByText('Delete Asset?')).toBeVisible()
await deleteDialog.getByRole('button', { name: 'Delete' }).click()
await expect(fileCard(page, latestVersionName)).not.toBeVisible()
// The whole stack is soft-deleted, including its versions
await expect
.poll(async () => {
const row = await prisma.asset.findUnique({ where: { id: stack.id } })
return row?.isDeleted
})
.toBe(true)
// Restore it from the Recently Deleted view
await page.getByText('Recently Deleted').click()
await expect(page).toHaveURL(/\/recently-deleted/)
await expect(fileCard(page, latestVersionName)).toBeVisible()
await fileCard(page, latestVersionName).click({ button: 'right' })
await page.getByRole('menuitem', { name: 'Restore' }).click()
await expect(fileCard(page, latestVersionName)).not.toBeVisible()
await expect
.poll(async () => {
const row = await prisma.asset.findUnique({ where: { id: stack.id } })
return row?.isDeleted
})
.toBe(false)
// Back at the project root the stack is visible again
await page.goto(`/projects/${projectId}`)
await expect(fileCard(page, latestVersionName)).toBeVisible()
})
+1 -1
View File
@@ -1,6 +1,6 @@
{
"name": "@shumai/web-app",
"version": "0.4.6",
"version": "0.5.3",
"private": true,
"type": "module",
"dependencies": {
+28
View File
@@ -12,10 +12,13 @@ import { metadataService } from '@shumai/core/src/metadata/metadata'
import { initTranscodeWorkflows } from '@shumai/transcode'
import { workflowService } from '@shumai/workflow-core'
import { migrateLegacyAgentAvatars } from '@shumai/core/src/agent/migration'
import { agentService } from '@shumai/core/src/agent/agent'
import { handleDaemonCommands } from '@shumai/core/src/utils/daemon'
import { authService } from '@shumai/core/src/auth/auth'
import { embeddingService } from '@shumai/core/src/search/embedding-service'
import { sandboxService } from '@shumai/core'
import { notificationJobService } from '@shumai/core/src/notification/notification-job'
if (process.argv.includes('--check')) {
console.log('✅ Web app evaluated successfully!')
@@ -46,6 +49,29 @@ if (resetCmdIndex !== -1) {
}
}
const regenCmdIndex = cliArgs.findIndex((arg) => arg === 'regenerate-embeddings')
if (regenCmdIndex !== -1) {
try {
console.log('🔄 Regenerating embeddings...')
const result = await embeddingService.regenerateEmbeddings()
console.log(
`✅ Successfully queued ${result.tasksCreated} embedding tasks for ${result.totalAssets} assets.`,
)
if (result.skippedTeams.length > 0) {
console.warn(
`⚠️ Skipped ${result.skippedTeams.length} team(s) without an active embedding agent.`,
)
}
console.log('The background workers will process the embedding tasks.')
process.exit(0)
} catch (err: unknown) {
const message = err instanceof Error ? err.message : 'Unknown error'
console.error(`Error: ${message}`)
process.exit(1)
}
}
async function run() {
// Initialize workflows and activities for local executor mode
initAgentWorkflows()
@@ -57,7 +83,9 @@ async function run() {
// Start services
await metadataService.syncSystemFields().catch(console.error)
await migrateLegacyAgentAvatars().catch(console.error)
await agentService.ensureSystemEmbeddingAgents().catch(console.error)
assetService.startCleanupJob()
notificationJobService.start()
workflowService.start()
if (process.env.WORKFLOW_EXECUTOR === 'temporal') {
const args = process.argv.slice(2)
+76 -53
View File
@@ -33,7 +33,7 @@
},
"apps/agent": {
"name": "@shumai/agent-app",
"version": "0.4.6",
"version": "0.5.3",
"bin": {
"shumai-agent": "./src/index.ts",
},
@@ -45,7 +45,7 @@
},
"apps/cli": {
"name": "@shumai/cli",
"version": "0.4.6",
"version": "0.5.3",
"bin": {
"shumai-cli": "./src/index.ts",
},
@@ -60,7 +60,7 @@
},
"apps/premiere-uxp": {
"name": "@shumai/premiere-uxp",
"version": "0.4.6",
"version": "0.5.3",
"dependencies": {
"@shumai/api": "workspace:*",
"@shumai/dtos": "workspace:*",
@@ -114,7 +114,7 @@
},
"apps/transcode": {
"name": "@shumai/transcode-app",
"version": "0.4.6",
"version": "0.5.3",
"bin": {
"shumai-transcode": "./src/index.ts",
},
@@ -126,7 +126,7 @@
},
"apps/web": {
"name": "@shumai/web-app",
"version": "0.4.6",
"version": "0.5.3",
"dependencies": {
"@shumai/agent": "workspace:*",
"@shumai/api": "workspace:*",
@@ -146,12 +146,12 @@
},
"packages/agent": {
"name": "@shumai/agent",
"version": "0.4.6",
"version": "0.5.3",
"dependencies": {
"@anthropic-ai/sandbox-runtime": "0.0.74",
"@earendil-works/pi-agent-core": "0.82.1",
"@earendil-works/pi-ai": "0.82.1",
"@google/genai": "2.19.0",
"@google/genai": "2.28.0",
"@shumai/core": "workspace:*",
"@shumai/db": "workspace:*",
"@shumai/dtos": "workspace:*",
@@ -161,6 +161,7 @@
"@temporalio/workflow": "1.23.0",
"adm-zip": "0.6.0",
"jittered-fractional-indexing": "1.0.1",
"sharp": "0.35.4",
"typebox": "1.3.22",
"ulid": "3.0.2",
"zod": "4.5.4",
@@ -171,7 +172,7 @@
},
"packages/api": {
"name": "@shumai/api",
"version": "0.4.6",
"version": "0.5.3",
"dependencies": {
"@hono/zod-validator": "0.7.6",
"@shumai/core": "workspace:*",
@@ -185,7 +186,7 @@
},
"packages/core": {
"name": "@shumai/core",
"version": "0.4.6",
"version": "0.5.3",
"dependencies": {
"@ai-sdk/alibaba": "2.0.38",
"@ai-sdk/amazon-bedrock": "5.0.67",
@@ -210,10 +211,13 @@
"@aws-sdk/s3-request-presigner": "3.1121.0",
"@earendil-works/pi-agent-core": "0.82.1",
"@earendil-works/pi-ai": "0.82.1",
"@google/genai": "2.28.0",
"@modelcontextprotocol/client": "2.0.0",
"@shumai/db": "workspace:*",
"@shumai/dtos": "workspace:*",
"@sinclair/typebox": "0.34.52",
"@temporalio/activity": "1.24.0",
"@temporalio/workflow": "1.24.0",
"adm-zip": "0.6.0",
"ai": "7.0.84",
"better-auth": "1.6.11",
@@ -221,9 +225,11 @@
"date-fns": "4.4.0",
"dotenv": "17.4.2",
"dotenv-expand": "13.0.0",
"exiftool-vendored": "38.1.0",
"extract-zip": "2.0.1",
"hono": "4.13.5",
"jittered-fractional-indexing": "1.0.1",
"nodemailer": "10.0.10",
"pdfkit": "0.20.1",
"pino": "10.3.1",
"sharp": "0.35.4",
@@ -237,12 +243,13 @@
"@modelcontextprotocol/server": "2.0.0",
"@types/adm-zip": "0.5.8",
"@types/extract-zip": "2.0.3",
"@types/nodemailer": "8.0.2",
"@types/pdfkit": "0.17.6",
},
},
"packages/db": {
"name": "@shumai/db",
"version": "0.4.6",
"version": "0.5.3",
"dependencies": {
"@earendil-works/pi-agent-core": "0.82.1",
"@prisma/adapter-pg": "7.10.0",
@@ -258,7 +265,7 @@
},
"packages/dtos": {
"name": "@shumai/dtos",
"version": "0.4.6",
"version": "0.5.3",
"dependencies": {
"@earendil-works/pi-agent-core": "0.82.1",
"zod": "4.5.4",
@@ -266,10 +273,10 @@
},
"packages/e2e": {
"name": "@shumai/e2e",
"version": "0.4.6",
"version": "0.5.3",
"dependencies": {
"@earendil-works/pi-agent-core": "0.82.1",
"@google/genai": "2.19.0",
"@google/genai": "2.28.0",
"@shumai/agent": "workspace:*",
"@shumai/core": "workspace:*",
"@shumai/db": "workspace:*",
@@ -281,11 +288,11 @@
},
"packages/tableprinter": {
"name": "@shumai/tableprinter",
"version": "0.4.6",
"version": "0.5.3",
},
"packages/transcode": {
"name": "@shumai/transcode",
"version": "0.4.6",
"version": "0.5.3",
"dependencies": {
"@shumai/core": "workspace:*",
"@shumai/db": "workspace:*",
@@ -299,7 +306,7 @@
},
"packages/webui": {
"name": "@shumai/webui",
"version": "0.4.6",
"version": "0.5.3",
"dependencies": {
"@base-ui/react": "1.7.0",
"@dnd-kit/abstract": "0.5.0",
@@ -359,6 +366,7 @@
"cmdk": "1.1.1",
"date-fns": "4.4.0",
"embla-carousel-react": "8.6.0",
"hls.js": "1.7.3",
"hono": "4.13.5",
"input-otp": "1.5.0",
"konva": "10.3.2",
@@ -382,7 +390,6 @@
"ulid": "3.0.2",
"use-image": "1.1.4",
"vaul": "1.1.2",
"video.js": "8.24.0",
"zod": "4.5.4",
"zustand": "5.0.15",
},
@@ -401,7 +408,7 @@
},
"packages/workflow-core": {
"name": "@shumai/workflow-core",
"version": "0.4.6",
"version": "0.5.3",
"dependencies": {
"@shumai/core": "workspace:*",
"@shumai/db": "workspace:*",
@@ -647,7 +654,7 @@
"@floating-ui/utils": ["@floating-ui/utils@0.2.12", "", {}, "sha512-HpCo8tmWzLVad5s2d19EhAz5zqrrQ6s69qd6moPMQvkOuSwDT1YgRfWSVuc4ennqrgv3OHppiOGMQ7oC13yIww=="],
"@google/genai": ["@google/genai@2.19.0", "", { "dependencies": { "google-auth-library": "^10.3.0", "p-retry": "^4.6.2", "protobufjs": "^7.5.4", "ws": "^8.18.0" }, "peerDependencies": { "@modelcontextprotocol/sdk": "^1.25.2" }, "optionalPeers": ["@modelcontextprotocol/sdk"] }, "sha512-IM3u1FjpOLyOPsPh0smGA6Ntj2olUWCPbTs/WQj7cA/2AKCGw/652W7IaySx0XtwuscFnMYBoraebQusSZ0zSQ=="],
"@google/genai": ["@google/genai@2.28.0", "", { "dependencies": { "google-auth-library": "^10.3.0", "p-retry": "^4.6.2", "protobufjs": "^7.5.4", "ws": "^8.18.0" }, "peerDependencies": { "@modelcontextprotocol/sdk": "^1.25.2" }, "optionalPeers": ["@modelcontextprotocol/sdk"] }, "sha512-hyNAJj+ywu3L46qlZod6KO/JvcT3QONz6VCiPevC/FWOjAD+LQb7DMLUoV8vzoLpt61Q0ruQX99iqbxEjPJWmw=="],
"@grpc/grpc-js": ["@grpc/grpc-js@1.14.4", "", { "dependencies": { "@grpc/proto-loader": "^0.8.0", "@js-sdsl/ordered-map": "^4.4.2" } }, "sha512-k9Dj3DV/itK9D06Y8f190Qgop7/Ui+D0njFV3LHMPwPT75DpXLQohE9Wmz0QElrJnzsjB7KPWiKJbOl7IPDArQ=="],
@@ -907,6 +914,8 @@
"@oxc-project/types": ["@oxc-project/types@0.147.0", "", {}, "sha512-IJ3s6ltHLp45S0bh7phkX+gJO7A1Wuz2EaqpAhb8WjqDwbzMiWKHhyyT42tskaWjEYXtHtVCPpnBJVT9+dcRLg=="],
"@photostructure/tz-lookup": ["@photostructure/tz-lookup@11.7.0", "", {}, "sha512-QbKHXvleluMIlAHf2JiQTJNU5rLQBeQhLobyRqgXZ6NNGKqNikaRkjYviJ4nMU7Y75QE7bBYM8xjiFGjAukyhw=="],
"@pinojs/redact": ["@pinojs/redact@0.4.0", "", {}, "sha512-k2ENnmBugE/rzQfEcdWHcCY+/FM3VLzH9cYEsbdsoqrvzAKRhUZeRNhAZvB8OitQJ1TBed3yqWtdjzS6wJKBwg=="],
"@pkgjs/parseargs": ["@pkgjs/parseargs@0.11.0", "", {}, "sha512-+1VkjdD0QBLPodGrJUeqarH8VAIvQODIbwh9XpP5Syisf7YoQgsJKPNFoqqLQlu+VQ/tVSshMR6loPMn8U+dPg=="],
@@ -1557,12 +1566,16 @@
"@types/lodash": ["@types/lodash@4.17.25", "", {}, "sha512-+K1NIO8I+F9/wNulfVvu23QYd0Pe9/OCqRrim4NoYIf1VoEDL90Ve4ClzpyqBLc7NpGGWRvYNCKZ1BE/Jpf8dQ=="],
"@types/luxon": ["@types/luxon@3.7.5", "", {}, "sha512-jJ41Q4z6ZVO260MNDdHfW7+7a5iMiX8Mr6ZJHcmgrvhZha6dz5704o/lF2kKl6URjH6ivEL97w9xS/MgpJEphg=="],
"@types/mdast": ["@types/mdast@4.0.4", "", { "dependencies": { "@types/unist": "*" } }, "sha512-kGaNbPh1k7AFzgpud/gMdvIm5xuECykRR+JnWKQno9TAXVa6WIVCGTPvYGekIDL4uwCZQSYbUxNBSb1aUo79oA=="],
"@types/ms": ["@types/ms@2.1.0", "", {}, "sha512-GsCCIZDE/p3i96vtEqx+7dBUGXrc7zeSK3wwPHIaRThS+9OhWIXRqzs4d6k1SVU8g91DrNRWxWUGhp5KXQb2VA=="],
"@types/node": ["@types/node@25.9.1", "", { "dependencies": { "undici-types": ">=7.24.0 <7.24.7" } }, "sha512-xfrlY7UD5rMJk3ZVJP8BNzS28J36YJg+xp+LPXV1TdWxr8uMH5A860QNxYDGQe/ylDSgjxE52Q9VnO7p75tJxg=="],
"@types/nodemailer": ["@types/nodemailer@8.0.2", "", { "dependencies": { "@types/node": "*" } }, "sha512-c7M5ox8p0nEOfbJ2E9Qcmt8/QCu/VzsiAhzBiZbvky2PhKZDHpKB3sQHBM5MEZNkCfBOaS6S9//AGPRTNB7IaA=="],
"@types/pdfkit": ["@types/pdfkit@0.17.6", "", { "dependencies": { "@types/node": "*" } }, "sha512-tIwzxk2uWKp0Cq9JIluQXJid77lYhF52EsIOwhsMF4iWLA6YneoBR1xVKYYdAysHuepUB0OX4tdwMiUDdGKmig=="],
"@types/pg": ["@types/pg@8.23.1", "", { "dependencies": { "@types/node": "*", "pg-protocol": "*", "pg-types": "^2.2.0" } }, "sha512-fKVHpikPdg4GKks3JuLEhvwSyvwzF23hnabPy6DD8ljVbC7+6J5dQzdv4arV6jqq57djnMgs1HKBxX4P8aBI3A=="],
@@ -1621,12 +1634,6 @@
"@vercel/oidc": ["@vercel/oidc@3.2.0", "", {}, "sha512-UycprH3T6n3jH0k44NHMa7pnFHGu/N05MjojYr+Mc6I7obkoLIJujSWwin1pCvdy/eOxrI/l3uDLQsmcrOb4ug=="],
"@videojs/http-streaming": ["@videojs/http-streaming@3.17.5", "", { "dependencies": { "@babel/runtime": "^7.12.5", "@videojs/vhs-utils": "^4.1.2", "aes-decrypter": "^4.0.2", "global": "^4.4.0", "m3u8-parser": "^7.2.0", "mpd-parser": "^1.4.0", "mux.js": "7.1.0", "video.js": "^7 || ^8" } }, "sha512-PRRQVu0hYW5h0XvRWpIyIpEatpMZzZtf7L1avV89Hnu+69f+0c7oLsSL9jwpg2nbvzGkdX5rQxQkz5Vc7tFR9w=="],
"@videojs/vhs-utils": ["@videojs/vhs-utils@4.1.2", "", { "dependencies": { "@babel/runtime": "^7.12.5", "global": "^4.4.0" } }, "sha512-1Dsv5dtWOq0LP1G37O46FQFA2tKAZON5UMe6dFPxT7JvPhEx8QIXcFQxan8+hHI3fX6aomFRFrUBJNgRYQDSXA=="],
"@videojs/xhr": ["@videojs/xhr@2.7.0", "", { "dependencies": { "@babel/runtime": "^7.5.5", "global": "~4.4.0", "is-function": "^1.0.1" } }, "sha512-giab+EVRanChIupZK7gXjHy90y3nncA2phIOyG3Ne5fvpiMJzvqYwiTOnEVW2S4CoYcuKJkomat7bMXA/UoUZQ=="],
"@visx/curve": ["@visx/curve@4.0.1-alpha.0", "", { "dependencies": { "@visx/vendor": "4.0.0-alpha.0" } }, "sha512-jRu61Uz274pV1zyioXmboyrLutYbnKsgjj4njSGCnhdXj5GkZvZbg+ThDb6oOzoAnJOBRLz4rzPlWvNJOzuVMg=="],
"@visx/event": ["@visx/event@4.0.1-alpha.0", "", { "dependencies": { "@types/react": "*", "@visx/point": "4.0.1-alpha.0" } }, "sha512-EQqCMSv/s8NbFjo+hz3FKsvvYfP+2QslsFJ/24/O5l/W+7UC6J6aAvO0ujVwrTwdYbuQ+vhxKi1xdPdKR/qj1g=="],
@@ -1691,8 +1698,6 @@
"@workflow/serde": ["@workflow/serde@4.1.0", "", {}, "sha512-pav4F2BoirECWR7Nf1TKt+2eETcBj7jj4cBefQ8VXQCA6NPkaKeLfj/zMgi+3zYV5ZIBT4GuUiphsj0/b9hPQQ=="],
"@xmldom/xmldom": ["@xmldom/xmldom@0.8.15", "", {}, "sha512-/5NV/vDALVFDXgLmfsy9TRCBlKwO2LNBFzpzvb9iIj+jR+eSc6DLYYvVOdivT/jm7MtU6TebYuRmzEOI7w40UA=="],
"@xtuc/ieee754": ["@xtuc/ieee754@1.2.0", "", {}, "sha512-DX8nKgqcGwsc0eJSqYt5lwP4DH5FlHnmuWWBRy7X0NcaGR0ZtuyeESgMwTYVEtxmsNGY+qit4QYT/MIYTOTPeA=="],
"@xtuc/long": ["@xtuc/long@4.2.2", "", {}, "sha512-NuHqBY1PB/D8xU6s/thBgOAiAP7HOYDQ32+BFZILJ8ivkUkAHQnWfn6WhL79Owj1qmUnoN/YPhktdIoucipkAQ=="],
@@ -1705,8 +1710,6 @@
"adm-zip": ["adm-zip@0.6.0", "", {}, "sha512-XleryMhbuksdKtofnWZ9Sk+4CUTbms4Mb/EU32SZwToAyZ5RgVos/ki8n+yr0LWHOGKuakbXTuuYNHLQjhddgg=="],
"aes-decrypter": ["aes-decrypter@4.0.2", "", { "dependencies": { "@babel/runtime": "^7.12.5", "@videojs/vhs-utils": "^4.1.1", "global": "^4.4.0", "pkcs7": "^1.0.4" } }, "sha512-lc+/9s6iJvuaRe5qDlMTpCFjnwpkeOXp8qP3oiZ5jsj1MRg+SBVUmmICrhxHvc8OELSmc+fEyyxAuppY6hrWzw=="],
"agent-base": ["agent-base@7.1.4", "", {}, "sha512-MnA+YT8fwfJPgBx3m60MNqakm30XOkyIoH1y6huTQvC0PwZG7ki8NacLBcrPbNoo8vEZy7Jpuk7+jMO+CUovTQ=="],
"ai": ["ai@7.0.84", "", { "dependencies": { "@ai-sdk/gateway": "4.0.68", "@ai-sdk/provider": "4.0.8", "@ai-sdk/provider-utils": "5.0.33" }, "peerDependencies": { "zod": "^3.25.76 || ^4.1.8" } }, "sha512-znHNBWADOy79k8EK7636kDm5LmnjF4tMzegC2TfmlLDobDZbs0z0OXD2sxtH3eDTp5Re+pJQESvjinWJ6+kadQ=="],
@@ -1789,6 +1792,8 @@
"baseline-browser-mapping": ["baseline-browser-mapping@2.11.20", "", { "bin": { "baseline-browser-mapping": "dist/cli.cjs" } }, "sha512-H0ulySigv6icDJ1F7SjtdCD6PrhTpdYCmP0CactWy1+ekh0AFd0o1Wn5T8b+hnTmdBx19u9yhL6wvCylXMY7zw=="],
"batch-cluster": ["batch-cluster@19.2.0", "", {}, "sha512-mvozaN/7TYxl+lmXDYH+HocCcRMvW25QmghrJux2y93o+NhHPQ4wir0nBS1p8oTlinjc/l5LaNX+VFxNISrqgQ=="],
"bcrypt-pbkdf": ["bcrypt-pbkdf@1.0.2", "", { "dependencies": { "tweetnacl": "^0.14.3" } }, "sha512-qeFIXtP4MSoi6NLqO12WfqARWWuCKi2Rn/9hJLEmtB5yTNr9DqFWkJRCf2qShWzPeAMRnOgCrq0sg/KLv5ES9w=="],
"better-auth": ["better-auth@1.6.11", "", { "dependencies": { "@better-auth/core": "1.6.11", "@better-auth/drizzle-adapter": "1.6.11", "@better-auth/kysely-adapter": "1.6.11", "@better-auth/memory-adapter": "1.6.11", "@better-auth/mongo-adapter": "1.6.11", "@better-auth/prisma-adapter": "1.6.11", "@better-auth/telemetry": "1.6.11", "@better-auth/utils": "0.4.0", "@better-fetch/fetch": "1.1.21", "@noble/ciphers": "^2.1.1", "@noble/hashes": "^2.0.1", "better-call": "1.3.5", "defu": "^6.1.4", "jose": "^6.1.3", "kysely": "^0.28.17", "nanostores": "^1.1.1", "zod": "^4.3.6" }, "peerDependencies": { "@lynx-js/react": "*", "@prisma/client": "^5.0.0 || ^6.0.0 || ^7.0.0", "@sveltejs/kit": "^2.0.0", "@tanstack/react-start": "^1.0.0", "@tanstack/solid-start": "^1.0.0", "better-sqlite3": "^12.0.0", "drizzle-kit": ">=0.31.4", "drizzle-orm": "^0.45.2", "mongodb": "^6.0.0 || ^7.0.0", "mysql2": "^3.0.0", "next": "^14.0.0 || ^15.0.0 || ^16.0.0", "pg": "^8.0.0", "prisma": "^5.0.0 || ^6.0.0 || ^7.0.0", "react": "^18.0.0 || ^19.0.0", "react-dom": "^18.0.0 || ^19.0.0", "solid-js": "^1.0.0", "svelte": "^4.0.0 || ^5.0.0", "vitest": "^2.0.0 || ^3.0.0 || ^4.0.0", "vue": "^3.0.0" }, "optionalPeers": ["@lynx-js/react", "@prisma/client", "@sveltejs/kit", "@tanstack/react-start", "@tanstack/solid-start", "better-sqlite3", "drizzle-kit", "drizzle-orm", "mongodb", "mysql2", "next", "pg", "prisma", "react", "react-dom", "solid-js", "svelte", "vitest", "vue"] }, "sha512-Wwt6+q07dwIhsp6XiM7L1qSXVUWBEtNl+eZvwM778CguFqDZFBN9Pt6LtFaHl55t8Z+Zc//5kxcbgDY8/79vFQ=="],
@@ -1995,8 +2000,6 @@
"dom-accessibility-api": ["dom-accessibility-api@0.5.16", "", {}, "sha512-X7BJ2yElsnOJ30pZF4uIIDfBEVgF4XEBxL9Bxhy6dnrm5hkzqmsWHGTiHqRiITNhMyFLyAiWndIJP7Z1NTteDg=="],
"dom-walk": ["dom-walk@0.1.2", "", {}, "sha512-6QvTW9mrGeIegrFXdtQi9pk7O/nSK6lSdXW2eqUspN5LWD7UTji2Fqw5V2YLjBpHEoU9Xl/eUWNpDeZvoyOv2w=="],
"dotenv": ["dotenv@17.4.2", "", {}, "sha512-nI4U3TottKAcAD9LLud4Cb7b2QztQMUEfHbvhTH09bqXTxnSie8WnjPALV/WMCrJZ6UV/qHJ6L03OqO3LcdYZw=="],
"dotenv-expand": ["dotenv-expand@13.0.0", "", { "dependencies": { "dotenv": "^17.4.2" } }, "sha512-aBfBS8eYIeXmpHI9ThIlA7/WLq+SLt18iXUZhb52rW89QLKQFoIpPG1bPeewoPZsTyjSSO3T7234FBVUM1V2rA=="],
@@ -2099,6 +2102,12 @@
"eventsource-parser": ["eventsource-parser@3.1.1", "", {}, "sha512-EKN1vKAMcZ8MlYMpaNuxN6R9yakzH6uajHcHVTqWJzvu5pWw9DyhbP35HH8MVBQ+dZjAfDxk+A8NiR9KWaXiyQ=="],
"exiftool-vendored": ["exiftool-vendored@38.1.0", "", { "dependencies": { "@photostructure/tz-lookup": "^11.7.0", "@types/luxon": "^3.7.5", "batch-cluster": "^19.2.0", "he": "^1.2.0", "luxon": "^3.7.2" }, "optionalDependencies": { "exiftool-vendored.exe": "13.59.2", "exiftool-vendored.pl": "13.59.2" } }, "sha512-LIERqM/Qb0oqJlSswLFKWsxGVH81skp7omsX2BPKA/dEzfLP5syNO5KL7k5UlMCf+m0jqEXrKTQr/xV5vScPow=="],
"exiftool-vendored.exe": ["exiftool-vendored.exe@13.59.2", "", { "os": "win32" }, "sha512-U0l0CApIGO+AjpFYCiNCApQsMdlf6sqCyEl4wlKKqvwVAEVcvMRw91k7zj32OPFZE+/+XlLFsc6876MpOqBLiw=="],
"exiftool-vendored.pl": ["exiftool-vendored.pl@13.59.2", "", { "os": "!win32", "bin": { "exiftool": "bin/exiftool" } }, "sha512-5B80JT9phZSUsc7zz77rQ6DfW0YssTpJ5qW/1bOH1N8XQq/jcn+9eRW9HFwhVT9uoIliw9Fg007ST3pUzQIzyw=="],
"expect-type": ["expect-type@1.4.0", "", {}, "sha512-KfYbmpRm0VbLjEvVa9yGwCi9GI34xvi7A/HXYWQO65CSD2u3MczUJSuwXKFIxlGsgBQizV9q5J9NHj4VG0n+pA=="],
"exsolve": ["exsolve@1.1.1", "", {}, "sha512-9U/jZUgjnSGyntRr6y5Muu1MJcwFl6kPu7k8qLF0IMNfLqvw0NZ4nnVDq0RVoZ0RvCyumib4Ez3KYrVfilrw+g=="],
@@ -2207,8 +2216,6 @@
"glob-to-regex.js": ["glob-to-regex.js@1.2.0", "", { "peerDependencies": { "tslib": "2" } }, "sha512-QMwlOQKU/IzqMUOAZWubUOT8Qft+Y0KQWnX9nK3ch0CJg0tTp4TvGZsTfudYKv2NzoQSyPcnA6TYeIQ3jGichQ=="],
"global": ["global@4.4.0", "", { "dependencies": { "min-document": "^2.19.0", "process": "^0.11.10" } }, "sha512-wv/LAoHdRE3BeTGz53FAamhGlPLhlssK45usmGFThIi4XqnBmjKQ16u+RNbP7WvigRZDxUsM0J3gcQ5yicaL0w=="],
"globals": ["globals@17.11.0", "", {}, "sha512-Z2I8hM+PbJDXQDq3Icgpzv+mPdwr68iZUU9d5WW4FuXfDUQfkZaZuvjMv42/5crNyw154+9+VWXbYrUgDXbxNw=="],
"globalthis": ["globalthis@1.0.4", "", { "dependencies": { "define-properties": "^1.2.1", "gopd": "^1.0.1" } }, "sha512-DpLKbNU4WylpxJykQujfCcwYWiV/Jhm50Goo0wrVILAv5jOr9d+H+UR3PhSCD2rCCEIg0uc+G+muBTwD54JhDQ=="],
@@ -2247,10 +2254,14 @@
"hast-util-whitespace": ["hast-util-whitespace@3.0.0", "", { "dependencies": { "@types/hast": "^3.0.0" } }, "sha512-88JUN06ipLwsnv+dVn+OIYOvAuvBMy/Qoi6O7mQHxdPXpjy+Cd6xRkWwux7DKO+4sYILtLBRIKgsdpS2gQc7qw=="],
"he": ["he@1.2.0", "", { "bin": { "he": "bin/he" } }, "sha512-F/1DnUGPopORZi0ni+CvrCgHQ5FyEAHRLSApuYWMmrbSwoN2Mn/7k+Gl38gJnR7yyDZk6WLXwiGod1JOWNDKGw=="],
"heap-js": ["heap-js@2.7.1", "", {}, "sha512-EQfezRg0NCZGNlhlDR3Evrw1FVL2G3LhU7EgPoxufQKruNBSYA8MiRPHeWbU+36o+Fhel0wMwM+sLEiBAlNLJA=="],
"help-me": ["help-me@5.0.0", "", {}, "sha512-7xgomUX6ADmcYzFik0HzAxh/73YlKR9bmFzf51CZwR+b6YtzU2m0u49hQCqV6SvlqIqsaxovfwdvbnsw3b/zpg=="],
"hls.js": ["hls.js@1.7.3", "", {}, "sha512-MsPlx6yVW4Qv4C7mEVou4/gk/5cN2dVxOTnrNjPSmyH2f0Ln+/UkG10Ax1PE6OvrmTVPKgQYozTtBaPOuCzq7A=="],
"hono": ["hono@4.13.5", "", {}, "sha512-O6+/eCYRkzzzy0rPWwKLiGBR1nFuUPZynnwjxN1MBA62NNqbT0wQEzQyK2gSO5yDIDB336sXQleAhOHrzlYyKw=="],
"html-url-attributes": ["html-url-attributes@3.0.1", "", {}, "sha512-ol6UPyBWqsrO6EJySPz2O7ZSr856WDrEzM5zMqp+FJJLGMW35cLYmmZnl0vztAZxRUoNZJFTCohfjuIJ8I4QBQ=="],
@@ -2311,8 +2322,6 @@
"is-fullwidth-code-point": ["is-fullwidth-code-point@3.0.0", "", {}, "sha512-zymm5+u+sCsSWyD9qNaejV3DFvhCKclKdizYaJUuHA83RLjb7nSuGnddCHGv0hk+KY7BMAlsWeK4Ueg6EV6XQg=="],
"is-function": ["is-function@1.0.2", "", {}, "sha512-lw7DUp0aWXYg+CBCN+JKkcE0Q2RayZnSvnZBlwgxHBQhqt5pZNVy4Ri7H9GmmXkdu7LUthszM+Tor1u/2iBcpQ=="],
"is-generator-function": ["is-generator-function@1.1.2", "", { "dependencies": { "call-bound": "^1.0.4", "generator-function": "^2.0.0", "get-proto": "^1.0.1", "has-tostringtag": "^1.0.2", "safe-regex-test": "^1.1.0" } }, "sha512-upqt1SkGkODW9tsGNG5mtXTXtECizwtS2kA161M+gJPc1xdb/Ax629af6YrTwcOeQHbewrPNlE5Dx7kzvXTizA=="],
"is-glob": ["is-glob@4.0.3", "", { "dependencies": { "is-extglob": "^2.1.1" } }, "sha512-xelSayHH36ZgE7ZWhli7pW34hNbNl8Ojv5KVmkJD4hBdD3th8Tfk9vYasLM+mXWOZhFkgZfxhLSnrwRr4elSSg=="],
@@ -2465,9 +2474,9 @@
"lucide-react": ["lucide-react@1.37.0", "", { "peerDependencies": { "react": "^16.5.1 || ^17.0.0 || ^18.0.0 || ^19.0.0" } }, "sha512-LPsB4rD1TD6wZu1djKOf9vUnS1jTNaHbolXebXDgiTdb6jeA1agIJhJsIybCmjKmQClcOaal1o1OaiYahEftyQ=="],
"lz-string": ["lz-string@1.5.0", "", { "bin": { "lz-string": "bin/bin.js" } }, "sha512-h5bgJWpxJNswbU7qCrV0tIKQCaS3blPDrqKWx+QxzuzL1zGUzij9XCWLrSLsJPu5t+eWA/ycetzYAO5IOMcWAQ=="],
"luxon": ["luxon@3.7.2", "", {}, "sha512-vtEhXh/gNjI9Yg1u4jX/0YVPMvxzHuGgCm6tC5kZyb08yjGWGnqAjGJvcXbqQR2P3MyMEFnRbpcdFS6PBcLqew=="],
"m3u8-parser": ["m3u8-parser@7.2.0", "", { "dependencies": { "@babel/runtime": "^7.12.5", "@videojs/vhs-utils": "^4.1.1", "global": "^4.4.0" } }, "sha512-CRatFqpjVtMiMaKXxNvuI3I++vUumIXVVT/JpCpdU/FynV/ceVw1qpPyyBNindL+JlPMSesx+WX1QJaZEJSaMQ=="],
"lz-string": ["lz-string@1.5.0", "", { "bin": { "lz-string": "bin/bin.js" } }, "sha512-h5bgJWpxJNswbU7qCrV0tIKQCaS3blPDrqKWx+QxzuzL1zGUzij9XCWLrSLsJPu5t+eWA/ycetzYAO5IOMcWAQ=="],
"magic-string": ["magic-string@0.30.21", "", { "dependencies": { "@jridgewell/sourcemap-codec": "^1.5.5" } }, "sha512-vd2F4YUyEXKGcLHoq+TEyCjxueSeHnFxyyjNp80yg0XV4vUhnDer/lvvlqM/arB5bXQN5K2/3oinyCRyx8T2CQ=="],
@@ -2539,8 +2548,6 @@
"mime-match": ["mime-match@1.0.2", "", { "dependencies": { "wildcard": "^1.1.0" } }, "sha512-VXp/ugGDVh3eCLOBCiHZMYWQaTNUHv2IJrut+yXA6+JbLPXHglHwfS/5A5L0ll+jkCY7fIzRJcH6OIunF+c6Cg=="],
"min-document": ["min-document@2.19.2", "", { "dependencies": { "dom-walk": "^0.1.0" } }, "sha512-8S5I8db/uZN8r9HSLFVWPdJCvYOejMcEC82VIzNUc6Zkklf/d1gg2psfE79/vyhWOj4+J8MtwmoOz3TmvaGu5A=="],
"minimatch": ["minimatch@10.2.6", "", { "dependencies": { "brace-expansion": "^5.0.8" } }, "sha512-vpLQEs+VLCr1nU0BXS07maYoFwlDAH0gngQuuttxIwutDFEMHq2blX+8vpgxDdK3J1PwjCJiep77OitTZ4Ll1A=="],
"minimist": ["minimist@1.2.8", "", {}, "sha512-2yyAR8qBkN3YuheJanUpWC5U3bb5osDywNB8RzDVlDwDHbocAJveqqj1u8+SVD7jkWT4yvsHCpWqqWqAxb0zCA=="],
@@ -2553,12 +2560,8 @@
"mkdirp-classic": ["mkdirp-classic@0.5.3", "", {}, "sha512-gKLcREMhtuZRwRAfqP3RFW+TK4JqApVBtOIftVgjuABpAtpxhPGaDcfvbhNvD0B8iD1oUr/txX35NjcaY6Ns/A=="],
"mpd-parser": ["mpd-parser@1.4.0", "", { "dependencies": { "@babel/runtime": "^7.12.5", "@videojs/vhs-utils": "^4.1.2", "@xmldom/xmldom": "^0.8.3", "global": "^4.4.0" }, "bin": { "mpd-to-m3u8-json": "bin/parse.js" } }, "sha512-blbA7XpAaOdC/PR0Tu8GiUxlx6CpBuRowQPYV7lhi9Yr9G9+dkPXIUjqjzc/NCn/uZZPopluaJlIYntIEI/VLw=="],
"ms": ["ms@2.1.3", "", {}, "sha512-6FlzubTLZG3J2a/NVCAleEhjzq5oxgHyaCU9yYXvcLsvoVaHJq/s5xXI6/XXP6tz7R9xAOtHnSO/tXtF3WRTlA=="],
"mux.js": ["mux.js@7.1.0", "", { "dependencies": { "@babel/runtime": "^7.11.2", "global": "^4.4.0" }, "bin": { "muxjs-transmux": "bin/transmux.js" } }, "sha512-NTxawK/BBELJrYsZThEulyUMDVlLizKdxyAsMuzoCD1eFj97BVaA8D/CvKsKu6FOLYkFojN5CbM9h++ZTZtknA=="],
"mysql2": ["mysql2@3.15.3", "", { "dependencies": { "aws-ssl-profiles": "^1.1.1", "denque": "^2.1.0", "generate-function": "^2.3.1", "iconv-lite": "^0.7.0", "long": "^5.2.1", "lru.min": "^1.0.0", "named-placeholders": "^1.1.3", "seq-queue": "^0.0.5", "sqlstring": "^2.3.2" } }, "sha512-FBrGau0IXmuqg4haEZRBfHNWB5mUARw6hNwPDXXGg0XzVJ50mr/9hb267lvpVMnhZ1FON3qNd4Xfcez1rbFwSg=="],
"named-placeholders": ["named-placeholders@1.1.6", "", { "dependencies": { "lru.min": "^1.1.0" } }, "sha512-Tz09sEL2EEuv5fFowm419c1+a/jSMiBjI9gHxVLrVdbUkkNUUfjsVYs9pVZu5oCon/kmRh9TfLEObFtkVxmY0w=="],
@@ -2591,6 +2594,8 @@
"node-releases": ["node-releases@2.0.54", "", {}, "sha512-YHs7BmmcsdAI5Ozuf8JZo6PT0mv2GIWC9vMfvUC3dp65M8hn7Ux8CPL+2oBI7juNuj9d0ndhTcznq2ODBps9cQ=="],
"nodemailer": ["nodemailer@10.0.10", "", {}, "sha512-He9XskOFms62SyAKkLu8CcGcYHAo+BSHSsORI8s4PJH8qZgZY4L4lDfvyN0twvmbpyG+eGrxpyBlskj9d9rJ8A=="],
"normalize-path": ["normalize-path@3.0.0", "", {}, "sha512-6eZs5Ls3WtCisHWp9S2GUy8dqkpGi4BVSz3GaqiE6ezub0512ESztXUwUB6C6IKbQkY2Pnb/mD4WYojCRwcwLA=="],
"object-assign": ["object-assign@4.1.1", "", {}, "sha512-rJgTQnkUnH1sFw8yT6VSU3zD3sWmu6sZhIseY8VX+GRu3P6F7Fu+JNDoXfklElbLJSnc3FUQHVe4cU5hj+BcUg=="],
@@ -2687,8 +2692,6 @@
"pkce-challenge": ["pkce-challenge@5.0.1", "", {}, "sha512-wQ0b/W4Fr01qtpHlqSqspcj3EhBvimsdh0KlHhH8HRZnMsEa0ea2fTULOXOS9ccQr3om+GcGRk4e+isrZWV8qQ=="],
"pkcs7": ["pkcs7@1.0.4", "", { "dependencies": { "@babel/runtime": "^7.5.5" }, "bin": { "pkcs7": "bin/cli.js" } }, "sha512-afRERtHn54AlwaF2/+LFszyAANTCggGilmcmILUzEjvs3XgFZT+xE6+QWQcAGmu4xajy+Xtj7acLOPdx5/eXWQ=="],
"pkg-types": ["pkg-types@2.3.1", "", { "dependencies": { "confbox": "^0.2.4", "exsolve": "^1.0.8", "pathe": "^2.0.3" } }, "sha512-y+ichcgc2LrADuhLNAx8DFjVfgz91pRxfZdI3UDhxHvcVEZsenLO+7XaU5vOp0u/7V/wZ+plyuQxtrDlZJ+yeg=="],
"playwright": ["playwright@1.61.1", "", { "dependencies": { "playwright-core": "1.61.1" }, "optionalDependencies": { "fsevents": "2.3.2" }, "bin": { "playwright": "cli.js" } }, "sha512-DWnY5o3YbLWK4GovuAVwpqL+1VwGNdUGrRr++8j8PtQQzvAVZUIMjKQ90fY689sEJZJBbZVw1rXaOKSTitkzPQ=="],
@@ -3093,14 +3096,6 @@
"victory-vendor": ["victory-vendor@37.3.6", "", { "dependencies": { "@types/d3-array": "^3.0.3", "@types/d3-ease": "^3.0.0", "@types/d3-interpolate": "^3.0.1", "@types/d3-scale": "^4.0.2", "@types/d3-shape": "^3.1.0", "@types/d3-time": "^3.0.0", "@types/d3-timer": "^3.0.0", "d3-array": "^3.1.6", "d3-ease": "^3.0.1", "d3-interpolate": "^3.0.1", "d3-scale": "^4.0.2", "d3-shape": "^3.1.0", "d3-time": "^3.0.0", "d3-timer": "^3.0.1" } }, "sha512-SbPDPdDBYp+5MJHhBCAyI7wKM3d5ivekigc2Dk2s7pgbZ9wIgIBYGVw4zGHBml/qTFbexrofXW6Gu4noGxrOwQ=="],
"video.js": ["video.js@8.24.0", "", { "dependencies": { "@babel/runtime": "^7.28.4", "@videojs/http-streaming": "^3.17.5", "@videojs/vhs-utils": "^4.1.2", "@videojs/xhr": "2.7.0", "aes-decrypter": "^4.0.2", "global": "4.4.0", "m3u8-parser": "^7.2.0", "mpd-parser": "^1.4.0", "mux.js": "^7.0.1", "videojs-contrib-quality-levels": "4.1.0", "videojs-font": "4.2.0", "videojs-vtt.js": "0.15.5" } }, "sha512-ZUuxTlDRkEWhpp2mvGCSFzI0Tc3zcC6IUZbIhBcZ3QW0Ms+IRp/tzYkhqexcMYwrcukRFiWWJmRANsLKvv8Vsg=="],
"videojs-contrib-quality-levels": ["videojs-contrib-quality-levels@4.1.0", "", { "dependencies": { "global": "^4.4.0" }, "peerDependencies": { "video.js": "^8" } }, "sha512-TfrXJJg1Bv4t6TOCMEVMwF/CoS8iENYsWNKip8zfhB5kTcegiFYezEA0eHAJPU64ZC8NQbxQgOwAsYU8VXbOWA=="],
"videojs-font": ["videojs-font@4.2.0", "", {}, "sha512-YPq+wiKoGy2/M7ccjmlvwi58z2xsykkkfNMyIg4xb7EZQQNwB71hcSsB3o75CqQV7/y5lXkXhI/rsGAS7jfEmQ=="],
"videojs-vtt.js": ["videojs-vtt.js@0.15.5", "", { "dependencies": { "global": "^4.3.1" } }, "sha512-yZbBxvA7QMYn15Lr/ZfhhLPrNpI/RmCSCqgIff57GC2gIrV5YfyzLfLyZMj0NnZSAz8syB4N0nHXpZg9MyrMOQ=="],
"vite": ["vite@8.2.2", "", { "dependencies": { "lightningcss": "^1.33.0", "picomatch": "^4.0.5", "postcss": "^8.5.26", "rolldown": "~1.2.4", "tinyglobby": "^0.2.17" }, "optionalDependencies": { "fsevents": "~2.3.3" }, "peerDependencies": { "@types/node": "^20.19.0 || >=22.12.0", "@vitejs/devtools": "^0.4.0 || ^0.5.0", "esbuild": "^0.27.0 || ^0.28.0", "jiti": ">=1.21.0", "less": "^4.0.0", "sass": "^1.70.0", "sass-embedded": "^1.70.0", "stylus": ">=0.54.8", "sugarss": "^5.0.0", "terser": "^5.16.0", "tsx": "^4.8.1", "yaml": "^2.4.2" }, "optionalPeers": ["@types/node", "@vitejs/devtools", "esbuild", "jiti", "less", "sass", "sass-embedded", "stylus", "sugarss", "terser", "tsx", "yaml"], "bin": { "vite": "bin/vite.js" } }, "sha512-cFKLV/PRgAUlIRm5WjMjJ86jrftzpqcgH+Us+DS8mI3CDNiH30Whrz8uHL3+MOLPAgqbMBAqWdAHAphOAM+z/Q=="],
"vite-tsconfig-paths": ["vite-tsconfig-paths@6.1.1", "", { "dependencies": { "debug": "^4.1.1", "globrex": "^0.1.2", "tsconfck": "^3.0.3" }, "peerDependencies": { "vite": "*" } }, "sha512-2cihq7zliibCCZ8P9cKJrQBkfgdvcFkOOc3Y02o3GWUDLgqjWsZudaoiuOwO/gzTzy17cS5F7ZPo4bsnS4DGkg=="],
@@ -3239,6 +3234,10 @@
"@reduxjs/toolkit/reselect": ["reselect@5.3.0", "", {}, "sha512-XGoLeRAVzUTcJ1qkxPQhDJyIZ5d6zzZD9nT7AEZOaaU9UbWclhycElmhO+VD5bFeLuzhPBaOV2oXC8uG35ZSpg=="],
"@shumai/core/@temporalio/activity": ["@temporalio/activity@1.24.0", "", { "dependencies": { "@temporalio/client": "1.24.0", "@temporalio/common": "1.24.0" } }, "sha512-eBBho/r9XHe2KczYGzmaSp6e9+6c9ta/GGOBldIBr00D/cV34kJ76iAY3Ue1pQTaP7YaX3phMuOz07uUg/IKyA=="],
"@shumai/core/@temporalio/workflow": ["@temporalio/workflow@1.24.0", "", { "dependencies": { "@temporalio/common": "1.24.0", "@temporalio/proto": "1.24.0", "nexus-rpc": "^0.0.3" } }, "sha512-pIA9nxkDCCbBnNuMSffQ44HWHYe3zXFfrDpDLwzh/FUSx8uLbsBM8jwf5oUMf+5eV/qVW9PknZhOu9IFO4XnPA=="],
"@shumai/webui/@types/react-dom": ["@types/react-dom@19.2.5", "", { "peerDependencies": { "@types/react": "^19.2.0" } }, "sha512-fMPwH9v7r/pp43yUd2/Mbiex5KouJwwR3dzHkhLREUC6764VyDsqxhAxv6OFEYR1RhjOyD1naqba8ECDBe7ZQg=="],
"@spectrum-web-components/action-button/@spectrum-web-components/icon": ["@spectrum-web-components/icon@0.37.0", "", { "dependencies": { "@spectrum-web-components/base": "^0.37.0", "@spectrum-web-components/iconset": "^0.37.0" } }, "sha512-DZ1yOKHxNQygzIJ7u9akYComDnMmxuTzfmYxxt9sHdUXmc1NbMkCJqEaAG/LYsdb+hc4xWaDTT9SzArLKZ/btA=="],
@@ -3489,6 +3488,14 @@
"@prisma/studio-core/@radix-ui/react-toggle/@radix-ui/react-use-controllable-state": ["@radix-ui/react-use-controllable-state@1.2.2", "", { "dependencies": { "@radix-ui/react-use-effect-event": "0.0.2", "@radix-ui/react-use-layout-effect": "1.1.1" }, "peerDependencies": { "@types/react": "*", "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc" }, "optionalPeers": ["@types/react"] }, "sha512-BjasUjixPFdS+NKkypcyyN5Pmg83Olst0+c6vGov0diwTEo6mgdqVR6hxcEgFuh4QrAs7Rc+9KuGJ9TVCj0Zzg=="],
"@shumai/core/@temporalio/activity/@temporalio/client": ["@temporalio/client@1.24.0", "", { "dependencies": { "@grpc/grpc-js": "^1.12.4", "@temporalio/common": "1.24.0", "@temporalio/proto": "1.24.0", "abort-controller": "^3.0.0", "long": "^5.3.2", "nexus-rpc": "^0.0.3", "uuid": "^11.1.0" } }, "sha512-/GGg6ZYl+pN/bQEvq414CyjDvyM+MM1WsMzi/7SCYb3UlivVXx9nOmH/6BGx8OfYFayp11fkmfSh6rt7vW/ePQ=="],
"@shumai/core/@temporalio/activity/@temporalio/common": ["@temporalio/common@1.24.0", "", { "dependencies": { "@temporalio/proto": "1.24.0", "long": "^5.3.2", "ms": "3.0.0-canary.1", "nexus-rpc": "^0.0.3", "protobufjs": "^8.7.1" } }, "sha512-jXp77fXwynBcVNc/CaGMT33TVYnhcN26S959HDv7Q71oc0rN6d8RHWR5z0nRCDAu5CQwZQRRhVK0TRdyAiEg7Q=="],
"@shumai/core/@temporalio/workflow/@temporalio/common": ["@temporalio/common@1.24.0", "", { "dependencies": { "@temporalio/proto": "1.24.0", "long": "^5.3.2", "ms": "3.0.0-canary.1", "nexus-rpc": "^0.0.3", "protobufjs": "^8.7.1" } }, "sha512-jXp77fXwynBcVNc/CaGMT33TVYnhcN26S959HDv7Q71oc0rN6d8RHWR5z0nRCDAu5CQwZQRRhVK0TRdyAiEg7Q=="],
"@shumai/core/@temporalio/workflow/@temporalio/proto": ["@temporalio/proto@1.24.0", "", { "dependencies": { "long": "^5.3.2", "protobufjs": "^8.7.1" } }, "sha512-vhlhB9GpoyJx+78r3lDahHywpdri6LKx7kt7djLIw6XzHyb8FQTIi4lX8DSJBqKp3+HVfL1CLwiqO26KFRzg/g=="],
"@spectrum-web-components/action-button/@spectrum-web-components/icon/@spectrum-web-components/iconset": ["@spectrum-web-components/iconset@0.37.0", "", { "dependencies": { "@spectrum-web-components/base": "^0.37.0" } }, "sha512-s1TBjWeQtGdGf6n+r+aG2glv77IgR0pNs0HLusGyYt6jzWFkP276fE7Neqq/2XvG9a9zvAwngffnWS3td/Gtmg=="],
"@spectrum-web-components/action-group/@spectrum-web-components/icons-workflow/@spectrum-web-components/icon": ["@spectrum-web-components/icon@0.37.0", "", { "dependencies": { "@spectrum-web-components/base": "^0.37.0", "@spectrum-web-components/iconset": "^0.37.0" } }, "sha512-DZ1yOKHxNQygzIJ7u9akYComDnMmxuTzfmYxxt9sHdUXmc1NbMkCJqEaAG/LYsdb+hc4xWaDTT9SzArLKZ/btA=="],
@@ -3599,6 +3606,20 @@
"@prisma/studio-core/@radix-ui/react-toggle/@radix-ui/react-use-controllable-state/@radix-ui/react-use-layout-effect": ["@radix-ui/react-use-layout-effect@1.1.1", "", { "peerDependencies": { "@types/react": "*", "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc" }, "optionalPeers": ["@types/react"] }, "sha512-RbJRS4UWQFkzHTTwVymMTUv8EqYhOp8dOOviLj2ugtTiXRaRQS7GLGxZTLL1jWhMeoSCf5zmcZkqTl9IiYfXcQ=="],
"@shumai/core/@temporalio/activity/@temporalio/client/@temporalio/proto": ["@temporalio/proto@1.24.0", "", { "dependencies": { "long": "^5.3.2", "protobufjs": "^8.7.1" } }, "sha512-vhlhB9GpoyJx+78r3lDahHywpdri6LKx7kt7djLIw6XzHyb8FQTIi4lX8DSJBqKp3+HVfL1CLwiqO26KFRzg/g=="],
"@shumai/core/@temporalio/activity/@temporalio/common/@temporalio/proto": ["@temporalio/proto@1.24.0", "", { "dependencies": { "long": "^5.3.2", "protobufjs": "^8.7.1" } }, "sha512-vhlhB9GpoyJx+78r3lDahHywpdri6LKx7kt7djLIw6XzHyb8FQTIi4lX8DSJBqKp3+HVfL1CLwiqO26KFRzg/g=="],
"@shumai/core/@temporalio/activity/@temporalio/common/ms": ["ms@3.0.0-canary.1", "", {}, "sha512-kh8ARjh8rMN7Du2igDRO9QJnqCb2xYTJxyQYK7vJJS4TvLLmsbyhiKpSW+t+y26gyOyMd0riphX0GeWKU3ky5g=="],
"@shumai/core/@temporalio/activity/@temporalio/common/protobufjs": ["protobufjs@8.8.0", "", { "dependencies": { "long": "^5.3.2" } }, "sha512-N3xhQ5yyBx3vQq4gubBfASzYhJGNzeDbjqBpu61g7UVylsN/qyffU96TKWD3GbbLOKF82VGNRNvv1+BFgE31Eg=="],
"@shumai/core/@temporalio/workflow/@temporalio/common/ms": ["ms@3.0.0-canary.1", "", {}, "sha512-kh8ARjh8rMN7Du2igDRO9QJnqCb2xYTJxyQYK7vJJS4TvLLmsbyhiKpSW+t+y26gyOyMd0riphX0GeWKU3ky5g=="],
"@shumai/core/@temporalio/workflow/@temporalio/common/protobufjs": ["protobufjs@8.8.0", "", { "dependencies": { "long": "^5.3.2" } }, "sha512-N3xhQ5yyBx3vQq4gubBfASzYhJGNzeDbjqBpu61g7UVylsN/qyffU96TKWD3GbbLOKF82VGNRNvv1+BFgE31Eg=="],
"@shumai/core/@temporalio/workflow/@temporalio/proto/protobufjs": ["protobufjs@8.8.0", "", { "dependencies": { "long": "^5.3.2" } }, "sha512-N3xhQ5yyBx3vQq4gubBfASzYhJGNzeDbjqBpu61g7UVylsN/qyffU96TKWD3GbbLOKF82VGNRNvv1+BFgE31Eg=="],
"@spectrum-web-components/action-group/@spectrum-web-components/icons-workflow/@spectrum-web-components/icon/@spectrum-web-components/iconset": ["@spectrum-web-components/iconset@0.37.0", "", { "dependencies": { "@spectrum-web-components/base": "^0.37.0" } }, "sha512-s1TBjWeQtGdGf6n+r+aG2glv77IgR0pNs0HLusGyYt6jzWFkP276fE7Neqq/2XvG9a9zvAwngffnWS3td/Gtmg=="],
"@spectrum-web-components/card/@spectrum-web-components/icons-workflow/@spectrum-web-components/icon/@spectrum-web-components/iconset": ["@spectrum-web-components/iconset@0.37.0", "", { "dependencies": { "@spectrum-web-components/base": "^0.37.0" } }, "sha512-s1TBjWeQtGdGf6n+r+aG2glv77IgR0pNs0HLusGyYt6jzWFkP276fE7Neqq/2XvG9a9zvAwngffnWS3td/Gtmg=="],
@@ -3620,5 +3641,7 @@
"readdir-glob/minimatch/brace-expansion/balanced-match": ["balanced-match@1.0.2", "", {}, "sha512-3oSeUO0TMV67hN1AmbXsK4yaqU7tjiHlbxRDZOpH0KW9+CeX4bRAaX0Anxt0tx2MrpRpWwQaPwIlISEJhYU5Pw=="],
"@prisma/studio-core/@radix-ui/react-toggle/@radix-ui/react-primitive/@radix-ui/react-slot/@radix-ui/react-compose-refs": ["@radix-ui/react-compose-refs@1.1.2", "", { "peerDependencies": { "@types/react": "*", "react": "^16.8 || ^17.0 || ^18.0 || ^19.0 || ^19.0.0-rc" }, "optionalPeers": ["@types/react"] }, "sha512-z4eqJvfiNnFMHIIvXP3CY57y2WJs5g2v3X0zm9mEJkrkNv4rDxu+sg9Jh8EkXyeqBkB7SOcboo9dMVqhyrACIg=="],
"@shumai/core/@temporalio/activity/@temporalio/client/@temporalio/proto/protobufjs": ["protobufjs@8.8.0", "", { "dependencies": { "long": "^5.3.2" } }, "sha512-N3xhQ5yyBx3vQq4gubBfASzYhJGNzeDbjqBpu61g7UVylsN/qyffU96TKWD3GbbLOKF82VGNRNvv1+BFgE31Eg=="],
}
}
@@ -0,0 +1,69 @@
services:
postgres:
image: docker.m.daocloud.io/pgvector/pgvector:pg18
container_name: shumai_postgres
environment:
POSTGRES_USER: shumai
POSTGRES_PASSWORD: shumai_password
POSTGRES_DB: shumai_db
volumes:
- postgres_data:/var/lib/postgresql/18/docker
healthcheck:
test: ['CMD-SHELL', 'pg_isready -U shumai -d shumai_db']
interval: 5s
timeout: 5s
retries: 5
embedding-server:
image: shumaione/embedding-server:latest
container_name: shumai_embedding
restart: unless-stopped
shumai:
image: crpi-000nkafzu6xvpwsc.cn-hongkong.personal.cr.aliyuncs.com/shumai-one/shumai:latest
cap_add:
- SYS_ADMIN
security_opt:
- apparmor=unconfined
- seccomp=unconfined
container_name: shumai_app
ports:
- '3000:3000'
volumes:
- ./data:/app/data
# Hardware Acceleration (NVENC / QuickSync / VA-API / RKMPP)
# Uncomment the extends section below for hardware acceleration:
# extends:
# file: hwaccel.transcoding.yaml
# service: cpu # set to one of [nvenc, quicksync, rkmpp, vaapi, vaapi-wsl] for accelerated transcoding
environment:
# Database Configuration
DATABASE_URL: postgres://shumai:shumai_password@shumai_postgres:5432/shumai_db?search_path=public&sslmode=disable
# Server Configuration
SHUMAI_SERVER_PORT: 3000
# S3 Configuration
STORAGE_BACKEND: local
AWS_ENDPOINT_URL_S3: http://localhost:3000
# Better Auth
BETTER_AUTH_SECRET: qSxs9DxzHDZBbeeTNPEwBuspYwipBqz5Gk5XdBjNhWw=
# Sandbox Configuration
ENABLE_WEAKER_NESTED_SANDBOX: 'true'
# Embedding Configuration
EMBEDDING_PROVIDER: gemma-embedding-2-openai
GEMMA_EMBEDDING_OPENAI_BASE_URL: http://embedding-server:9379/v1
ENABLE_EMBEDDING_AGENT: 'true'
CONCURRENCY_AGENT: 1
depends_on:
postgres:
condition: service_healthy
embedding-server:
condition: service_started
volumes:
postgres_data:
@@ -0,0 +1,89 @@
services:
postgres:
image: docker.m.daocloud.io/pgvector/pgvector:pg18
container_name: shumai_postgres
environment:
POSTGRES_USER: shumai
POSTGRES_PASSWORD: shumai_password
POSTGRES_DB: shumai_db
volumes:
- postgres_data:/var/lib/postgresql/18/docker
healthcheck:
test: ['CMD-SHELL', 'pg_isready -U shumai -d shumai_db']
interval: 5s
timeout: 5s
retries: 5
embedding-server:
image: shumaione/embedding-server:latest
container_name: shumai_embedding
restart: unless-stopped
turntable-renderer:
image: shumaione/turntable-renderer:latest
container_name: shumai_turntable
restart: unless-stopped
gotenberg:
image: gotenberg/gotenberg:8
container_name: shumai_gotenberg
restart: unless-stopped
shumai:
image: crpi-000nkafzu6xvpwsc.cn-hongkong.personal.cr.aliyuncs.com/shumai-one/shumai:latest
cap_add:
- SYS_ADMIN
security_opt:
- apparmor=unconfined
- seccomp=unconfined
container_name: shumai_app
ports:
- '3000:3000'
volumes:
- ./data:/app/data
# Hardware Acceleration (NVENC / QuickSync / VA-API / RKMPP)
# Uncomment the extends section below for hardware acceleration:
# extends:
# file: hwaccel.transcoding.yaml
# service: cpu # set to one of [nvenc, quicksync, rkmpp, vaapi, vaapi-wsl] for accelerated transcoding
environment:
# Database Configuration
DATABASE_URL: postgres://shumai:shumai_password@shumai_postgres:5432/shumai_db?search_path=public&sslmode=disable
# Server Configuration
SHUMAI_SERVER_PORT: 3000
# S3 Configuration
STORAGE_BACKEND: local
AWS_ENDPOINT_URL_S3: http://localhost:3000
# Better Auth
BETTER_AUTH_SECRET: qSxs9DxzHDZBbeeTNPEwBuspYwipBqz5Gk5XdBjNhWw=
# Sandbox Configuration
ENABLE_WEAKER_NESTED_SANDBOX: 'true'
# Embedding Configuration
EMBEDDING_PROVIDER: gemma-embedding-2-openai
GEMMA_EMBEDDING_OPENAI_BASE_URL: http://embedding-server:9379/v1
ENABLE_EMBEDDING_AGENT: 'true'
CONCURRENCY_AGENT: 1
# 3D Turntable Renderer Configuration
TURNTABLE_RENDERER_URL: http://turntable-renderer:3000
# Gotenberg Document Converter Configuration
GOTENBERG_URL: http://gotenberg:3000
depends_on:
postgres:
condition: service_healthy
embedding-server:
condition: service_started
turntable-renderer:
condition: service_started
gotenberg:
condition: service_started
volumes:
postgres_data:
@@ -6,8 +6,6 @@ services:
POSTGRES_USER: shumai
POSTGRES_PASSWORD: shumai_password
POSTGRES_DB: shumai_db
ports:
- '5432:5432'
volumes:
- postgres_data:/var/lib/postgresql/18/docker
healthcheck:
@@ -28,6 +26,11 @@ services:
- '3000:3000'
volumes:
- ./data:/app/data
# Hardware Acceleration (NVENC / QuickSync / VA-API / RKMPP)
# Uncomment the extends section below for hardware acceleration:
# extends:
# file: hwaccel.transcoding.yaml
# service: cpu # set to one of [nvenc, quicksync, rkmpp, vaapi, vaapi-wsl] for accelerated transcoding
environment:
# Database Configuration
DATABASE_URL: postgres://shumai:shumai_password@shumai_postgres:5432/shumai_db?search_path=public&sslmode=disable
@@ -0,0 +1,69 @@
services:
postgres:
image: pgvector/pgvector:pg18
container_name: shumai_postgres
environment:
POSTGRES_USER: shumai
POSTGRES_PASSWORD: shumai_password
POSTGRES_DB: shumai_db
volumes:
- postgres_data:/var/lib/postgresql/18/docker
healthcheck:
test: ['CMD-SHELL', 'pg_isready -U shumai -d shumai_db']
interval: 5s
timeout: 5s
retries: 5
embedding-server:
image: shumaione/embedding-server:latest
container_name: shumai_embedding
restart: unless-stopped
shumai:
image: shumaione/shumai:latest
cap_add:
- SYS_ADMIN
security_opt:
- apparmor=unconfined
- seccomp=unconfined
container_name: shumai_app
ports:
- '3000:3000'
volumes:
- ./data:/app/data
# Hardware Acceleration (NVENC / QuickSync / VA-API / RKMPP)
# Uncomment the extends section below for hardware acceleration:
# extends:
# file: hwaccel.transcoding.yaml
# service: cpu # set to one of [nvenc, quicksync, rkmpp, vaapi, vaapi-wsl] for accelerated transcoding
environment:
# Database Configuration
DATABASE_URL: postgres://shumai:shumai_password@shumai_postgres:5432/shumai_db?search_path=public&sslmode=disable
# Server Configuration
SHUMAI_SERVER_PORT: 3000
# S3 Configuration
STORAGE_BACKEND: local
AWS_ENDPOINT_URL_S3: http://localhost:3000
# Better Auth
BETTER_AUTH_SECRET: qSxs9DxzHDZBbeeTNPEwBuspYwipBqz5Gk5XdBjNhWw=
# Sandbox Configuration
ENABLE_WEAKER_NESTED_SANDBOX: 'true'
# Embedding Configuration
EMBEDDING_PROVIDER: gemma-embedding-2-openai
GEMMA_EMBEDDING_OPENAI_BASE_URL: http://embedding-server:9379/v1
ENABLE_EMBEDDING_AGENT: 'true'
CONCURRENCY_AGENT: 1
depends_on:
postgres:
condition: service_healthy
embedding-server:
condition: service_started
volumes:
postgres_data:
@@ -0,0 +1,89 @@
services:
postgres:
image: pgvector/pgvector:pg18
container_name: shumai_postgres
environment:
POSTGRES_USER: shumai
POSTGRES_PASSWORD: shumai_password
POSTGRES_DB: shumai_db
volumes:
- postgres_data:/var/lib/postgresql/18/docker
healthcheck:
test: ['CMD-SHELL', 'pg_isready -U shumai -d shumai_db']
interval: 5s
timeout: 5s
retries: 5
embedding-server:
image: shumaione/embedding-server:latest
container_name: shumai_embedding
restart: unless-stopped
turntable-renderer:
image: shumaione/turntable-renderer:latest
container_name: shumai_turntable
restart: unless-stopped
gotenberg:
image: gotenberg/gotenberg:8
container_name: shumai_gotenberg
restart: unless-stopped
shumai:
image: shumaione/shumai:latest
cap_add:
- SYS_ADMIN
security_opt:
- apparmor=unconfined
- seccomp=unconfined
container_name: shumai_app
ports:
- '3000:3000'
volumes:
- ./data:/app/data
# Hardware Acceleration (NVENC / QuickSync / VA-API / RKMPP)
# Uncomment the extends section below for hardware acceleration:
# extends:
# file: hwaccel.transcoding.yaml
# service: cpu # set to one of [nvenc, quicksync, rkmpp, vaapi, vaapi-wsl] for accelerated transcoding
environment:
# Database Configuration
DATABASE_URL: postgres://shumai:shumai_password@shumai_postgres:5432/shumai_db?search_path=public&sslmode=disable
# Server Configuration
SHUMAI_SERVER_PORT: 3000
# S3 Configuration
STORAGE_BACKEND: local
AWS_ENDPOINT_URL_S3: http://localhost:3000
# Better Auth
BETTER_AUTH_SECRET: qSxs9DxzHDZBbeeTNPEwBuspYwipBqz5Gk5XdBjNhWw=
# Sandbox Configuration
ENABLE_WEAKER_NESTED_SANDBOX: 'true'
# Embedding Configuration
EMBEDDING_PROVIDER: gemma-embedding-2-openai
GEMMA_EMBEDDING_OPENAI_BASE_URL: http://embedding-server:9379/v1
ENABLE_EMBEDDING_AGENT: 'true'
CONCURRENCY_AGENT: 1
# 3D Turntable Renderer Configuration
TURNTABLE_RENDERER_URL: http://turntable-renderer:3000
# Gotenberg Document Converter Configuration
GOTENBERG_URL: http://gotenberg:3000
depends_on:
postgres:
condition: service_healthy
embedding-server:
condition: service_started
turntable-renderer:
condition: service_started
gotenberg:
condition: service_started
volumes:
postgres_data:
+5 -2
View File
@@ -6,8 +6,6 @@ services:
POSTGRES_USER: shumai
POSTGRES_PASSWORD: shumai_password
POSTGRES_DB: shumai_db
ports:
- '5432:5432'
volumes:
- postgres_data:/var/lib/postgresql/18/docker
healthcheck:
@@ -28,6 +26,11 @@ services:
- '3000:3000'
volumes:
- ./data:/app/data
# Hardware Acceleration (NVENC / QuickSync / VA-API / RKMPP)
# Uncomment the extends section below for hardware acceleration:
# extends:
# file: hwaccel.transcoding.yaml
# service: cpu # set to one of [nvenc, quicksync, rkmpp, vaapi, vaapi-wsl] for accelerated transcoding
environment:
# Database Configuration
DATABASE_URL: postgres://shumai:shumai_password@shumai_postgres:5432/shumai_db?search_path=public&sslmode=disable
@@ -0,0 +1,44 @@
# Configurations for hardware-accelerated transcoding
services:
cpu: {}
nvenc:
deploy:
resources:
reservations:
devices:
- driver: nvidia
count: 1
capabilities:
- gpu
- compute
- video
quicksync:
devices:
- /dev/dri:/dev/dri
rkmpp:
security_opt: # enables full access to /sys and /proc, still far better than privileged: true
- systempaths=unconfined
- apparmor=unconfined
group_add:
- video
devices:
- /dev/rga:/dev/rga
- /dev/dri:/dev/dri
- /dev/dma_heap:/dev/dma_heap
- /dev/mpp_service:/dev/mpp_service
vaapi:
devices:
- /dev/dri:/dev/dri
vaapi-wsl: # use this for VAAPI if running in WSL2
devices:
- /dev/dri:/dev/dri
- /dev/dxg:/dev/dxg
volumes:
- /usr/lib/wsl:/usr/lib/wsl
environment:
- LIBVA_DRIVER_NAME=d3d12
@@ -7,8 +7,6 @@ services:
POSTGRES_USER: shumai
POSTGRES_PASSWORD: shumai_password
POSTGRES_DB: shumai_db
ports:
- '5432:5432'
volumes:
- postgres_data:/var/lib/postgresql
healthcheck:
+908
View File
@@ -0,0 +1,908 @@
---
title: "Changelog"
description: "Recent updates, new features, and improvements in Shumai"
---
<Update label="v0.5.3" description="October 16, 2026">
### Added
* **cli**: Add administrative `shumai regenerate-embeddings` command to regenerate multimodal vector embeddings for all existing media files across active teams
* **docker**: Provide three Docker Compose deployment tiers (Lite, with Embedding Server, and Full) with recommended memory specifications
* **agent**: Save video chunk embeddings incrementally to the database as each chunk finishes, preserving completed progress on failure and resuming from where it left off on retry
* **agent**: Add support for local LiteRT-LM Gemma Embedding 2 (`gemma-embedding-2-openai`) provider and environment-enforced read-only embedding agents with automatic 768-dimensional vector search indexing
### Fixed
### Changed
* **search**: Display search results in a responsive card grid with interactive hover video sprite scrubbing, semantic match chunk time range previews, and new-tab opening on click
* **search**: Significantly improve semantic search response times by generating query text embeddings directly with in-memory caching instead of dispatching background tasks
</Update>
<Update label="v0.5.2" description="October 13, 2026">
### Added
* **3d asset**: Add support for 3D model files (.glb, .gltf, .obj, .fbx, .stl, .dae, .usd, .usda, .usdc, .usdz) with automated turntable video rendering, hover sprite scrubbing, and an interactive 3D turntable viewer featuring 360° ruler scrubbing and degree-based markup comments
* **kanban**: Related assets on a task are now clickable, taking you directly to that file or folder in its project
* **settings**: Add an Integrations section to Team Settings with dedicated configuration for Gotenberg Document Converter, supporting connection testing and environment variable override protection
### Fixed
* **transcode**: Fix an issue where portrait photos from cameras and phones appeared sideways in thumbnails, previews, watermarked share links and annotated snapshots, and were listed with their width and height swapped
### Changed
</Update>
<Update label="v0.5.1" description="October 6, 2026">
### Added
* **notification**: Notification cards are now clickable, taking you directly to the relevant file
* **transcode**: Add hardware decoding and GPU scaling support for NVIDIA NVENC, Intel QuickSync Video (QSV), Rockchip (RKMPP), and VA-API (Intel and AMD GPUs), significantly speeding up video proxy and HLS rendition generation across supported platforms. Set `SHUMAI_HW_DECODE=false` to turn hardware decoding off
* **transcode**: Add `dcraw_emu` fallback to generate WebP previews for camera RAW files without embedded JPEG previews
### Fixed
### Changed
</Update>
<Update label="v0.5.0" description="October 2, 2026">
### Added
* **transcode**: Add configurable target resolution ladders (480p, 720p, 1080p, 1440p, 2160p) for MP4 Multi Resolutions transcoding in team settings, with unified resolution matching and lowest-resolution fallback for both MP4 and HLS
* **transcode**: Add HLS live streaming with fMP4/CMAF adaptive bitrate ladders (480p, 720p, 1080p, 1440p, 2160p), configurable in Transcode Settings, with automatic quality switching, manual resolution selection in the video viewer, seamless rendition switching, and full watermarking support for shared links
* **transcode**: Add basic camera RAW image support for popular camera formats (including Sony ARW, Canon CR2/CR3, Nikon NEF, Adobe DNG, Fujifilm RAF, and Panasonic RW2), extracting high-quality embedded JPEG previews and automatically applying camera orientation for fast thumbnail and proxy generation
### Fixed
- **asset**: Fix an issue where media files remained stuck in a perpetual "Preparing..." state when transcoding failed, by displaying a clear failed status and error details in the file list and file viewer
* **transcode**: Fix an issue where videos or audio-only files containing attached cover art or picture streams could cause transcoding to fail during poster generation or select a single static image instead of the video
* **webui**: Fix inconsistent video resolution labels across download menus, mobile file headers, and the video player by correctly deriving resolutions from the video's short side, including portrait videos and 480p videos
* **trash**: Fix an issue where the "Empty Trash" confirmation dialog would close immediately instead of showing a loading state while files were being permanently deleted
### Changed
* **settings**: Redesign team Video Transcoding settings with dedicated cards for Transcode Policy, Hardware Acceleration, and FFmpeg Threads, clearly explaining MP4 and HLS playback vs download behavior and storage implications
* **settings**: Rename AI Settings group to Team AI Settings in settings navigation
</Update>
<Update label="v0.4.9" description="September 28, 2026">
### Added
- **notification**: You can now receive notifications via email, with batched digest emails and configurable team-level SMTP settings
- **transcode**: Add VA-API hardware acceleration support for Linux hosts and Docker environments with Intel and AMD GPUs, featuring automatic DRI render node discovery, variable bitrate control, and structured transcoding observability logs
- **transcode**: Add automatic fallback to software transcoding (libx264) when hardware-accelerated video encoding encounters GPU resolution constraints or runtime driver errors
- **docker**: Add modular hardware-accelerated transcoding presets matching Immich's compose configuration via `hwaccel.transcoding.yaml` with support for NVENC, QuickSync, RKMPP, and VA-API
### Fixed
### Changed
- **transcode**: Improve video poster selection using intelligent scene detection and color variance filtering to automatically skip black frames, fade-ins, and blank intros in favor of representative video content
- **transcode**: Optimize video sprite generation for large and long videos using a concurrent fast-seek pool and image compositing, avoiding full-video decoding while preserving instant single-pass generation for short clips
</Update>
<Update label="v0.4.8" description="September 24, 2026">
### Added
- **transcode**: Add HDR video proxy support: HDR videos automatically generate HDR video proxies and preview videos preserving HDR color characteristics, tone-mapped SDR thumbnails and posters, and automatic HDR player playback on supported displays with hardware EDR acceleration
- **docs**: Add a Changelog section to the documentation site so users can explore release notes and updates directly in Mintlify
### Fixed
- **webui**: Only show the center play button and darkened tint on initial video open before playback starts; once played, pausing or scrubbing back to start keeps the video canvas clean and unobstructed for review
- **transcode**: Fix an issue where videos with orientation or rotation metadata (such as smartphone vertical videos) were incorrectly displayed with stretched aspect ratios in the video player
- **cli**: Only manage and check PID files in daemon mode, preventing stale PID files, crash-loop conflicts, or spurious process checks during normal foreground and Docker execution
### Changed
- **transcode**: Improve video proxy visual quality and disable B-frames for smoother frame-by-frame seeking and scrubbing
</Update>
<Update label="v0.4.7" description="September 19, 2026">
### Added
- **changelog**: Introduce Keep a Changelog support, historical release summaries, and automated release notes extraction
- **release**: Automatically populate GitHub release bodies from CHANGELOG.md and promote unreleased notes during version bumps
- **settings**: Add FFmpeg threads slider to team media processing settings
- **asset**: Show previews of a folder's contents and a "days left" countdown for each item in Recently Deleted
- **asset**: Show the poster and hover preview for a video in the file list while it is still transcoding, with a gently pulsing cover (or an animated waiting circle before the poster is ready), "Uploading..."/"Preparing..." labels, and the duration hidden until it is ready
### Fixed
- **asset**: Fix an issue where moving a video to trash while it was still transcoding could cause it to lose its trashed status
- **transcode**: Immediately stop background video processing and transcoding when a file is permanently deleted from the trash
- **asset**: Fix proxy download filename using original file extension instead of proxy extension
- **asset**: Fix an issue where deleting or restoring a version stack from the file list did nothing
- **webui**: Center breadcrumb file action popup menu on the chevron icon
</Update>
<Update label="v0.4.6" description="September 17, 2026">
### Added
- **comment**: Add emoji reaction support for comments ([#405](https://github.com/shumaiOne/shumai/pull/405))
- **premiere-uxp**: Introduce Adobe Premiere Pro UXP plugin for Shumai ([#407](https://github.com/shumaiOne/shumai/pull/407))
- **settings**: Add team appearance setting to hide agent features ([#411](https://github.com/shumaiOne/shumai/pull/411))
### Fixed
- **premiere-uxp**: Preserve remote cloud storage presigned URLs in asset resolution ([#408](https://github.com/shumaiOne/shumai/pull/408))
- **premiere-uxp**: Scope sequence links by server endpoint and surface sync errors ([#409](https://github.com/shumaiOne/shumai/pull/409))
- **api**: Enforce SHUMAI_DEMO_MODE for API token and CLI requests ([#413](https://github.com/shumaiOne/shumai/pull/413))
### Documentation
- Add NLE export and Premiere Pro UXP plugin documentation ([#412](https://github.com/shumaiOne/shumai/pull/412))
</Update>
<Update label="v0.4.5" description="September 13, 2026">
### Added
- **core**: Export comments to FCP, Premiere, Media Composer, and Resolve ([#400](https://github.com/shumaiOne/shumai/pull/400))
- **webui**: Add NLE comments export and migrate export API to POST ([#401](https://github.com/shumaiOne/shumai/pull/401))
- **webui**: Add space key quick preview for files and folders ([#402](https://github.com/shumaiOne/shumai/pull/402))
- **webui**: Show comments count badge in file card preview area ([#403](https://github.com/shumaiOne/shumai/pull/403))
- **agent**: Use deterministic storage keys for preset avatars ([#404](https://github.com/shumaiOne/shumai/pull/404))
</Update>
<Update label="v0.4.4" description="September 9, 2026">
### Added
- **upload**: S3 multipart upload with offline resume and wake lock ([#391](https://github.com/shumaiOne/shumai/pull/391))
- **transcode**: Optimize video operations using ffmpeg http range requests ([#393](https://github.com/shumaiOne/shumai/pull/393))
- **agent**: Enable read_asset tool for autofill agent ([#396](https://github.com/shumaiOne/shumai/pull/396))
- **sandbox**: Abstract SandboxService and hot-reload allowed domains ([#398](https://github.com/shumaiOne/shumai/pull/398))
### Fixed
- **s3**: Stream downloadToFile and putObject to prevent OOM on large files ([#392](https://github.com/shumaiOne/shumai/pull/392))
- **agent**: Pass agent thinkingLevel to session name generation ([#395](https://github.com/shumaiOne/shumai/pull/395))
- **agent**: Restrict pageCount and duration recommendations to matching media types ([#397](https://github.com/shumaiOne/shumai/pull/397))
- **transcode**: Co-locate screenshots and derived artifacts in asset storage directory ([#399](https://github.com/shumaiOne/shumai/pull/399))
</Update>
<Update label="v0.4.3" description="September 6, 2026">
### Added
- **asset**: Record and display agent provenance on assets and audit logs ([#380](https://github.com/shumaiOne/shumai/pull/380))
- **agent**: Store preset avatars as real files in s3 storage ([#381](https://github.com/shumaiOne/shumai/pull/381))
- **agent**: Add built-in image and video generation tools and settings ([#382](https://github.com/shumaiOne/shumai/pull/382))
- **media-generation**: Add gemini-omni-1.1-flash video model and redesign settings UI ([#383](https://github.com/shumaiOne/shumai/pull/383))
- **settings**: Redesign image/video settings to provider dialog with model toggles ([#385](https://github.com/shumaiOne/shumai/pull/385))
- **agent**: Add asset management tools and CLI commands ([#387](https://github.com/shumaiOne/shumai/pull/387))
- **cli**: Control asset deletion via SHUMAI_ALLOW_DELETE env var ([#388](https://github.com/shumaiOne/shumai/pull/388))
### Fixed
- **webui**: Clamp file card creation info to 2 lines and show tooltip on hover ([#390](https://github.com/shumaiOne/shumai/pull/390))
### Documentation
- **agents**: Add documentation for image and video generation settings ([#389](https://github.com/shumaiOne/shumai/pull/389))
### Maintenance
- **deps**: Upgrade bun to 1.4.2 and update packages ([#384](https://github.com/shumaiOne/shumai/pull/384))
- **i18n**: Update paraglide-js to 2.25.0 and stop checking in generated code ([#386](https://github.com/shumaiOne/shumai/pull/386))
</Update>
<Update label="v0.4.2" description="September 3, 2026">
### Added
- **chat**: Support markups, attachments, and timestamps in agent 1-to-1 chat ([#368](https://github.com/shumaiOne/shumai/pull/368))
- **agent**: Move thread metadata to structured context XML ([#373](https://github.com/shumaiOne/shumai/pull/373))
- **agent**: Improve timestamp precision and proxy selection for video screenshots ([#374](https://github.com/shumaiOne/shumai/pull/374))
- **agent**: Support dynamic annotationId resolution across chat and comments ([#375](https://github.com/shumaiOne/shumai/pull/375))
- **agent**: Consolidate media tools into read_asset and add S3 key support to download_asset ([#376](https://github.com/shumaiOne/shumai/pull/376))
- **provider**: Optimize provider and model configuration management ([#377](https://github.com/shumaiOne/shumai/pull/377))
- **provider**: Add provider and model sync feature with vendored generator ([#378](https://github.com/shumaiOne/shumai/pull/378))
### Fixed
- **agent**: Set empty details in asset tool results to avoid data duplication ([#370](https://github.com/shumaiOne/shumai/pull/370))
- **webui**: Strip styles and preserve multiline text on paste in chat input ([#371](https://github.com/shumaiOne/shumai/pull/371))
- **agent**: Hide main comment sessions in dashboard sessions list ([#372](https://github.com/shumaiOne/shumai/pull/372))
### Changed
- **webui**: Use semantic system colors in notification settings ([#379](https://github.com/shumaiOne/shumai/pull/379))
</Update>
<Update label="v0.4.1" description="August 30, 2026">
### Added
- **webui**: Add manage versions option to file card context menu ([#364](https://github.com/shumaiOne/shumai/pull/364))
- **agent**: Refactor message context into unified shumai_message ([#365](https://github.com/shumaiOne/shumai/pull/365))
- **webui**: Redesign comment attachment display into compact rows ([#366](https://github.com/shumaiOne/shumai/pull/366))
### Fixed
- **asset,webui**: Disallow dropping version stacks onto files and version stacks ([#363](https://github.com/shumaiOne/shumai/pull/363))
- **webui**: Prevent empty folder flash during initial loading ([#367](https://github.com/shumaiOne/shumai/pull/367))
### Changed
- **agent**: Remove executeAgentToolWorkflow and invoke core services directly ([#362](https://github.com/shumaiOne/shumai/pull/362))
</Update>
<Update label="v0.4.0" description="August 28, 2026">
### Added
- **transcode**: Add hardware-accelerated video transcoding support ([#352](https://github.com/shumaiOne/shumai/pull/352))
- **webui**: Truncate long file names with 100ms hover tooltip ([#355](https://github.com/shumaiOne/shumai/pull/355))
- **webui**: Add simplified mobile-friendly layout for file and share lists ([#356](https://github.com/shumaiOne/shumai/pull/356))
- **webui**: Add mobile friendly file detail and public share view ([#359](https://github.com/shumaiOne/shumai/pull/359))
### Fixed
- **transcode**: Apply team hardware acceleration and source bitrate capping to watermark transcoding ([#353](https://github.com/shumaiOne/shumai/pull/353))
- **share**: Hide and soft-delete symlinks when target assets are trashed ([#354](https://github.com/shumaiOne/shumai/pull/354))
- **webui**: Fix breadcrumb file dropdown trigger stretching and alignment ([#360](https://github.com/shumaiOne/shumai/pull/360))
- **webui**: Align mobile comment cards and thread vertical indicator ([#361](https://github.com/shumaiOne/shumai/pull/361))
### Documentation
- Add SECURITY.md with a private disclosure path ([#357](https://github.com/shumaiOne/shumai/pull/357))
</Update>
<Update label="v0.3.11" description="August 25, 2026">
### Added
- **auth**: Add admin password reset CLI command and reset password web UI ([#347](https://github.com/shumaiOne/shumai/pull/347))
- **project**: Add recent viewed files section to project folder tree ([#348](https://github.com/shumaiOne/shumai/pull/348))
### Fixed
- **webui**: Prevent left carousel remounting and scroll jump on file selection ([#351](https://github.com/shumaiOne/shumai/pull/351))
</Update>
<Update label="v0.3.10" description="August 23, 2026">
### Added
- **agentic kanban**: Phase 1 not-agentic kanban board ([#330](https://github.com/shumaiOne/shumai/pull/330))
- **kanban**: Resolve latest version for linked version stack assets and update demo login text ([#342](https://github.com/shumaiOne/shumai/pull/342))
- **kanban**: Update task status permissions, reporter defaults, and participant access ([#345](https://github.com/shumaiOne/shumai/pull/345))
### Fixed
- **kanban**: Use true thumbnail and poster for task asset previews ([#343](https://github.com/shumaiOne/shumai/pull/343))
- **webui**: Sidebar selection styling, kanban active state, pointer cursors, and owner dashboard ([#344](https://github.com/shumaiOne/shumai/pull/344))
- **kanban**: Improve drag-drop to column bottom, remove empty drop line, and fix timeline alignment ([#346](https://github.com/shumaiOne/shumai/pull/346))
</Update>
<Update label="v0.3.9" description="August 18, 2026">
### Added
- **agent**: Add permission controls for chat agents ([#313](https://github.com/shumaiOne/shumai/pull/313))
- **authz**: Enforce project-aware roles for agent/skill/mcp permissions ([#314](https://github.com/shumaiOne/shumai/pull/314))
- **webui**: Use project members for asset comment mentions ([#315](https://github.com/shumaiOne/shumai/pull/315))
- **quota**: Add 2-level database-backed resource quota system ([#318](https://github.com/shumaiOne/shumai/pull/318))
- **settings**: Reorganize settings into personal, team, and AI sections ([#324](https://github.com/shumaiOne/shumai/pull/324))
- **quota**: Move usage monitoring to team dashboard ([#327](https://github.com/shumaiOne/shumai/pull/327))
### Fixed
- **webui**: Preserve selected chatbot agent preference on reload ([#323](https://github.com/shumaiOne/shumai/pull/323))
- **quota**: Exclude agents from quota rules and usage records ([#325](https://github.com/shumaiOne/shumai/pull/325))
### Documentation
- **team-settings**: Add resource quotas documentation and update readme features ([#326](https://github.com/shumaiOne/shumai/pull/326))
</Update>
<Update label="v0.3.8" description="August 15, 2026">
### Added
- **webui**: Replace legacy markdown editor with new lexical wysiwyg editor ([#308](https://github.com/shumaiOne/shumai/pull/308))
- **agent**: Add nested agents.md support across backend and webui ([#309](https://github.com/shumaiOne/shumai/pull/309))
- **webui**: Improve AGENTS.md editor placeholder and read-only notice ([#310](https://github.com/shumaiOne/shumai/pull/310))
- **version-stack**: Add manage versions dialog and version removal ([#311](https://github.com/shumaiOne/shumai/pull/311))
### Fixed
- **core**: Rename latest version child record on version stack rename ([#312](https://github.com/shumaiOne/shumai/pull/312))
</Update>
<Update label="v0.3.7" description="August 14, 2026">
### Added
- **metadata**: Replace aiAutofill boolean with autofillSource enum ([#302](https://github.com/shumaiOne/shumai/pull/302))
- **agent**: Type-safe CREATION_CONTEXT metadata on create_file/create_version ([#303](https://github.com/shumaiOne/shumai/pull/303))
- **sandbox**: Default Network Sandbox to off and add settings toggle switch ([#304](https://github.com/shumaiOne/shumai/pull/304))
- **webui**: Redesign select metadata fields and improve long text field widget ([#305](https://github.com/shumaiOne/shumai/pull/305))
- **agent**: Support dynamic newOption in select and selectMulti metadata fields ([#306](https://github.com/shumaiOne/shumai/pull/306))
### Fixed
- **webui**: Fix select field height and multi-select overflow/interaction issues ([#307](https://github.com/shumaiOne/shumai/pull/307))
</Update>
<Update label="v0.3.6" description="August 12, 2026">
### Added
- **mcp**: Add MCP server support with proxy tool and OAuth flow ([#289](https://github.com/shumaiOne/shumai/pull/289))
- **mcp**: Support tool-level directTools array and 3-state UI configuration ([#297](https://github.com/shumaiOne/shumai/pull/297))
- **webui**: Direct canvas PDF rendering with selectable text ([#299](https://github.com/shumaiOne/shumai/pull/299))
### Fixed
- **agent**: Prevent agents from accessing disabled skills ([#292](https://github.com/shumaiOne/shumai/pull/292))
- **agent**: Resolve version stack name and size in list_assets ([#296](https://github.com/shumaiOne/shumai/pull/296))
- **webui**: Content-only pinch-zoom and two-finger pan in media viewers ([#298](https://github.com/shumaiOne/shumai/pull/298))
- **i18n**: Update Chinese translation for dashboard to 仪表板 ([#301](https://github.com/shumaiOne/shumai/pull/301))
### Documentation
- **mcp**: Add MCP support documentation to README and docs ([#300](https://github.com/shumaiOne/shumai/pull/300))
</Update>
<Update label="v0.3.5" description="August 9, 2026">
### Added
- **transcode**: Add Gotenberg support for Office, HTML, CSV, and Markdown PDF proxies ([#283](https://github.com/shumaiOne/shumai/pull/283))
### Fixed
- **webui**: Resolve upload card freeze for unsupported files ([#284](https://github.com/shumaiOne/shumai/pull/284))
- **webui**: Exclude file extension when selecting file name during inline rename ([#285](https://github.com/shumaiOne/shumai/pull/285))
- Use renamed asset name in download Content-Disposition ([#286](https://github.com/shumaiOne/shumai/pull/286))
- **webui**: Batch download no longer flashes a tab per file ([#287](https://github.com/shumaiOne/shumai/pull/287))
- **webui**: Enable rename and delete actions in file detail top bar breadcrumb ([#288](https://github.com/shumaiOne/shumai/pull/288))
</Update>
<Update label="v0.3.4" description="August 8, 2026">
### Added
- **sharelink**: Add sharelink watermark support ([#270](https://github.com/shumaiOne/shumai/pull/270))
- **webui**: Remove original-file display fallbacks, use proxies only ([#273](https://github.com/shumaiOne/shumai/pull/273))
- **share**: Add allow download config to share links ([#275](https://github.com/shumaiOne/shumai/pull/275))
- **transcode**: Update preview thumbnail resolution to 300p with max length constraint ([#281](https://github.com/shumaiOne/shumai/pull/281))
### Fixed
- **share**: Resolve symlink asset downloads in public share pages ([#274](https://github.com/shumaiOne/shumai/pull/274))
- **webui**: Persist project sorting and prevent full page flash ([#276](https://github.com/shumaiOne/shumai/pull/276))
### Documentation
- **share-links**: Document allow download and lazy watermark settings ([#278](https://github.com/shumaiOne/shumai/pull/278))
- **agent**: Clarify create_file/create_version context param guidance ([#282](https://github.com/shumaiOne/shumai/pull/282))
### Maintenance
- **e2e**: Watermarked download-disabled share link coverage (image + video) ([#277](https://github.com/shumaiOne/shumai/pull/277))
- **e2e**: Stabilize watermark share spec on Firefox ([#279](https://github.com/shumaiOne/shumai/pull/279))
- Split e2e scripts into app and webui targets ([#280](https://github.com/shumaiOne/shumai/pull/280))
</Update>
<Update label="v0.3.3" description="August 5, 2026">
### Added
- **agent**: Enable LLM API call retries in AgentHarness ([#250](https://github.com/shumaiOne/shumai/pull/250))
- **audit**: Add audit log support for APIs and WebUI dashboard ([#251](https://github.com/shumaiOne/shumai/pull/251))
- **webui**: Persist agent preference in localStorage ([#254](https://github.com/shumaiOne/shumai/pull/254))
- **e2e**: Add file domain comment and draw tests ([#260](https://github.com/shumaiOne/shumai/pull/260))
- **agent**: Restrict bash access for non-owner users ([#262](https://github.com/shumaiOne/shumai/pull/262))
- **agent**: Allow create_file tool to create files directly from content ([#263](https://github.com/shumaiOne/shumai/pull/263))
- **agent**: Propagate autofill context from chat to autofill workflow ([#264](https://github.com/shumaiOne/shumai/pull/264))
- **webui**: Render select options in field creation mode and pick random option color ([#265](https://github.com/shumaiOne/shumai/pull/265))
- **project**: Include team-scoped members in project member list and add 3-section dialog layout ([#267](https://github.com/shumaiOne/shumai/pull/267))
- **webui**: Disable role dropdown for project-scoped members in team view ([#269](https://github.com/shumaiOne/shumai/pull/269))
### Fixed
- **webui**: Mention popup avatars, keyboard auto-scroll, and non-sticky section headers ([#256](https://github.com/shumaiOne/shumai/pull/256))
- **webui**: Persist field order on drag and drop ([#266](https://github.com/shumaiOne/shumai/pull/266))
### Changed
- **agent**: Load models directly from database without pi fallback ([#253](https://github.com/shumaiOne/shumai/pull/253))
- **core**: Replace generated provider files with pi-ai dynamic methods ([#255](https://github.com/shumaiOne/shumai/pull/255))
### Maintenance
- **core**: Add missing tests for methods introduced in 97069042 ([#252](https://github.com/shumaiOne/shumai/pull/252))
- **e2e**: Add domain-based auth and project web app e2e tests ([#257](https://github.com/shumaiOne/shumai/pull/257))
- **e2e**: Add file, share, and collection web app e2e tests ([#258](https://github.com/shumaiOne/shumai/pull/258))
- **e2e**: Add project-level invite and file-move e2e tests ([#259](https://github.com/shumaiOne/shumai/pull/259))
- **e2e**: Add search in file list and save-search-as-collection cases ([#261](https://github.com/shumaiOne/shumai/pull/261))
- **e2e**: Add member permission change tests ([#268](https://github.com/shumaiOne/shumai/pull/268))
</Update>
<Update label="v0.3.2" description="July 30, 2026">
### Added
- **agent**: Add output truncation and streaming limits to sandboxed bash tool ([#248](https://github.com/shumaiOne/shumai/pull/248))
### Fixed
- **agent**: Download original document asset instead of pdf proxy ([#247](https://github.com/shumaiOne/shumai/pull/247))
- **webui**: Persist chatbot sidebar input text across route navigation ([#249](https://github.com/shumaiOne/shumai/pull/249))
</Update>
<Update label="v0.3.1" description="July 30, 2026">
### Added
- **transcode**: Trigger ai metadata autofill post transcode ([#234](https://github.com/shumaiOne/shumai/pull/234))
- **dashboard**: Add agent sessions view for team owners ([#237](https://github.com/shumaiOne/shumai/pull/237))
- **db**: Preserve agent sessions on asset deletion by setting onDelete to SetNull ([#240](https://github.com/shumaiOne/shumai/pull/240))
- **agent**: Add download_asset tool and prompt instructions ([#242](https://github.com/shumaiOne/shumai/pull/242))
- **db**: Update Asset.sizeByte and AiUsage token fields to BigInt ([#244](https://github.com/shumaiOne/shumai/pull/244))
### Fixed
- **agent**: Include select option labels and field name in autofill schema ([#235](https://github.com/shumaiOne/shumai/pull/235))
- **agent**: Scope agent comment sessions to thread sessions anchored at root comments ([#239](https://github.com/shumaiOne/shumai/pull/239))
- **agent**: Disable tools for session naming and make naming non-blocking ([#241](https://github.com/shumaiOne/shumai/pull/241))
- **webui**: Update file size formatting to SI decimal standard ([#245](https://github.com/shumaiOne/shumai/pull/245))
- **s3**: Enforce property syntax in S3Service interface and fix ReadableStream handling in S3StorageService ([#246](https://github.com/shumaiOne/shumai/pull/246))
</Update>
<Update label="v0.3.0" description="July 28, 2026">
### Added
- **file-browser**: Add copy name and path option to context menu ([#225](https://github.com/shumaiOne/shumai/pull/225))
- **agent**: Inject user name and role into chat context ([#226](https://github.com/shumaiOne/shumai/pull/226))
- **agent**: Refactor database session storage to normalized DAG entries and lazy thread sync ([#229](https://github.com/shumaiOne/shumai/pull/229))
- **agent**: Support ENABLE_WEAKER_NESTED_SANDBOX environment variable ([#233](https://github.com/shumaiOne/shumai/pull/233))
### Fixed
- **agent**: Ensure atomic appendEntry and persist leaf entries in DatabaseSessionStorage ([#224](https://github.com/shumaiOne/shumai/pull/224))
- Sharp version mismatch bug ([#230](https://github.com/shumaiOne/shumai/pull/230))
- **agent**: Resolve S3 MIME type detection and prompt parameter in image analysis ([#232](https://github.com/shumaiOne/shumai/pull/232))
### Changed
- **agent**: Remove implicit agent mention trigger in comment threads ([#227](https://github.com/shumaiOne/shumai/pull/227))
### Maintenance
- **workflow**: Add docker compose deployment verification job ([#231](https://github.com/shumaiOne/shumai/pull/231))
</Update>
<Update label="v0.2.3" description="July 25, 2026">
### Fixed
- **webui**: Show right sidebar by default on file detail page ([#219](https://github.com/shumaiOne/shumai/pull/219))
- **webui**: Refresh file list and detail when agent executes asset creation tools ([#220](https://github.com/shumaiOne/shumai/pull/220))
- **webui**: Fix space key drag overlay and shift range selection for version stacks ([#221](https://github.com/shumaiOne/shumai/pull/221))
- **transcode**: Handle timestamp rounding tolerance for screenshot annotations ([#223](https://github.com/shumaiOne/shumai/pull/223))
### Changed
- **e2e**: Split transcode workflow E2E tests into per-workflow test files ([#222](https://github.com/shumaiOne/shumai/pull/222))
</Update>
<Update label="v0.2.2" description="July 23, 2026">
### Added
- **webui**: Integrate AI usage API into dashboard page ([#216](https://github.com/shumaiOne/shumai/pull/216))
- **ai-usage**: Add team and member token usage tracking and APIs ([#215](https://github.com/shumaiOne/shumai/pull/215))
- **transcode**: Add PSD support via ImageMagick and enforce sRGB conversion ([#217](https://github.com/shumaiOne/shumai/pull/217))
### Fixed
- **search**: Fix buildSqlCondition date range, numeric metadata, and string parsing bugs ([#213](https://github.com/shumaiOne/shumai/pull/213))
- **webui**: Prevent horizontal overflow in search dialog result rows ([#214](https://github.com/shumaiOne/shumai/pull/214))
- **webui**: Invalidate upload tasks query when upload task starts and completes ([#218](https://github.com/shumaiOne/shumai/pull/218))
</Update>
<Update label="v0.2.1" description="July 22, 2026">
### Added
- **skill**: Add team owner skill permission control ([#209](https://github.com/shumaiOne/shumai/pull/209))
### Fixed
- **agent**: Handle non-builtin models in getModelFromDb and upgrade pi-ai ([#211](https://github.com/shumaiOne/shumai/pull/211))
- **provider**: Preserve model ids on provider update ([#212](https://github.com/shumaiOne/shumai/pull/212))
### Changed
- **core**: Split generated providers into per-provider files ([#210](https://github.com/shumaiOne/shumai/pull/210))
</Update>
<Update label="v0.2.0" description="July 21, 2026">
### Added
- **transcode**: Add PDF preview, sprite generation, and annotation support ([#193](https://github.com/shumaiOne/shumai/pull/193))
- **web**: Register pdf.worker route in production ([#195](https://github.com/shumaiOne/shumai/pull/195))
- **agent**: Add read_pdf_pages tool and PDF page context support ([#196](https://github.com/shumaiOne/shumai/pull/196))
- **transcode**: Split monolithic transcodeMedia into focused workflows ([#197](https://github.com/shumaiOne/shumai/pull/197))
- **transcode**: Add PDF proxy support for TXT and CSV files ([#199](https://github.com/shumaiOne/shumai/pull/199))
- **asset**: Replace mediaType/mimeType logic with proxyType ([#200](https://github.com/shumaiOne/shumai/pull/200))
- **docker**: Install fonts-noto-cjk in runner stage ([#202](https://github.com/shumaiOne/shumai/pull/202))
- **transcode**: Add markdown file to pdf proxy support ([#203](https://github.com/shumaiOne/shumai/pull/203))
- **chat**: Track context asset changes and update agent context ([#204](https://github.com/shumaiOne/shumai/pull/204))
- **agent**: Refactor media tools to static injection with mandatory assetId and authz ([#208](https://github.com/shumaiOne/shumai/pull/208))
### Fixed
- **transcode**: Mark pdfkit as external and extract for docker runtime ([#201](https://github.com/shumaiOne/shumai/pull/201))
- **agent**: Include skill id in formatSkillsForPrompt ([#205](https://github.com/shumaiOne/shumai/pull/205))
- **asset**: Exclude soft-deleted children from latestChildren preview ([#206](https://github.com/shumaiOne/shumai/pull/206))
- **agent**: Pass latest version asset from version stack to prompt builder ([#207](https://github.com/shumaiOne/shumai/pull/207))
</Update>
<Update label="v0.1.10" description="July 16, 2026">
### Added
- **workflow-e2e**: Add E2E tests for agentAutofillMedia workflow ([#188](https://github.com/shumaiOne/shumai/pull/188))
- **comments**: Support comment deletion and video player shortcuts ([#191](https://github.com/shumaiOne/shumai/pull/191))
- **agent**: Support disabling specific tools for agents ([#192](https://github.com/shumaiOne/shumai/pull/192))
### Maintenance
- **e2e**: Add E2E workflow tests for remaining agent and transcode services ([#189](https://github.com/shumaiOne/shumai/pull/189))
</Update>
<Update label="v0.1.9" description="July 14, 2026">
### Added
- **e2e**: Set up fullstack app-level e2e test framework and signup test case ([#184](https://github.com/shumaiOne/shumai/pull/184))
- **chat**: Chatbot UI ([#185](https://github.com/shumaiOne/shumai/pull/185))
- **agent**: Support aborting agent execution ([#186](https://github.com/shumaiOne/shumai/pull/186))
### Changed
- **api/chat**: Remove direct Prisma query from API layer ([#187](https://github.com/shumaiOne/shumai/pull/187))
</Update>
<Update label="v0.1.8" description="July 10, 2026">
### Added
- **chatbot**: Implement type-safe Chat APIs with compaction support ([#180](https://github.com/shumaiOne/shumai/pull/180))
### Fixed
- **webui**: Enable build splitting and code-split all page routes ([#183](https://github.com/shumaiOne/shumai/pull/183))
### Documentation
- **readme**: Clarify displayed file size represents original file size ([#182](https://github.com/shumaiOne/shumai/pull/182))
</Update>
<Update label="v0.1.7" description="July 6, 2026">
### Added
- **audio**: Support frame-accurate audio playback, comments, and AI safeguards ([#175](https://github.com/shumaiOne/shumai/pull/175))
- **i18n**: Localize members dialog, notifications, uploads, user menus, dates, and players ([#178](https://github.com/shumaiOne/shumai/pull/178))
- **comments**: Allow marking comments as complete or incomplete ([#179](https://github.com/shumaiOne/shumai/pull/179))
### Fixed
- **upload**: Resolve empty and generic MIME types in confirmFileUpload ([#176](https://github.com/shumaiOne/shumai/pull/176))
### Changed
- **webui**: Modularize file viewer components into registry-driven architecture ([#174](https://github.com/shumaiOne/shumai/pull/174))
### Documentation
- Improve documentation and examples for S3 compatible storage ([#177](https://github.com/shumaiOne/shumai/pull/177))
</Update>
<Update label="v0.1.6" description="July 4, 2026">
### Added
- **file-browser**: Show video duration on file card preview ([#170](https://github.com/shumaiOne/shumai/pull/170))
- **webui**: Compare versions view for version stacks ([#171](https://github.com/shumaiOne/shumai/pull/171))
### Fixed
- **webui**: Resolve conditional hooks and setState-in-render errors ([#167](https://github.com/shumaiOne/shumai/pull/167))
- **webui**: Resolve video player total frames mismatch ([#172](https://github.com/shumaiOne/shumai/pull/172))
- **version-stack**: Resolve correct latest version ID of version stack ([#173](https://github.com/shumaiOne/shumai/pull/173))
### Maintenance
- **webui**: Video player e2e framework + Safari playhead fix ([#166](https://github.com/shumaiOne/shumai/pull/166))
- **webui**: Add frame-accurate comment seek roundtrip e2e ([#168](https://github.com/shumaiOne/shumai/pull/168))
- **webui**: Verify video player restarts playback after end ([#169](https://github.com/shumaiOne/shumai/pull/169))
</Update>
<Update label="v0.1.5" description="July 2, 2026">
### Changed
- **storage**: Use original filenames for stored files ([#165](https://github.com/shumaiOne/shumai/pull/165))
### Maintenance
- **release**: Add Aliyun ACR push to release workflow ([#164](https://github.com/shumaiOne/shumai/pull/164))
</Update>
<Update label="v0.1.4" description="July 1, 2026">
### Added
- **transcode**: Change video poster resolution to 480p ([#161](https://github.com/shumaiOne/shumai/pull/161))
- **auth**: Allow setting user locale on signup ([#163](https://github.com/shumaiOne/shumai/pull/163))
### Fixed
- **webui**: Preserve video ended state by skipping playhead snap on ended ([#162](https://github.com/shumaiOne/shumai/pull/162))
</Update>
<Update label="v0.1.3" description="June 30, 2026">
### Added
- **webui**: Add i18n support using paraglide-js ([#156](https://github.com/shumaiOne/shumai/pull/156))
- **video**: Add frame-accurate video player, SMPTE timecodes, and format switcher ([#157](https://github.com/shumaiOne/shumai/pull/157))
### Fixed
- **webui**: Align video timeline duration and formatting with frameio ([#158](https://github.com/shumaiOne/shumai/pull/158))
- **webui**: Fix missing left sidebar for nested folders ([#159](https://github.com/shumaiOne/shumai/pull/159))
- **api**: Fix update team settings and make request schema type-safe ([#160](https://github.com/shumaiOne/shumai/pull/160))
</Update>
<Update label="v0.1.2" description="June 25, 2026">
### Changed
- **s3**: Remove automatic port appending to AWS_ENDPOINT_URL_S3 ([#155](https://github.com/shumaiOne/shumai/pull/155))
</Update>
<Update label="v0.1.1" description="June 23, 2026">
### Added
- **docker**: Add root entrypoint with privilege dropping for data volumes ([#151](https://github.com/shumaiOne/shumai/pull/151))
</Update>
<Update label="v0.1.0" description="June 22, 2026">
### Added
- **share**: Secure share info API and fix wrong password feedback ([#150](https://github.com/shumaiOne/shumai/pull/150))
### Fixed
- **webui**: Truncate long folder, collection, and share link names in sidebar ([#149](https://github.com/shumaiOne/shumai/pull/149))
</Update>
<Update label="v0.0.11" description="June 22, 2026">
### Added
- **cli**: Add shumai-cli and token authentication ([#144](https://github.com/shumaiOne/shumai/pull/144))
- **settings**: Use singular headers and remove collapse/counters for single agent types ([#145](https://github.com/shumaiOne/shumai/pull/145))
- **folders**: Auto expand folder tree in left sidebar on navigation ([#146](https://github.com/shumaiOne/shumai/pull/146))
- **metadata**: Add single user and multiple user custom fields ([#147](https://github.com/shumaiOne/shumai/pull/147))
### Maintenance
- **cli**: Rename npm package to @shumai-one/shumai-cli ([#148](https://github.com/shumaiOne/shumai/pull/148))
</Update>
<Update label="v0.0.10" description="June 21, 2026">
### Changed
- **web**: Dynamically register production HTML bundle routes ([#142](https://github.com/shumaiOne/shumai/pull/142))
### Documentation
- **readme**: Polish structure, correct grammar, and add dependencies table ([#143](https://github.com/shumaiOne/shumai/pull/143))
</Update>
<Update label="v0.0.9" description="June 20, 2026">
### Added
- **webui**: Show app version in avatar menu ([#139](https://github.com/shumaiOne/shumai/pull/139))
- **cli**: Add background daemonization and subcommands ([#141](https://github.com/shumaiOne/shumai/pull/141))
### Maintenance
- **release**: Add smoke tests to local release stage ([#140](https://github.com/shumaiOne/shumai/pull/140))
</Update>
<Update label="v0.0.8" description="June 19, 2026">
### Added
- **api**: Implement read-only mode for reviewers ([#122](https://github.com/shumaiOne/shumai/pull/122))
- **server**: Allow customizing server port via SHUMAI_SERVER_PORT ([#124](https://github.com/shumaiOne/shumai/pull/124))
- **embedding**: Defer embedding workflow, use transcoded assets, and execute chunks in separate activities ([#125](https://github.com/shumaiOne/shumai/pull/125))
- **webui**: Add external drag-and-drop file and folder upload support ([#126](https://github.com/shumaiOne/shumai/pull/126))
- **agent**: Implement overlapping video chunks for semantic search ([#127](https://github.com/shumaiOne/shumai/pull/127))
- **webui**: Support jumping to video timestamp from search results ([#128](https://github.com/shumaiOne/shumai/pull/128))
- **auth**: Add demo mode support ([#131](https://github.com/shumaiOne/shumai/pull/131))
- **auth**: Refine demo mode support ([#132](https://github.com/shumaiOne/shumai/pull/132))
- **webui**: Default right sidebar to collapsed for new users ([#133](https://github.com/shumaiOne/shumai/pull/133))
- **webui**: Use Shadcn Switch in toggle field and Rating component in search filter panel ([#137](https://github.com/shumaiOne/shumai/pull/137))
- **upload**: Track byte-level progress with XMLHttpRequest ([#138](https://github.com/shumaiOne/shumai/pull/138))
### Fixed
- **transcode**: Handle high-precision FPS using rational fractions ([#123](https://github.com/shumaiOne/shumai/pull/123))
- **db**: Automatically update project updatedAt on asset modification ([#129](https://github.com/shumaiOne/shumai/pull/129))
- **webui**: Use system colors for fields and improve search filter select options ([#130](https://github.com/shumaiOne/shumai/pull/130))
- **search**: Fix hasAny and hasNone operators on metadata filter ([#134](https://github.com/shumaiOne/shumai/pull/134))
- **webui**: Resolve folder tree sync, search dialog reset, and sidebar actions issues ([#136](https://github.com/shumaiOne/shumai/pull/136))
### Changed
- **webui**: Fix search dialog layout height, scroll areas, and restrict filter toggle click ([#135](https://github.com/shumaiOne/shumai/pull/135))
</Update>
<Update label="v0.0.7" description="June 17, 2026">
### Added
- **workflow**: Configure transcode and agent concurrency via environment variables ([#116](https://github.com/shumaiOne/shumai/pull/116))
- **webui**: Replace hardcoded colors with system tokens ([#119](https://github.com/shumaiOne/shumai/pull/119))
- **webui**: Replace native overflow with shadcn ScrollArea in settings ([#120](https://github.com/shumaiOne/shumai/pull/120))
- **api**: Use enum for API protocol in provider config ([#121](https://github.com/shumaiOne/shumai/pull/121))
### Fixed
- **webui**: Reload video player and reset state when switching assets ([#114](https://github.com/shumaiOne/shumai/pull/114))
### Changed
- **transcode**: Relocate TranscodeService to core and flatten API package ([#115](https://github.com/shumaiOne/shumai/pull/115))
### Documentation
- Improve and restructure shumai documentation ([#118](https://github.com/shumaiOne/shumai/pull/118))
</Update>
<Update label="v0.0.6" description="June 15, 2026">
### Added
- **sandbox**: Add pending domains support for network sandbox ([#109](https://github.com/shumaiOne/shumai/pull/109))
- **webui**: Always show right sidebar and implement left vertical carousel ([#111](https://github.com/shumaiOne/shumai/pull/111))
### Fixed
- **build**: Propagate termination signals to active child process in wrappers ([#102](https://github.com/shumaiOne/shumai/pull/102))
- **transcode**: Make deterministic transcode activity failures non-retryable in Temporal ([#104](https://github.com/shumaiOne/shumai/pull/104))
- **player**: Implement seekbar dragging and match hover highlight color ([#108](https://github.com/shumaiOne/shumai/pull/108))
- **webui**: Avoid refetching all pages on file/folder reorder ([#112](https://github.com/shumaiOne/shumai/pull/112))
- **webui**: Force video-player mime type to video/mp4 ([#113](https://github.com/shumaiOne/shumai/pull/113))
### Changed
- **webui**: Use system color design tokens in video player components ([#106](https://github.com/shumaiOne/shumai/pull/106))
- **webui**: Use requestAnimationFrame for smooth video progress bar updates ([#107](https://github.com/shumaiOne/shumai/pull/107))
- **webui**: Fetch agents in mention input list dynamically ([#110](https://github.com/shumaiOne/shumai/pull/110))
### Documentation
- **readme**: Add comprehensive README.md ([#103](https://github.com/shumaiOne/shumai/pull/103))
- **readme**: Improve remote deployment and temporal execution instructions ([#105](https://github.com/shumaiOne/shumai/pull/105))
</Update>
<Update label="v0.0.5" description="June 13, 2026">
### Added
- **npm**: Auto deploy database migrations on package start ([#101](https://github.com/shumaiOne/shumai/pull/101))
</Update>
<Update label="v0.0.4" description="June 12, 2026">
### Fixed
- **cli/docker**: Resolve global CLI external dependencies and fix Docker 404 error ([#100](https://github.com/shumaiOne/shumai/pull/100))
</Update>
<Update label="v0.0.3" description="June 12, 2026">
### Fixed
- **cli**: Resolve platform binary path resolution under nested node_modules ([#99](https://github.com/shumaiOne/shumai/pull/99))
</Update>
<Update label="v0.0.2" description="June 12, 2026">
### Fixed
- **npm**: Add repository metadata to package.json for provenance verification ([#98](https://github.com/shumaiOne/shumai/pull/98))
### Maintenance
- Add release-npm.yaml workflow ([#97](https://github.com/shumaiOne/shumai/pull/97))
</Update>
<Update label="v0.0.1" description="June 12, 2026">
### Added
- Initial release of Shumai
</Update>
+50 -11
View File
@@ -24,12 +24,13 @@ Shumai is configured via environment variables. You can define these variables i
---
## Authentication Config
## Authentication & Application Config
| Variable | Description | Example / Default |
| :------------------- | :--------------------------------------------------------------------- | :--------------------------------------------------------------- |
| `BETTER_AUTH_SECRET` | A secure, random 32-byte secret used to sign session cookies and JWTs. | Generate with `openssl rand -base64 32` |
| `BETTER_AUTH_URL` | The public URL of your Hono backend API. Used for redirect flows. | `http://localhost:3000` (respects `SHUMAI_SERVER_PORT` fallback) |
| Variable | Description | Example / Default |
| :------------------- | :--------------------------------------------------------------------------------------------------------------- | :--------------------------------------------------------------- |
| `BETTER_AUTH_SECRET` | A secure, random 32-byte secret used to sign session cookies and JWTs. | Generate with `openssl rand -base64 32` |
| `BETTER_AUTH_URL` | The public URL of your Hono backend API. Used for redirect flows and notification email links fallback. | `http://localhost:3000` (respects `SHUMAI_SERVER_PORT` fallback) |
| `APP_URL` | The public URL of the web application. Used for notification email deep links (falls back to `BETTER_AUTH_URL`). | `https://shumai.example.com` |
---
@@ -65,12 +66,24 @@ For example, if you deploy Shumai on a remote server at `http://123.456.7.8` and
Configures whether background transcoding and AI agents are executed locally or offloaded to a Temporal cluster.
| Variable | Description | Example / Default |
| :------------------------- | :------------------------------------------------------------------------------------------------- | :------------------------------ |
| `WORKFLOW_EXECUTOR` | Engine choice. Use `local` for in-process database polling, or `temporal` for distributed workers. | `local` (default) or `temporal` |
| `TEMPORAL_ADDRESS` | Host and port of the Temporal cluster gRPC service. | `localhost:7233` (default) |
| `EMBEDDING_CHUNK_DURATION` | The duration in seconds for each video chunk when generating video embeddings. | `60.0` (default) |
| `EMBEDDING_CHUNK_OVERLAP` | The duration in seconds to overlap video chunks when generating video embeddings. | `5.0` (default) |
| Variable | Description | Example / Default |
| :---------------------- | :------------------------------------------------------------------------------------------------- | :------------------------------ |
| `WORKFLOW_EXECUTOR` | Engine choice. Use `local` for in-process database polling, or `temporal` for distributed workers. | `local` (default) or `temporal` |
| `TEMPORAL_ADDRESS` | Host and port of the Temporal cluster gRPC service. | `localhost:7233` (default) |
| `CONCURRENCY_AGENT` | Maximum concurrent AI tasks executed by workers. Set to `1` when running local `embedding-server`. | `5` (default) |
| `CONCURRENCY_TRANSCODE` | Maximum concurrent media transcoding tasks executed by workers. | `1` (default) |
---
## Embedding Config
| Variable | Description | Example / Default |
| :-------------------------------- | :------------------------------------------------------------------------------------------------------------ | :-------------------------------- |
| `EMBEDDING_PROVIDER` | Multimodal embedding provider. Options: `gemini-embedding-2`, `gemma-embedding-2-openai`. | `gemini-embedding-2` (default) |
| `GEMMA_EMBEDDING_OPENAI_BASE_URL` | Base URL of the OpenAI-compatible LiteRT-LM embedding server. | `http://embedding-server:9379/v1` |
| `ENABLE_EMBEDDING_AGENT` | When set to `true`, automatically provisions and enables the system embedding agent for all teams on startup. | `false` (default) |
| `EMBEDDING_CHUNK_DURATION` | The duration in seconds for each video chunk when generating video embeddings. | `60.0` (default) |
| `EMBEDDING_CHUNK_OVERLAP` | The duration in seconds to overlap video chunks when generating video embeddings. | `5.0` (default) |
## Sandbox Config
@@ -98,3 +111,29 @@ API keys to AI providers on startup.
| `GEMINI_API_KEY` | API key for Google Gemini provider. | `AIzaSy...` |
| `OPENAI_API_KEY` | API key for OpenAI provider. | `sk-proj-...` |
| `ANTHROPIC_API_KEY` | API key for Anthropic provider. | `sk-ant-...` |
---
## External Integrations Config
Environment variables can configure external microservices globally. When set, these environment variables take precedence over any team-level settings configured in **Team Settings → Integrations**, and lock the corresponding settings fields in the WebUI to read-only.
### Gotenberg Document Converter
Used to convert Office documents, HTML, Markdown, and CSV files into high-fidelity PDF proxies for in-browser previewing. See the [Gotenberg Guide](/using-shumai/gotenberg) for setup instructions.
| Variable | Description | Example / Default |
| :------------------------------ | :----------------------------------------------------------------------- | :---------------------- |
| `GOTENBERG_URL` | Base URL of the external Gotenberg conversion API. | `http://gotenberg:3000` |
| `GOTENBERG_BASIC_AUTH_USERNAME` | Username when Basic Authentication is enabled on the Gotenberg instance. | `admin` |
| `GOTENBERG_BASIC_AUTH_PASSWORD` | Password when Basic Authentication is enabled on the Gotenberg instance. | `secret` |
### 3D Turntable Renderer
Used to convert 3D model files (GLB, OBJ, FBX, USD, STL, DAE) into interactive 360° turntable video proxies, posters, and hover sprites. See the [3D Asset Support Guide](/using-shumai/3d-asset-support) for setup instructions.
| Variable | Description | Example / Default |
| :--------------------------------------- | :-------------------------------------------------------------------------------------------------------------- | :------------------------------- |
| `TURNTABLE_RENDERER_URL` | Base URL of the external turntable-renderer service. | `http://turntable-renderer:3000` |
| `TURNTABLE_RENDERER_BASIC_AUTH_USERNAME` | Username when Basic Authentication is enabled on the turntable renderer (alias: `TURNTABLE_RENDERER_USERNAME`). | `admin` |
| `TURNTABLE_RENDERER_BASIC_AUTH_PASSWORD` | Password when Basic Authentication is enabled on the turntable renderer (alias: `TURNTABLE_RENDERER_PASSWORD`). | `secret` |
+233
View File
@@ -0,0 +1,233 @@
---
title: Hardware Transcoding
description: Accelerate video transcoding and reduce CPU load with GPU hardware acceleration.
---
High-resolution camera recordings, phone videos, and animation media files require proxy transcoding to play smoothly in web browsers. Shumai transcodes these assets in the background, generating a universal `180p` proxy for instant scrubbing alongside full-resolution preview streams.
Hardware-accelerated transcoding allows you to offload video encoding to a dedicated or integrated GPU, dramatically speeding up processing and freeing host CPU cores for other services.
Shumai includes a built-in usability probe that tests hardware encoders at startup, alongside automatic runtime fallback during video processing. If GPU drivers or device nodes are unavailable, or if a hardware encoder encounters hardware limits (such as resolution or dimension constraints on certain GPUs) or runtime driver errors, Shumai seamlessly falls back to software encoding (`libx264`) without failing user tasks.
---
## Supported Backends
| Backend | Hardware Platform | Container Driver / Devices |
| :-------------- | :--------------------------------- | :----------------------------------------------- |
| **`nvenc`** | NVIDIA GPUs | NVIDIA Container Toolkit (`gpu, compute, video`) |
| **`quicksync`** | Intel Integrated & Arc GPUs | `/dev/dri` render nodes |
| **`vaapi`** | AMD & Intel GPUs | `/dev/dri` render nodes |
| **`rkmpp`** | Rockchip ARM SoCs (e.g. RK3588) | `/dev/rga`, `/dev/dri`, `/dev/mpp_service` |
| **`vaapi-wsl`** | Windows Subsystem for Linux (WSL2) | `/dev/dri`, `/dev/dxg`, `/usr/lib/wsl` |
---
## Limitations
- The instructions on this page focus on **Docker Compose**. Other container orchestrators (such as Kubernetes or Podman) require equivalent device passthrough configurations.
- **Docker Desktop on macOS** runs inside a Linux virtual machine and does not support Apple Silicon VideoToolbox passthrough. (VideoToolbox hardware acceleration is supported when running Shumai natively via NPM or from source).
- Hardware transcoding produces slightly different bitrate characteristics compared to software encoding. Shumai uses adaptive variable bitrate (VBR) rate control tailored to web streaming to maintain visual fidelity.
---
## Prerequisites
### NVENC (NVIDIA)
1. Ensure the official NVIDIA graphics driver is installed on the host:
```bash
nvidia-smi
```
2. Install the [NVIDIA Container Toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html) on the host and restart Docker:
```bash
sudo apt-get install -y nvidia-container-toolkit
sudo nvidia-ctk runtime configure --runtime=docker
sudo systemctl restart docker
```
### QuickSync & VA-API (Intel / AMD)
1. Verify that the Linux kernel has initialized the Direct Rendering Infrastructure (DRI) render node:
```bash
ls -la /dev/dri
```
You should see `card0` and `renderD128` (or similar).
2. Shumai's Docker image automatically adjusts group permissions on container startup so the non-root application user can access host `/dev/dri` devices without permission errors.
#### GPU decoding
With VA-API, Shumai decodes and resizes video on the GPU as well as encoding it, so frames never leave GPU memory. This is much faster than decoding on the CPU, especially for 4K sources on integrated graphics. (Other backends currently decode on the CPU and encode on the GPU.)
Some videos automatically use CPU decoding with GPU encoding instead:
- Codecs your GPU cannot decode, such as HEVC or 10-bit video on older Intel chips. Shumai tries the GPU first and retries on the CPU within a second.
- Rotated videos, such as portrait phone recordings, so they keep the correct orientation.
- Watermarked share links and HDR sources.
If a GPU-decode attempt stops making progress for 2 minutes (some driver bugs make ffmpeg hang instead of failing), Shumai stops it and retries on the CPU. Change the limit with `SHUMAI_HW_DECODE_STALL_TIMEOUT` (in seconds).
To always decode on the CPU, set `SHUMAI_HW_DECODE=false` on the `shumai` service.
### WSL2 (Windows)
- Ensure you are running Windows 11 with WSL2 and have installed the appropriate GPU drivers on the Windows host.
- Use the `vaapi-wsl` preset which maps `/dev/dxg` and `/usr/lib/wsl` into the container.
---
## Setup with Docker Compose
Shumai uses a modular `hwaccel.transcoding.yaml` companion file to configure hardware devices.
### Step 1: Download Configuration Files
In your Shumai deployment directory, download the companion `hwaccel.transcoding.yaml` alongside `docker-compose.yaml`:
```bash
# Global / Official
curl -o hwaccel.transcoding.yaml https://raw.githubusercontent.com/shumaiOne/shumai/main/docker-compose/local/hwaccel.transcoding.yaml
# Accelerated Mirror (Mainland China)
curl -o hwaccel.transcoding.yaml https://gh-proxy.com/https://raw.githubusercontent.com/shumaiOne/shumai/main/docker-compose/local/hwaccel.transcoding.yaml
```
### Step 2: Enable Hardware Acceleration in Docker Compose
Open `docker-compose.yaml` (or `docker-compose-aliyun.yaml`), find the commented `extends` block under the `shumai` service, and uncomment it:
```yaml
services:
shumai:
# ...
# Hardware Acceleration (NVENC / QuickSync / VA-API / RKMPP)
extends:
file: hwaccel.transcoding.yaml
service: nvenc # set to one of [nvenc, quicksync, rkmpp, vaapi, vaapi-wsl]
```
Replace `nvenc` with your target backend:
- `nvenc` for NVIDIA GPUs
- `quicksync` for Intel Quick Sync
- `vaapi` for AMD or Intel VA-API
- `rkmpp` for Rockchip SoCs
- `vaapi-wsl` for WSL2 on Windows
### Step 3: Start the Stack
Restart or recreate the Shumai container:
```bash
docker compose up -d
```
---
## Single Compose File Setup (Unraid / Portainer)
If your environment or container manager does not support multi-file `extends`, you can inline the hardware configuration directly into the `shumai` service in `docker-compose.yaml`:
<AccordionGroup>
<Accordion title="NVIDIA NVENC">
```yaml
services:
shumai:
# ...
deploy:
resources:
reservations:
devices:
- driver: nvidia
count: 1
capabilities:
- gpu
- compute
- video
```
</Accordion>
<Accordion title="Intel QuickSync / AMD VA-API">
```yaml services: shumai: # ... devices: - /dev/dri:/dev/dri ```
</Accordion>
<Accordion title="WSL2 (Windows)">
```yaml services: shumai: # ... devices: - /dev/dri:/dev/dri - /dev/dxg:/dev/dxg volumes: -
/usr/lib/wsl:/usr/lib/wsl environment: - LIBVA_DRIVER_NAME=d3d12 ```
</Accordion>
<Accordion title="Rockchip RKMPP">
```yaml
services:
shumai:
# ...
security_opt:
- systempaths=unconfined
- apparmor=unconfined
group_add:
- video
devices:
- /dev/rga:/dev/rga
- /dev/dri:/dev/dri
- /dev/dma_heap:/dev/dma_heap
- /dev/mpp_service:/dev/mpp_service
```
</Accordion>
</AccordionGroup>
---
## Web UI Settings
In the Shumai Web UI, navigate to **Team Settings** > **Transcoding Settings**.
- **Auto (Recommended)**: Shumai automatically probes candidate hardware encoders in optimal order (`h264_nvenc` &rarr; `h264_vaapi` &rarr; `h264_qsv` &rarr; `h264_amf` &rarr; `libx264`). If hardware acceleration is unavailable, it gracefully uses software encoding.
- **Explicit Encoder**: You can pin the transcode pipeline to a specific encoder (e.g. `NVENC`, `Quick Sync`, `VA-API`).
- **Off**: Forces CPU software encoding (`libx264`) for all video jobs.
---
## Verification & Monitoring
### Check Transcode Logs
When video transcoding begins on an uploaded asset, Shumai writes a structured log indicating the chosen encoder:
```bash
docker compose logs -f shumai
```
Look for the `Starting video transcoding` log entry:
```json
{
"level": 30,
"time": 1790414665096,
"encoder": "h264_nvenc",
"hardwareAcceleration": "auto",
"inputFile": "/tmp/transcode-.../video.mp4",
"outputFile": "/tmp/transcode-.../video-180p.mp4",
"msg": "Starting video transcoding"
}
```
If hardware acceleration was successfully initialized, `encoder` will show `h264_nvenc`, `h264_vaapi`, or `h264_qsv`.
### Monitor GPU Utilization
While transcoding a video, you can observe hardware utilization on the host:
- **NVIDIA**:
```bash
nvidia-smi
# or
nvtop
```
- **Intel**:
```bash
sudo intel_gpu_top
```
- **AMD**:
```bash
sudo radeontop
```
+17 -2
View File
@@ -12,6 +12,14 @@
"light": "#F5B895",
"dark": "#A8384D"
},
"navbar": {
"links": [
{
"label": "Changelog",
"href": "/changelog"
}
]
},
"navigation": {
"groups": [
{
@@ -36,7 +44,9 @@
"using-shumai/collections",
"using-shumai/share-links",
"using-shumai/nle-and-premiere",
"using-shumai/gotenberg"
"using-shumai/3d-asset-support",
"using-shumai/gotenberg",
"using-shumai/embedding-server"
]
},
{
@@ -57,16 +67,21 @@
"pages": [
"using-shumai/transcode-settings",
"using-shumai/quotas-settings",
"using-shumai/email-notifications",
"using-shumai/password-reset"
]
},
{
"group": "Configuration",
"pages": ["configuration/env-variables"]
"pages": ["configuration/env-variables", "configuration/hardware-transcoding"]
},
{
"group": "Temporal Deployment",
"pages": ["configuration/temporal"]
},
{
"group": "Releases",
"pages": ["changelog"]
}
]
}
+8
View File
@@ -12,9 +12,14 @@ We support three installation options. For most administrators, **Docker Compose
Docker Compose automatically configures and spins up all required services (the web application, a database, and local storage).
1. **Download the Configuration**:
```bash
curl -o docker-compose.yaml https://raw.githubusercontent.com/shumaiOne/shumai/main/docker-compose/local/docker-compose.yaml
```
> [!TIP]
> **Hardware Acceleration:** By default, the standard Compose stack uses CPU-based software transcoding. If your server has a dedicated or integrated GPU (NVIDIA, Intel, AMD, or Rockchip), see the [Hardware Transcoding Guide](/configuration/hardware-transcoding) to enable GPU acceleration via the companion `hwaccel.transcoding.yaml`.
2. **Start the Services**:
```bash
docker compose up -d
@@ -39,6 +44,9 @@ To handle media processing and security isolation, the host machine must have th
- **FFmpeg & FFprobe**: Required for video transcoding and metadata extraction.
- _macOS_: `brew install ffmpeg`
- _Linux (Ubuntu/Debian)_: `sudo apt-get install ffmpeg`
- **dcraw_emu (LibRaw)**: Required for camera RAW image decoding and preview generation.
- _macOS_: `brew install libraw`
- _Linux (Ubuntu/Debian)_: `sudo apt-get install libraw-bin`
- **Agent Sandbox Utilities**: Used to run custom automation scripts securely.
- _macOS_: `brew install ripgrep`
- _Linux (Ubuntu/Debian)_: `sudo apt-get install bubblewrap socat ripgrep`
+153
View File
@@ -0,0 +1,153 @@
---
title: 3D Asset Support with Turntable Renderer
description: Render interactive 3D model turntable previews, sprite scrubbing, and video proxies using turntable-renderer.
---
3D models and mesh assets cannot always be rendered consistently or inspected reliably across diverse client web browsers and mobile devices. Shumai integrates with [turntable-renderer](https://github.com/shumaiOne/turntable-renderer) — a standalone service that executes isolated, headless Blender 5.2 LTS subprocesses — to convert 3D model uploads into standardized, high-fidelity turntable previews.
Turntable Renderer automatically produces:
- **1080×1080 Turntable Video (`turntable.mp4`)**: A 4-second, 24-frame, 6 fps video rotating the model 360° (15° per frame) clockwise under dark studio lighting with a pure black background.
- **Poster Image (`poster.webp`)**: A crisp still poster for grid cards and immediate file display.
- **Hover Sprite Sheet (`sprite.webp`)**: A fast scrubbing sheet allowing users to rotate and inspect the model directly from the file browser on hover.
- **Interactive 3D Viewer**: A specialized web viewer featuring 360° ruler scrubbing, rotation angle control (0° to 345°), playback controls, and angle-specific review comments and visual markups.
---
## Supported File Types
When turntable-renderer is configured, Shumai automatically generates 3D turntable proxies for the following formats:
| Format / Standard | Extensions |
| --------------------------- | --------------------------------- |
| Binary glTF | `.glb` |
| Universal Scene Description | `.usd`, `.usda`, `.usdc`, `.usdz` |
| Autodesk FBX | `.fbx` |
| Wavefront OBJ | `.obj` |
| Stereolithography (STL) | `.stl` |
| COLLADA | `.dae` |
---
## Configuration
Shumai supports configuring the turntable renderer either through the **WebUI** or via **Environment Variables**.
### 1. WebUI (Team Settings → Integrations)
Team owners can configure the turntable renderer directly from the web interface:
1. Navigate to **Team Settings** from the sidebar.
2. Select the **Integrations** tab.
3. Under **3D Turntable Renderer**, enter the **Server URL** (e.g. `http://turntable-renderer:3000` or `http://localhost:3003`).
4. Enter the **Username** and **Password** if Basic Authentication is enabled on the service.
5. Click **Test Connection** to verify that Shumai can communicate with the renderer and check the detected Blender version.
6. Click **Save Changes**.
### 2. Environment Variables
You can also configure turntable-renderer globally across the system using environment variables on the machine running Shumai or the transcode workers:
| Variable | Required | Description |
| ---------------------------------------- | --------------- | ------------------------------------------------------------------------------------- |
| `TURNTABLE_RENDERER_URL` | Yes (to enable) | Base URL of the turntable renderer service, e.g. `http://turntable-renderer:3000`. |
| `TURNTABLE_RENDERER_BASIC_AUTH_USERNAME` | No | Username when Basic Authentication is enabled (alias: `TURNTABLE_RENDERER_USERNAME`). |
| `TURNTABLE_RENDERER_BASIC_AUTH_PASSWORD` | No | Password when Basic Authentication is enabled (alias: `TURNTABLE_RENDERER_PASSWORD`). |
### Configuration Priority & Read-Only Mode
> [!IMPORTANT]
> **Environment variables take highest priority.** If `TURNTABLE_RENDERER_URL` is set in your environment, Shumai will always use the environment variable configuration.
>
> When configured via environment variables, the settings in **Team Settings → Integrations** will display a **Configured via environment variables (Read-only)** badge, and the input fields will be locked to prevent accidental changes or configuration drift.
---
## Deploying Turntable Renderer
### Docker Quickstart
The easiest way to run `turntable-renderer` is using the prebuilt Docker image:
```bash
docker run -d --rm \
--name turntable-renderer \
-p 3003:3000 \
-e TURNTABLE_RENDERER_API_BASIC_AUTH_USERNAME=admin \
-e TURNTABLE_RENDERER_API_BASIC_AUTH_PASSWORD=your_password \
shumaione/turntable-renderer:latest
```
> **Note**: Host port `3003` maps to internal container port `3000`.
### Docker Compose Setup
Add a `turntable-renderer` service to your `docker-compose.yaml`:
```yaml
services:
turntable-renderer:
image: shumaione/turntable-renderer:latest
restart: unless-stopped
ports:
- '3003:3000'
environment:
TURNTABLE_RENDERER_API_BASIC_AUTH_USERNAME: admin
TURNTABLE_RENDERER_API_BASIC_AUTH_PASSWORD: your_password
```
Then point Shumai at it:
```bash
TURNTABLE_RENDERER_URL=http://localhost:3003
TURNTABLE_RENDERER_BASIC_AUTH_USERNAME=admin
TURNTABLE_RENDERER_BASIC_AUTH_PASSWORD=your_password
```
_(If both Shumai and turntable-renderer run within the same Docker Compose network, use `http://turntable-renderer:3000` as the URL)._
### GPU Acceleration
A dedicated GPU is optional — turntable-renderer's CPU rendering pipeline is fast enough for modest hardware. When a GPU is passed into the container, it is automatically detected and leveraged.
#### Intel / AMD (Open Source Drivers)
Pass the DRI render node:
```bash
docker run -d --rm \
--name turntable-renderer \
--shm-size=1gb \
-p 3003:3000 \
--device /dev/dri \
-e TURNTABLE_RENDERER_API_BASIC_AUTH_USERNAME=admin \
-e TURNTABLE_RENDERER_API_BASIC_AUTH_PASSWORD=your_password \
shumaione/turntable-renderer:latest
```
#### NVIDIA (Proprietary Drivers)
Requires host driver 580+ and the NVIDIA Container Toolkit (`--runtime nvidia`):
```bash
docker run -d --rm \
--name turntable-renderer \
--shm-size=1gb \
-p 3003:3000 \
--runtime nvidia \
--gpus all \
--device /dev/nvidia-modeset \
-e TURNTABLE_RENDERER_API_BASIC_AUTH_USERNAME=admin \
-e TURNTABLE_RENDERER_API_BASIC_AUTH_PASSWORD=your_password \
shumaione/turntable-renderer:latest
```
---
## Behavior Without Turntable Renderer
If turntable-renderer is not configured or unreachable:
- The 3D model upload completes successfully with status `processed`.
- No turntable video proxy, sprite, or poster is generated.
- The original raw 3D file is safely preserved and can be downloaded by team members.
+127
View File
@@ -0,0 +1,127 @@
---
title: Email Notifications
description: Configure team SMTP settings, deliver batched activity digest emails, and send test emails.
---
Shumai includes a built-in email notification system that delivers updates on team activities—such as comments, mentions, new asset uploads, status changes, and Kanban tasks—directly to team members' inboxes.
To prevent inbox clutter and notification fatigue, Shumai does not send emails for every single action individually. Instead, a background delivery service aggregates pending notifications and sends periodic, batched digest emails with inline thumbnails and deep links.
---
## Configuring SMTP Settings
Email notification settings are managed at the team level by **Team Owners**.
To configure outgoing mail:
1. Click on **Team Settings** in the sidebar.
2. Select the **Email Notifications** tab.
3. Turn on the **Enable Email Notifications** switch.
4. Fill in your SMTP server credentials and sender information.
5. Click **Save** or **Save and Send Test Email**.
### SMTP Server Settings
| Field | Description | Example / Default |
| :---------------------------- | :---------------------------------------------------------------------------------------------------------------------------------------------------------- | :-------------------------------------------------- |
| **SMTP Host** | The hostname or IP address of your outgoing SMTP server. | `smtp.example.com` or `smtp.sendgrid.net` |
| **SMTP Port** | The port used to connect to your SMTP server. | `587` (default, STARTTLS), `465` (SSL/TLS), or `25` |
| **Username** | The username or API key required to authenticate with the SMTP server. | `user@example.com` or `apikey` |
| **Password** | The authentication password or API secret. | `••••••••` |
| **SSL / TLS (SMTPS)** | Connect directly using SSL/TLS encryption. Typically required when connecting over port `465`. | `Disabled` (default) |
| **Ignore Certificate Errors** | Allows connections with self-signed, invalid, or unverified TLS certificates. Useful for local testing (e.g. Mailpit, MailHog) or private corporate relays. | `Disabled` (default) |
> [!TIP]
> **Password Retention**: When updating existing email settings, you can leave the password input blank to keep the currently stored password. For security, stored passwords are never returned in cleartext to the client.
### Sender Information
| Field | Description | Example / Default |
| :----------- | :-------------------------------------------------------------------------------------------------------------------------------------------------- | :----------------------------- |
| **From** | The sender name and email address that will appear in the `From` header. Ensure your SMTP provider is authorized to send on behalf of this address. | `Shumai <noreply@example.com>` |
| **Reply-To** | _(Optional)_ An email address to receive replies if members respond directly to a notification email. | `support@example.com` |
---
## Testing SMTP Delivery
Before relying on email delivery, you can verify your configuration and credentials directly from the settings page:
1. Click **Save and Send Test Email**.
2. In the **Send Test Email** modal, verify or adjust the sender `From` address.
3. Click **Send Test Email**.
Shumai will perform an active handshake verification (`transport.verify()`) against your SMTP server and dispatch a styled test message to the email address of the currently logged-in administrator.
If the handshake or delivery fails (for example, due to an invalid password, blocked outbound port, or rejected sender address), an error toast is displayed with the exact message returned by the SMTP server.
---
## Delivery Architecture & Digest Batching
Shumai uses an asynchronous, batched architecture for email delivery:
```
[ Team Activity ]
(Comments, Uploads, Mentions, Kanban)
│
▼
[ Database Queue ] (status: pending)
│
▼ (Every 3 minutes)
[ Notification Background Job ]
│
├─ Filter recipients (check user preferences & project scope)
├─ Exclude self-actions and AI agents
├─ Group & collapse items (clusters > 3 summarized)
├─ Fetch avatars & thumbnails as inline CID attachments
│
▼
[ Nodemailer Transport ] ──> SMTP Server ──> [ Member Inboxes ]
```
### 1. Periodic Background Delivery
Notifications are queued in PostgreSQL with a `pending` status. A background job runs periodically every **3 minutes**, acquiring pending notifications using atomic queue locking (`FOR UPDATE SKIP LOCKED`). This ensures safe delivery without duplicate emails even when multiple server instances run concurrently.
### 2. Digest Bundling
Rather than sending an email immediately for each action, all notifications collected for a recipient during the 3-minute window are combined into a single digest email. This keeps team communication organized and eliminates inbox spam during heavy collaboration sessions.
### 3. Smart Activity Collapsing
When an event cluster contains a high volume of items (for example, when an editor uploads dozens of video clips at once), Shumai automatically groups and collapses the items:
- If **3 or fewer** items are in a group, each item is displayed with its individual thumbnail and details.
- If **more than 3** items are uploaded together, Shumai summarizes them into a single concise notification (e.g., _"Alice uploaded 12 new files to Project Alpha"_) with a preview of the first three file names.
### 4. Inline CID Attachments
Avatars and video/image asset thumbnails are retrieved from storage and embedded directly into the email as inline MIME attachments (`cid:...`).
Unlike external image links, inline CID attachments:
- Display immediately in desktop and mobile email clients (such as Apple Mail, Gmail, and Outlook) without being blocked by "show remote content" security warnings.
- Do not require publicly accessible S3 URLs or reverse-proxy authentication to render.
### 5. Deep Linking
Every notification card in the email includes deep-linking action buttons (such as **View Asset**, **View Project**, or **View Task**) that navigate members straight to the corresponding resource in Shumai.
To ensure links point to your public domain instead of `localhost`, configure the `APP_URL` or `BETTER_AUTH_URL` environment variable on your Shumai server:
```bash
APP_URL=https://shumai.example.com
```
---
## Member Preferences & Filtering
Email delivery respects team permissions and member choices:
- **Personal Preferences**: Team members can independently choose which types of activities they receive by opening **Team Settings > Notifications**. If a user disables notifications for a category (e.g. general comments or status updates), Shumai will not include those items in their email digests.
- **Self-Action Suppression**: Members never receive notification emails for their own comments, uploads, or task updates (unless they explicitly enable the **Your Uploads** preference to be notified when their own media transcoding completes).
- **Project Scope**: Members whose access is restricted to specific projects only receive email digests for activities within those assigned projects, unless they are explicitly `@mentioned`.
- **AI Agents**: Automated AI agent accounts are automatically excluded from receiving email notifications.
+131
View File
@@ -0,0 +1,131 @@
---
title: Embedding Server Setup (EmbeddingGemma 2)
description: Run local, offline multimodal vector embeddings for semantic search using shumaiOne/embedding-server.
---
Shumai supports local, self-hosted multimodal vector embeddings for semantic search through [shumaiOne/embedding-server](https://github.com/shumaiOne/embedding-server) — a standalone Docker service powered by Google AI Edge's **LiteRT-LM** and **EmbeddingGemma 2 740M**.
The embedding server provides a unified 768-dimensional vector space across text, images, and video frames with an OpenAI-compatible HTTP API (`/v1/embeddings`). Because the model weights are pre-baked directly into the Docker image, the server operates completely offline and air-gapped without requiring external network access or third-party cloud API keys.
---
## Model Architecture & Specs
| Property | Value |
| :------------------------ | :------------------------------------------------------- |
| **Model** | `EmbeddingGemma 2 740M` (`google/embeddinggemma-2-740m`) |
| **Runtime** | Google AI Edge LiteRT-LM (Python 3.14 Debian slim) |
| **Vector Dimensionality** | 768 dimensions (multimodal: text, image, and video) |
| **Context Window** | 2,048 tokens |
| **Network Mode** | Completely offline capable (`--network none`) |
| **API Standard** | OpenAI-compatible `/v1/embeddings` |
---
## Video Processing & Sampling
For video assets, Shumai breaks videos into standardized segments and extracts keyframes:
- **Chunking**: Videos are divided into **60-second chunks** with a **5-second overlap** at segment boundaries to preserve narrative and visual context.
- **Frame Extraction Rate**: Shumai extracts **15 frames per minute** from each chunk.
- **Adaptive Frame Bounding**:
- For standard aspect ratios (1:1, 4:3, 3:4, 16:9, 9:16, 21:9), Shumai extracts 15 frames, which produces ~1,982 prompt tokens and stays safely within the 2,048-token context window.
- For non-standard or custom aspect ratios (e.g. 5:3, 1.4:1), Shumai automatically samples 14 frames (~1,990 tokens) to mathematically guarantee that prompt length never exceeds 2,048 tokens, preventing out-of-memory errors on the embedding server.
---
## Important: Changing Embedding Providers Breaks Existing Embeddings
> [!WARNING]
> **Vector spaces from different embedding models are completely incompatible.**
>
> If you change your embedding provider (for example, switching from Google Gemini's 1536-dimensional embeddings to EmbeddingGemma's 768-dimensional embeddings, or vice versa), **all existing vector embeddings in your database become invalid**.
>
> To protect production databases from accidental corruption and search failures, **Shumai intentionally does not provide a WebUI toggle to change the embedding provider**. The provider can only be configured by an administrator via environment variables (`EMBEDDING_PROVIDER` and `GEMMA_EMBEDDING_OPENAI_BASE_URL`).
---
## Regenerating Embeddings via CLI
When you do switch embedding providers or configure the embedding server on an existing Shumai deployment, you can regenerate embeddings for all existing media files using the administrative CLI command:
### Inside Docker
```bash
docker exec -it <shumai-container-name> shumai regenerate-embeddings
```
### Local / Bare Metal
```bash
shumai regenerate-embeddings
```
_(Alternatively, `bun apps/web/src/index.ts regenerate-embeddings`)_
### What the Command Does
1. **Scans Teams**: Checks all teams in your workspace. If a team does not have an active (enabled) embedding agent, it logs a warning and safely skips that team.
2. **Clears Existing Embeddings**: Purges all previous embeddings and cancels any pending/in-progress embedding tasks for eligible assets to ensure a clean transition to the new provider.
3. **Enqueues Fresh Tasks**: Queues new background `ai_embedding` tasks for all eligible assets (images, videos, and 3D assets with rendered previews) across active teams.
4. **Background Execution**: Background workers process the queued tasks according to your configured concurrency limits.
---
## Deployment & Configuration
### Docker Quickstart
To run the standalone embedding server container:
```bash
docker run -d \
--name shumai_embedding \
-p 9379:9379 \
shumaione/embedding-server:latest
```
### Docker Compose Setup
In Docker Compose environments, add `embedding-server` to your `docker-compose.yaml` and configure Shumai to connect to it over Docker's internal network:
```yaml
services:
embedding-server:
image: shumaione/embedding-server:latest
container_name: shumai_embedding
restart: unless-stopped
shumai:
image: shumaione/shumai:latest
environment:
# Embedding Server Configuration
EMBEDDING_PROVIDER: gemma-embedding-2-openai
GEMMA_EMBEDDING_OPENAI_BASE_URL: http://embedding-server:9379/v1
ENABLE_EMBEDDING_AGENT: 'true'
CONCURRENCY_AGENT: 1
depends_on:
- embedding-server
```
> [!NOTE]
> Since Shumai communicates with `embedding-server` over Docker's internal network, `embedding-server` does not need to expose any ports to the host machine.
### Environment Variables
| Variable | Required | Description | Default |
| :-------------------------------- | :---------- | :---------------------------------------------------------------------------------------------------------------------------- | :------------------------- |
| `EMBEDDING_PROVIDER` | Yes | Set to `gemma-embedding-2-openai` to use the local LiteRT-LM server. | `gemini-embedding-2` |
| `GEMMA_EMBEDDING_OPENAI_BASE_URL` | Yes | HTTP base URL of the embedding server (e.g. `http://embedding-server:9379/v1`). | `http://localhost:9379/v1` |
| `ENABLE_EMBEDDING_AGENT` | No | Automatically provisions and enables the system embedding agent for all teams on startup. | `false` |
| `CONCURRENCY_AGENT` | Recommended | Maximum concurrent AI tasks executed by workers. **Set to `1`** when running `embedding-server` to prevent high memory usage. | `5` |
---
## Hardware & Resource Requirements
- **Memory (RAM)**:
- We recommend at least **4 GB+ of dedicated system RAM** for the `embedding-server` container.
- Overall host system memory should be **4 GB+** when running the Lite+Embedding compose stack, and **8 GB+** for the full stack.
- **Concurrency**:
- Always set `CONCURRENCY_AGENT=1` in your Shumai configuration when using `embedding-server`. Parallel multimodal embedding generations consume significant memory and can cause Docker containers to be terminated by the OS Out-Of-Memory (OOM) killer on low-resource machines.
+25 -1
View File
@@ -32,7 +32,22 @@ When Gotenberg is configured, Shumai automatically generates a PDF proxy for the
## Configuration
Gotenberg support is opt-in and controlled by the following environment variables on the machine running the transcode workers:
Shumai supports configuring Gotenberg either through the **WebUI** or via **Environment Variables**.
### 1. WebUI (Team Settings → Integrations)
Team owners can configure Gotenberg directly from the web interface:
1. Navigate to **Team Settings** from the sidebar.
2. Select the **Integrations** tab.
3. Under **Gotenberg Document Converter**, enter the **Server URL** (e.g. `http://gotenberg:3000` or `http://localhost:3000`).
4. Enter the **Username** and **Password** if Basic Authentication is enabled on the service.
5. Click **Test Connection** to test communication with Gotenberg and verify its health.
6. Click **Save Changes**.
### 2. Environment Variables
Gotenberg can also be configured globally across the system using environment variables on the machine running Shumai or the transcode workers:
| Variable | Required | Description |
| ------------------------------- | --------------- | ---------------------------------------------------------------- |
@@ -40,6 +55,15 @@ Gotenberg support is opt-in and controlled by the following environment variable
| `GOTENBERG_BASIC_AUTH_USERNAME` | No | Username when Basic Authentication is enabled on Gotenberg. |
| `GOTENBERG_BASIC_AUTH_PASSWORD` | No | Password when Basic Authentication is enabled on Gotenberg. |
### Configuration Priority & Read-Only Mode
> [!IMPORTANT]
> **Environment variables take highest priority.** If `GOTENBERG_URL` is set in your environment, Shumai will always use the environment variable configuration.
>
> When configured via environment variables, the settings in **Team Settings → Integrations** will display a **Configured via environment variables (Read-only)** badge, and the input fields will be locked to prevent accidental changes or configuration drift.
### Deploying Gotenberg
### Docker Quickstart
You can start a Gotenberg container with a single command:
+8
View File
@@ -44,3 +44,11 @@ If you want to mute a specific project entirely (except for direct `@mentions` w
- The **Bell** icon in the navigation bar displays a red badge when you have unread notifications.
- Click the bell to open your notification feed panel.
- Click any notification card to mark it as read and jump directly to the relevant file and comment thread.
---
## Email Notifications
In addition to in-app notifications, Team Owners can configure outgoing SMTP settings to deliver batched activity digest emails to team members. The email delivery engine honors each member's individual notification preferences configured above.
See the [Email Notifications Guide](/using-shumai/email-notifications) for full SMTP setup instructions, test email workflows, and delivery architecture details.
+4 -5
View File
@@ -12,11 +12,10 @@ Traditional search relies on matching exact characters (e.g., searching "dog" wi
Semantic search translates text queries and file contents into conceptual profiles. Similar ideas and visuals are mapped near each other.
1. **Index Generation**: When you upload an asset, Shumai's background agent scans the file's visual elements, text contents, or descriptions, translating them into a concept profile.
- **Video Processing**: Shumai uses **gemini-embedding-2** for video indexing.
- **Chunking**: For long videos, Shumai breaks them into chunks to ensure high-quality indexing. By default, chunks are **60 seconds** long with a **5-second overlap** to ensure no context is lost at the boundaries.
- **Sampling**: The model processes a maximum of 32 frames per chunk: short chunks (≤32s) are sampled at 1 fps, while longer chunks are uniformly sampled to 32 frames.
- **Optimization**: To optimize processing, Shumai automatically transcodes each chunk to **1 FPS** and **removes the audio track** before sending it to the embedding agent.
- **Configuration**: Administrators can customize these settings using the `EMBEDDING_CHUNK_DURATION` and `EMBEDDING_CHUNK_OVERLAP` environment variables.
- **Multimodal Embedding Providers**: Shumai supports two embedding providers:
- **EmbeddingGemma 2 (Local)**: Powered by `shumaiOne/embedding-server` with 768-dimensional multimodal vectors running completely offline. Videos are sampled at **15 frames per minute** (14 frames for non-standard aspect ratios) per 60-second chunk. See the [Embedding Server Setup Guide](./embedding-server) for full details.
- **Gemini Embedding 2 (Cloud)**: Powered by Google's `gemini-embedding-2` API with 1536-dimensional vectors. Videos are chunked into 60-second segments (sampled up to 32 frames per chunk at 1 fps with audio stripped).
- **Chunking Configuration**: Administrators can customize chunk duration and overlap using the `EMBEDDING_CHUNK_DURATION` (default: 60s) and `EMBEDDING_CHUNK_OVERLAP` (default: 5s) environment variables.
2. **Concept Matching**: When you search with the "Semantic" toggle enabled, Shumai matches the conceptual meaning of your query against your assets.
3. **Result**: Searching "sunset at the beach" will successfully find a video file of ocean waves at dusk, even if the filename is `IMG_4029.mp4`.
+61 -15
View File
@@ -1,31 +1,77 @@
---
title: Media Transcoding Settings
description: Configure video transcode strategies and preview proxy generation.
title: Video Transcoding
description: Configure video transcode policies, MP4 and HLS streaming resolutions, hardware acceleration, and FFmpeg threads.
---
High-resolution camera formats and raw media files cannot be rendered natively in web browsers. Shumai transcodes these assets into web-friendly streaming formats in the background.
High-resolution camera formats and raw media files cannot be rendered natively in web browsers. Shumai transcodes video assets into web-friendly streaming and preview formats in the background.
Team Owners can customize video transcoding behavior from the **Team Settings > Video Transcoding** page.
Regardless of the selected transcoding strategy, Shumai's pipeline will **always** generate at least one lightweight `180p` proxy version of the raw file to ensure fast scrubbing, thumbnail generation, and instant previews.
---
## Video Transcode Strategies
## Transcode Policy
Team Owners can select how the transcoding engine handles uploaded files. Shumai evaluates the original file's dimensions against its standard target resolutions (2160p, 1080p, 720p, 540p, and 360p) to determine the output.
The **Transcode Policy** card configures both MP4 video transcoding and HLS adaptive streaming.
### 1. Best Match
### Playback and Download Behavior
- **Behavior**: Shumai checks the longest dimension (width or height) of the uploaded file. It then finds the highest standard resolution that is equal to or smaller than the original file, preventing unnecessary upscaling. If the original video is exceptionally small, it defaults to a minimum target of `360p`.
- **Output**: Generates exactly two web-optimized files: the universal `180p` proxy and the single calculated "best match" resolution.
- **Use Case**: This is the recommended option. It ensures web compatibility without wasting processing time and storage space on redundant lower-resolution versions.
Shumai uses a dual-format strategy for video delivery:
### 2. Transcode All
- **When HLS is disabled**: Both in-browser video playback and file downloads will use the generated MP4 file(s).
- **When HLS is enabled**: In-browser video playback uses HLS (HTTP Live Streaming) for smooth, adaptive bitrate playback, while file downloads will use the generated MP4 file(s).
- **Behavior**: Shumai calculates the "best match" resolution exactly as it does in the previous strategy. However, instead of stopping at one version, it processes the video into the best match _and_ every standard resolution smaller than it.
- **Output**: Generates a full cascade of files. For example, if the best match for an upload is `1080p`, Shumai will generate the `180p` proxy, `1080p`, `720p`, `540p`, and `360p` versions.
- **Use Case**: Ideal for environments requiring adaptive bitrate streaming, ensuring consistent playback performance across all client devices regardless of varying network conditions.
<Warning>
**Storage Usage Notice**: Enabling HLS generates multi-bitrate segment files (`.ts` / `.m4s`) and
playlist manifests in addition to standard MP4 proxies. This can significantly increase storage
consumption on your storage bucket.
</Warning>
![Transcode Settings](../transcode.webp)
### MP4 Transcoding
Generate web-compatible MP4 files for file downloads and standard video playback. Team Owners can select how MP4 transcodes are generated for uploaded files:
#### 1. Best Match
- **Behavior**: Shumai checks the longest dimension of the uploaded video and generates a single optimal resolution matching or smaller than the source quality, preventing unnecessary upscaling.
- **Output**: Generates the lightweight `180p` proxy and the calculated best-match MP4 file.
- **Use Case**: Recommended default. Minimizes processing time and storage space while ensuring universal playback and download compatibility.
#### 2. Multi Resolutions
- **Behavior**: Select specific target resolution ladders to generate: `480p`, `720p`, `1080p`, `1440p`, and `2160p`. Resolutions are only generated up to the source video resolution (content is never upscaled).
- **Output**: Generates the `180p` proxy plus all selected resolution ladders that do not exceed the source resolution.
- **Use Case**: Ideal when users need to download multiple specific resolutions (e.g. 720p for fast sharing and 1080p for review).
### HLS Streaming
When **HLS Streaming** is toggled on, Shumai produces an adaptive bitrate HLS package for uploaded videos:
- **Resolution Ladders**: Configure which resolutions to generate for HLS streaming (`480p`, `720p`, `1080p`, `1440p`, `2160p`). Ladders are only generated up to the original video resolution.
- **Player Integration**: The web player will automatically stream via HLS when available, dynamically switching bitrates based on client network bandwidth.
![Video Transcoding](../transcode.webp)
---
## Hardware Acceleration
To offload encoding workloads from CPUs to GPUs, Shumai supports hardware-accelerated video encoding:
- **Disabled (Software)**: Uses the standard CPU `libx264` software encoder with Constant Rate Factor (CRF).
- **Auto Detect**: Automatically probes and leverages available GPU/hardware acceleration (NVIDIA NVENC, Intel QuickSync, AMD/Intel VA-API, Apple VideoToolbox, Rockchip RKMPP).
See the [Hardware Transcoding Guide](/configuration/hardware-transcoding) for Docker Compose configuration, GPU passthrough, and driver prerequisites.
---
## FFmpeg Threads
The **FFmpeg Threads** slider controls the maximum number of CPU threads used by FFmpeg during transcoding tasks:
- **Auto (0)**: FFmpeg automatically selects the optimal thread count based on host CPU core availability.
- **Manual (1 - 32)**: Restricts FFmpeg to a fixed number of threads, useful for throttling CPU consumption on shared worker nodes.
---
@@ -33,4 +79,4 @@ Team Owners can select how the transcoding engine handles uploaded files. Shumai
Transcoding is resource-intensive. Depending on your scaling needs, your administrator can run the transcode engine on the main server (Single-Server mode) or delegate tasks to specialized transcoding worker machines (Temporal Worker mode).
Note: In all modes, the processing machine must have the standard system media utilities (`ffmpeg` and `ffprobe`) installed.
In all modes, the processing machine must have system media utilities (`ffmpeg` and `ffprobe`) installed. See the [Temporal Deployment Guide](/configuration/temporal) for setting up distributed transcoding workers.
+5 -2
View File
@@ -1,6 +1,6 @@
{
"name": "shumai",
"version": "0.4.6",
"version": "0.5.3",
"private": true,
"type": "module",
"repository": {
@@ -38,7 +38,10 @@
"release:local": "bun run scripts/release-local.ts",
"release:patch": "bun run scripts/release.ts patch",
"release:minor": "bun run scripts/release.ts minor",
"release:major": "bun run scripts/release.ts major"
"release:major": "bun run scripts/release.ts major",
"changelog:extract": "bun run scripts/extract-changelog.ts",
"docs:changelog": "bun run scripts/generate-docs-changelog.ts",
"docs:changelog:check": "bun run scripts/generate-docs-changelog.ts --check"
},
"devDependencies": {
"@eslint/js": "10.0.1",
+3 -2
View File
@@ -1,6 +1,6 @@
{
"name": "@shumai/agent",
"version": "0.4.6",
"version": "0.5.3",
"private": true,
"type": "module",
"main": "./src/index.ts",
@@ -9,7 +9,7 @@
"@anthropic-ai/sandbox-runtime": "0.0.74",
"@earendil-works/pi-agent-core": "0.82.1",
"@earendil-works/pi-ai": "0.82.1",
"@google/genai": "2.19.0",
"@google/genai": "2.28.0",
"@shumai/core": "workspace:*",
"@shumai/db": "workspace:*",
"@shumai/dtos": "workspace:*",
@@ -19,6 +19,7 @@
"@temporalio/workflow": "1.23.0",
"adm-zip": "0.6.0",
"jittered-fractional-indexing": "1.0.1",
"sharp": "0.35.4",
"typebox": "1.3.22",
"ulid": "3.0.2",
"zod": "4.5.4"
+132
View File
@@ -662,6 +662,23 @@ describe('Agent Activities', () => {
const pdfPrompt = mockHarness.prompt.mock.calls[1]?.[0] ?? ''
expect(pdfPrompt).toContain('Document pages: 50')
expect(pdfPrompt).not.toContain('Video duration')
// For a 3D asset, should produce 3D Model recommendation
await autofillAiActivity({
teamId: 't1',
assetId: 'model-3',
assetName: 'robot.glb',
mediaType: 'model/gltf-binary',
images: [],
fields: [{ id: 'f1', config: { name: 'F1', type: 'text' } }],
context,
})
const modelPrompt = mockHarness.prompt.mock.calls[2]?.[0] ?? ''
expect(modelPrompt).toContain('3D Model: This asset has an orthogonal 4-view composite image')
expect(modelPrompt).toContain(
'Front view at top-left, Right side view at top-right, Back view at bottom-left, Left side view at bottom-right',
)
})
})
@@ -1170,6 +1187,43 @@ describe('Agent Database Activities Integration', () => {
expect((rootEntry as any)?.details?.thread).toBeUndefined()
})
it('should set media position to degree for 3D asset comments', async () => {
const asset3d = await prisma.asset.create({
data: {
name: 'robot.glb',
type: 'file',
status: 'processed',
mediaType: 'model/gltf-binary',
media: { proxyType: '3d' } as unknown as PrismaJson.MediaInfo,
projectId: project.id,
},
})
const comment3d = await prisma.assetComment.create({
data: {
assetId: asset3d.id,
message: 'Nice angle',
creatorId: user.id,
second: 0.5, // 45°
},
})
const sessionId = await initializeAgentSessionActivity({
teamId: team.id,
agentId: 'test-agent-id',
userCommentId: comment3d.id,
userId: user.id,
})
const storage = new DatabaseSessionStorage(sessionId)
const pathEntries = await storage.getPathToRoot(comment3d.id)
const entry = pathEntries.find((e) => e.id === comment3d.id)
expect(entry).toBeDefined()
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- test entry position parsing
expect((entry as any)?.details?.position).toEqual({ type: 'degree', degree: 45 })
})
it('should return user name and role for team member', async () => {
const info = await getUserTeamInfoActivity({
userId: user.id,
@@ -2105,5 +2159,83 @@ describe('Agent Database Activities Integration', () => {
expect(systemPromptArg).toContain('<thread id="..." reply_count="..." />')
expect(systemPromptArg).toContain('read_thread')
})
it('agentChatActivity injects # 3D Model Inspection into systemPrompt', async () => {
const user = await prisma.user.create({
data: { name: 'User_3d_Prompt', email: `prompt-3d-${Date.now()}@example.com` },
})
const team = await prisma.team.create({
data: { name: 'Team_3d_Prompt_' + Date.now() },
})
const project = await prisma.project.create({
data: { name: 'Project_3d_Prompt', teamId: team.id },
})
const asset = await prisma.asset.create({
data: {
name: 'robot.glb',
type: 'file',
status: 'processed',
mediaType: 'model/gltf-binary',
media: { proxyType: '3d' } as unknown as PrismaJson.MediaInfo,
projectId: project.id,
},
})
const comment = await prisma.assetComment.create({
data: {
assetId: asset.id,
creatorId: user.id,
message: 'Check this 3d model',
},
})
const mockSession = {
id: 'mock-session-id',
subscribe: vi.fn(),
getEntries: vi.fn().mockResolvedValue([]),
getStorage: vi.fn().mockReturnValue({ sessionId: 'mock-session-id' }),
}
const mockHarness = {
subscribe: vi.fn(),
prompt: vi.fn().mockResolvedValue({
content: [{ type: 'text', text: 'Response' }],
usage: { input: 10, output: 20 },
}),
}
vi.mocked(piAgent.createAgentSession).mockResolvedValue({
session: mockSession as unknown as Session<DatabaseSessionMetadata>,
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- Mock AgentHarness instance for activity test
harness: mockHarness as unknown as AgentHarness<any, any, any, any>,
})
const context = {
agent: { id: 'agent-1', provider: { name: 'google' }, modelRef: { modelId: 'gemini' } },
dbProviders: [],
teamSkills: [],
allowedDomains: [],
} as unknown as AgentExecutionContext
await agentChatActivity({
teamId: team.id,
agentId: 'agent-1',
message: 'Hello',
imageUrls: [],
projectId: project.id,
folderId: asset.id,
assetId: asset.id,
sessionId: 'mock-session-id',
userCommentId: comment.id,
userId: user.id,
context,
})
const lastCall = vi.mocked(piAgent.createAgentSession).mock.calls.at(-1)
const systemPromptArg = lastCall?.[0].systemPrompt as string
expect(systemPromptArg).toContain('# 3D Model Inspection')
expect(systemPromptArg).toContain('media_type="3d"')
expect(systemPromptArg).toContain('config3d')
expect(systemPromptArg).toContain('<position type="degree" degree="..." />')
})
})
})
+21 -3
View File
@@ -21,6 +21,7 @@ import {
import { aiUsageService } from '@shumai/core/src/ai-usage/ai-usage'
import { quotaService, QuotaExceededError } from '@shumai/core/src/quota/quota-service'
import { getProxyType } from '@shumai/core/src/utils/mime'
import { secondToDegree } from '@shumai/core/src/utils/model3d'
import {
UpdateAssetMetadataRequest,
type ShumaiMessageContext,
@@ -179,6 +180,8 @@ User messages may contain a <context> block detailing the user, active asset loc
systemPrompt += `\n\n# Comment Threads\nWhen operating in comment mode, previous top-level messages in the conversation history may contain a <thread id="..." reply_count="..." /> tag indicating an earlier discussion thread with replies. If a user's question refers to or depends on earlier comments or discussions, you can use the 'read_thread' tool with the thread ID to inspect all replies in that thread.`
}
systemPrompt += `\n\n# 3D Model Inspection\nWhen interacting with 3D model assets (media_type="3d"), you can visually inspect the asset using 'read_asset'. Calling 'read_asset' with config3d: null (or config3d: {}) returns an orthogonal 4-view composite image (Front at top-left, Right side at top-right, Back at bottom-left, Left side at bottom-right). If a <position type="degree" degree="..." /> tag is present or a specific angle is discussed, pass 'config3d: { degree: ... }' (integer 0, 15, 30, ..., 345) to inspect that view angle. If an <annotation id="..." /> tag is present on a 3D asset, pass 'annotationId' to inspect the angle with the user's visual markup drawings overlaid.`
if (agent.type === 'autofill') {
systemPrompt += `\n\n# Autofill Instructions\nWhen performing metadata autofill, inspect the asset content using the 'read_asset' tool. If 'read_asset' fails or returns an error, do not autofill any metadata and report to the user directly. Do not try anything fancy, for example downloading the file or using python/ffmpeg to extract content.`
}
@@ -564,6 +567,7 @@ export async function autofillAiActivity(params: AutofillAiParams) {
const proxyType = getProxyType(params.mediaType, params.assetName)
const isVideoOrAudio = proxyType === 'video' || proxyType === 'audio'
const isPdf = proxyType === 'pdf'
const is3d = proxyType === '3d'
if (params.duration !== undefined && params.duration > 0 && (!proxyType || isVideoOrAudio)) {
promptLines.push(
@@ -575,11 +579,16 @@ export async function autofillAiActivity(params: AutofillAiParams) {
`- Document pages: ${params.pageCount}. Recommended inspection: call read_asset with docConfig: { mode: "pages", startPage: 1, endPage: ${Math.min(params.pageCount, 20)} } or mode: "text".`,
)
}
if (is3d) {
promptLines.push(
'- 3D Model: This asset has an orthogonal 4-view composite image (2x2 grid showing Front view at top-left, Right side view at top-right, Back view at bottom-left, Left side view at bottom-right). Call read_asset to inspect the 4 views.',
)
}
promptLines.push(
'',
'Instructions:',
`1. Call the "read_asset" tool with assetId: "${params.assetId || ''}" to inspect the asset content (use imageConfig, videoConfig, or docConfig according to the media type).`,
`1. Call the "read_asset" tool with assetId: "${params.assetId || ''}" to inspect the asset content (use imageConfig, videoConfig, or docConfig according to the media type; for 3D assets, call read_asset directly to inspect the 4-view composite image).`,
'2. After analyzing the retrieved content, call the "autofill_metadata" tool to provide the extracted metadata values.',
'3. If "read_asset" fails or returns an error, do not autofill any metadata and report to the user directly. Do not try anything fancy, for example downloading the file or using python/ffmpeg to extract content.',
)
@@ -834,7 +843,9 @@ export async function initializeAgentSessionActivity(params: {
let position: ShumaiMediaPosition | undefined = undefined
if (c.second !== null && c.second !== undefined) {
if (proxyType === 'pdf') {
if (proxyType === '3d') {
position = { type: 'degree', degree: secondToDegree(c.second) }
} else if (proxyType === 'pdf') {
position = { type: 'page', page: Math.round(c.second) }
} else {
position = { type: 'time', seconds: c.second }
@@ -852,7 +863,14 @@ export async function initializeAgentSessionActivity(params: {
id: att.asset.id,
name: att.asset.name,
type: att.asset.type,
mediaType: attProxyType as 'image' | 'video' | 'pdf' | 'audio' | 'other' | undefined,
mediaType: attProxyType as
| 'image'
| 'video'
| 'pdf'
| 'audio'
| '3d'
| 'other'
| undefined,
mimeType: att.asset.mediaType || undefined,
})
}
+493 -30
View File
@@ -1,17 +1,19 @@
import { describe, it, expect, vi, beforeEach } from 'vitest'
import { describe, it, expect, vi, beforeEach, afterEach } from 'vitest'
import {
generateTextEmbeddingActivity,
extractAiMetadataActivity,
getEmbeddingContextActivity,
saveAssetEmbeddingsActivity,
deleteAssetEmbeddingsActivity,
generateImageEmbeddingActivity,
generateVideoChunkEmbeddingActivity,
generateEmbeddingActivity,
getGemmaEmbeddingTargetFrameCount,
} from './ai'
import { s3Service } from '@shumai/core/src/s3/s3'
import { transcodeService } from '@shumai/core'
import { prisma, AssetType, AssetStatus } from '@shumai/db'
import { setupTestDbHooks } from '@shumai/db/test'
import sharp from 'sharp'
vi.mock('fs', async (importOriginal) => {
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- importOriginal returns the raw fs module, typed dynamically as any here
@@ -54,7 +56,6 @@ vi.mock('child_process', () => {
const execFileMock: any = vi.fn((file: any, args: any, cb: any) => {
cb(null, '120.0\n', '')
})
// Setup custom promisify behavior to return { stdout, stderr }
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- mock functions dynamically return varying types based on test command calls
execFileMock[Symbol.for('nodejs.util.promisify.custom')] = vi.fn(async (file: any) => {
if (file === 'ffprobe') {
@@ -90,6 +91,14 @@ describe('AI Activities Unit Tests', () => {
beforeEach(() => {
vi.clearAllMocks()
process.env.GEMINI_API_KEY = 'test-key'
delete process.env.EMBEDDING_PROVIDER
delete process.env.GEMMA_EMBEDDING_OPENAI_BASE_URL
})
afterEach(() => {
vi.unstubAllGlobals()
delete process.env.EMBEDDING_PROVIDER
delete process.env.GEMMA_EMBEDDING_OPENAI_BASE_URL
})
it('should generate image embedding', async () => {
@@ -126,33 +135,6 @@ describe('AI Activities Unit Tests', () => {
)
})
it('should resolve video input via resolveInput and avoid downloadToFile in generateEmbeddingActivity', async () => {
vi.mocked(s3Service.downloadToFile).mockClear()
vi.mocked(s3Service.resolveInput).mockClear()
vi.mocked(s3Service.resolveInput).mockResolvedValue(
'https://mock-storage/files/asset-123/video.mp4',
)
const res = await generateEmbeddingActivity({
teamId: 't1',
assetId: 'asset-123',
context: {
agent: {},
asset: {
mediaType: 'video/mp4',
name: 'sample.mp4',
storageKey: { key: 'files/asset-123/video.mp4' },
media: { proxyType: 'video' },
},
},
})
expect(s3Service.resolveInput).toHaveBeenCalledWith('shumai', 'files/asset-123/video.mp4')
expect(s3Service.downloadToFile).not.toHaveBeenCalled()
expect(s3Service.getObject).not.toHaveBeenCalledWith('shumai', 'files/asset-123/video.mp4')
expect(res.embeddings.length).toBeGreaterThan(0)
})
it('should generate text embedding successfully', async () => {
const res = await generateTextEmbeddingActivity({
text: 'Search query',
@@ -172,6 +154,356 @@ describe('AI Activities Unit Tests', () => {
)
})
it('should generate text embedding successfully with gemma-embedding-2-openai', async () => {
process.env.EMBEDDING_PROVIDER = 'gemma-embedding-2-openai'
const mockFetch = vi.fn().mockResolvedValue({
ok: true,
json: async () => ({
model: 'embeddinggemma-2',
data: [{ embedding: Array(768).fill(0.1) }],
// eslint-disable-next-line @typescript-eslint/naming-convention
usage: { prompt_tokens: 15 },
}),
})
vi.stubGlobal('fetch', mockFetch)
const res = await generateTextEmbeddingActivity({
text: 'Search query',
teamId: 't1',
})
expect(res.embedding).toHaveLength(768)
expect(res.usage.model).toBe('embeddinggemma-2')
expect(res.usage.inputTokens).toBe(15)
expect(mockFetch).toHaveBeenCalledWith(
'http://localhost:9379/v1/embeddings',
expect.objectContaining({
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
model: 'embeddinggemma-2',
input: 'task: search query | text: Search query',
}),
}),
)
})
it('should generate image embedding successfully with gemma-embedding-2-openai', async () => {
process.env.EMBEDDING_PROVIDER = 'gemma-embedding-2-openai'
const mockFetch = vi.fn().mockResolvedValue({
ok: true,
json: async () => ({
model: 'embeddinggemma-2',
data: [{ embedding: Array(768).fill(0.2) }],
// eslint-disable-next-line @typescript-eslint/naming-convention
usage: { prompt_tokens: 140 },
}),
})
vi.stubGlobal('fetch', mockFetch)
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: Buffer.from('test-image'),
contentType: 'image/png',
} as unknown as { buffer: Buffer; contentType: string })
const res = await generateImageEmbeddingActivity({
teamId: 't1',
assetKey: 'img.png',
mediaType: 'image/png',
})
expect(res.embedding).toHaveLength(768)
expect(res.usage.model).toBe('embeddinggemma-2')
expect(res.usage.inputTokens).toBe(140)
expect(mockFetch).toHaveBeenCalledWith(
'http://localhost:9379/v1/embeddings',
expect.objectContaining({
method: 'POST',
body: JSON.stringify({
model: 'embeddinggemma-2',
input: [
{
type: 'image_url',
// eslint-disable-next-line @typescript-eslint/naming-convention
image_url: {
url: `data:image/png;base64,${Buffer.from('test-image').toString('base64')}`,
},
},
],
}),
}),
)
})
it('should respect custom GEMMA_EMBEDDING_OPENAI_BASE_URL with or without /v1 suffix', async () => {
process.env.EMBEDDING_PROVIDER = 'gemma-embedding-2-openai'
const mockFetch = vi.fn().mockResolvedValue({
ok: true,
json: async () => ({
model: 'embeddinggemma-2',
data: [{ embedding: Array(768).fill(0.2) }],
// eslint-disable-next-line @typescript-eslint/naming-convention
usage: { prompt_tokens: 10 },
}),
})
vi.stubGlobal('fetch', mockFetch)
// Test with /v1
process.env.GEMMA_EMBEDDING_OPENAI_BASE_URL = 'http://custom-host:9379/v1'
await generateTextEmbeddingActivity({ text: 'test query', teamId: 't1' })
expect(mockFetch).toHaveBeenLastCalledWith(
'http://custom-host:9379/v1/embeddings',
expect.anything(),
)
// Test without /v1
process.env.GEMMA_EMBEDDING_OPENAI_BASE_URL = 'http://custom-host:9379'
await generateTextEmbeddingActivity({ text: 'test query', teamId: 't1' })
expect(mockFetch).toHaveBeenLastCalledWith(
'http://custom-host:9379/v1/embeddings',
expect.anything(),
)
})
it('should convert webp image to jpeg with gemma-embedding-2-openai', async () => {
process.env.EMBEDDING_PROVIDER = 'gemma-embedding-2-openai'
const mockFetch = vi.fn().mockResolvedValue({
ok: true,
json: async () => ({
model: 'embeddinggemma-2',
data: [{ embedding: Array(768).fill(0.2) }],
// eslint-disable-next-line @typescript-eslint/naming-convention
usage: { prompt_tokens: 140 },
}),
})
vi.stubGlobal('fetch', mockFetch)
const webpBuffer = await sharp({
create: { width: 10, height: 10, channels: 3, background: { r: 255, g: 0, b: 0 } },
})
.webp()
.toBuffer()
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: webpBuffer,
contentType: 'image/webp',
} as unknown as { buffer: Buffer; contentType: string })
const res = await generateImageEmbeddingActivity({
teamId: 't1',
assetKey: 'img.webp',
mediaType: 'image/webp',
})
expect(res.embedding).toHaveLength(768)
expect(mockFetch).toHaveBeenCalledWith(
'http://localhost:9379/v1/embeddings',
expect.objectContaining({
method: 'POST',
body: expect.stringContaining('data:image/jpeg;base64,'),
}),
)
})
it('should generate video chunk embedding from tar archive with gemma-embedding-2-openai', async () => {
process.env.EMBEDDING_PROVIDER = 'gemma-embedding-2-openai'
const mockFetch = vi.fn().mockResolvedValue({
ok: true,
json: async () => ({
model: 'embeddinggemma-2',
data: [{ embedding: Array(768).fill(0.3) }],
// eslint-disable-next-line @typescript-eslint/naming-convention
usage: { prompt_tokens: 140 },
}),
})
vi.stubGlobal('fetch', mockFetch)
const sampleArchive = new Bun.Archive({
'frame_0001.jpg': Buffer.from('frame-1'),
})
const archiveBytes = await sampleArchive.bytes()
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: Buffer.from(archiveBytes.buffer, archiveBytes.byteOffset, archiveBytes.byteLength),
contentType: 'application/x-tar',
} as unknown as { buffer: Buffer; contentType: string })
const res = await generateVideoChunkEmbeddingActivity({
teamId: 't1',
chunkKey: 'files/asset-1/tmp-embedding-chunks/frames-0-60.tar',
})
expect(res.embedding).toHaveLength(768)
expect(res.usage.model).toBe('embeddinggemma-2')
expect(mockFetch).toHaveBeenCalled()
})
describe('getGemmaEmbeddingTargetFrameCount', () => {
it('should return 15 for standard aspect ratios', () => {
// 16:9
expect(getGemmaEmbeddingTargetFrameCount(1920, 1080)).toBe(15)
expect(getGemmaEmbeddingTargetFrameCount(1280, 720)).toBe(15)
expect(getGemmaEmbeddingTargetFrameCount(854, 480)).toBe(15)
// 9:16
expect(getGemmaEmbeddingTargetFrameCount(1080, 1920)).toBe(15)
// 1:1
expect(getGemmaEmbeddingTargetFrameCount(1080, 1080)).toBe(15)
expect(getGemmaEmbeddingTargetFrameCount(480, 480)).toBe(15)
// 4:3
expect(getGemmaEmbeddingTargetFrameCount(1440, 1080)).toBe(15)
expect(getGemmaEmbeddingTargetFrameCount(640, 480)).toBe(15)
// 3:4
expect(getGemmaEmbeddingTargetFrameCount(1080, 1440)).toBe(15)
// 21:9
expect(getGemmaEmbeddingTargetFrameCount(2560, 1080)).toBe(15)
expect(getGemmaEmbeddingTargetFrameCount(3440, 1440)).toBe(15)
})
it('should return 14 for non-standard aspect ratios and missing/invalid dimensions', () => {
expect(getGemmaEmbeddingTargetFrameCount(1000, 600)).toBe(14) // 5:3
expect(getGemmaEmbeddingTargetFrameCount(700, 500)).toBe(14) // 1.4:1
expect(getGemmaEmbeddingTargetFrameCount(854, 672)).toBe(14) // ~1.27:1
expect(getGemmaEmbeddingTargetFrameCount(undefined, undefined)).toBe(14)
expect(getGemmaEmbeddingTargetFrameCount(0, 0)).toBe(14)
expect(getGemmaEmbeddingTargetFrameCount(-100, 100)).toBe(14)
})
})
it('should sample uniform 15 frames when chunk archive contains standard 16:9 frames', async () => {
process.env.EMBEDDING_PROVIDER = 'gemma-embedding-2-openai'
const mockFetch = vi.fn().mockResolvedValue({
ok: true,
json: async () => ({
model: 'embeddinggemma-2',
data: [{ embedding: Array(768).fill(0.3) }],
// eslint-disable-next-line @typescript-eslint/naming-convention
usage: { prompt_tokens: 140 },
}),
})
vi.stubGlobal('fetch', mockFetch)
const sampleJpg = await sharp({
create: {
width: 160,
height: 90,
channels: 3,
background: { r: 0, g: 0, b: 0 },
},
})
.jpeg()
.toBuffer()
const frames: Record<string, Buffer> = {}
for (let i = 1; i <= 30; i++) {
frames[`frame_${String(i).padStart(4, '0')}.jpg`] = sampleJpg
}
const sampleArchive = new Bun.Archive(frames)
const archiveBytes = await sampleArchive.bytes()
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: Buffer.from(archiveBytes.buffer, archiveBytes.byteOffset, archiveBytes.byteLength),
contentType: 'application/x-tar',
} as unknown as { buffer: Buffer; contentType: string })
const res = await generateVideoChunkEmbeddingActivity({
teamId: 't1',
chunkKey: 'files/asset-1/tmp-embedding-chunks/frames-0-60.tar',
})
expect(res.embedding).toHaveLength(768)
const requestBody = JSON.parse(mockFetch.mock.calls[0][1].body)
expect(requestBody.input).toHaveLength(15)
})
it('should fallback to 14 frames when chunk archive contains non-standard or unparseable frames', async () => {
process.env.EMBEDDING_PROVIDER = 'gemma-embedding-2-openai'
const mockFetch = vi.fn().mockResolvedValue({
ok: true,
json: async () => ({
model: 'embeddinggemma-2',
data: [{ embedding: Array(768).fill(0.3) }],
// eslint-disable-next-line @typescript-eslint/naming-convention
usage: { prompt_tokens: 140 },
}),
})
vi.stubGlobal('fetch', mockFetch)
const frames: Record<string, Buffer> = {}
for (let i = 1; i <= 30; i++) {
frames[`frame_${String(i).padStart(4, '0')}.jpg`] = Buffer.from(`frame-${i}`)
}
const sampleArchive = new Bun.Archive(frames)
const archiveBytes = await sampleArchive.bytes()
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: Buffer.from(archiveBytes.buffer, archiveBytes.byteOffset, archiveBytes.byteLength),
contentType: 'application/x-tar',
} as unknown as { buffer: Buffer; contentType: string })
const res = await generateVideoChunkEmbeddingActivity({
teamId: 't1',
chunkKey: 'files/asset-1/tmp-embedding-chunks/frames-0-60.tar',
})
expect(res.embedding).toHaveLength(768)
const requestBody = JSON.parse(mockFetch.mock.calls[0][1].body)
expect(requestBody.input).toHaveLength(14)
})
it('should throw non-retryable ApplicationFailure when chunk archive contains no valid frames', async () => {
process.env.EMBEDDING_PROVIDER = 'gemma-embedding-2-openai'
const sampleArchive = new Bun.Archive({
'info.txt': Buffer.from('not an image'),
})
const archiveBytes = await sampleArchive.bytes()
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: Buffer.from(archiveBytes.buffer, archiveBytes.byteOffset, archiveBytes.byteLength),
contentType: 'application/x-tar',
} as unknown as { buffer: Buffer; contentType: string })
await expect(
generateVideoChunkEmbeddingActivity({
teamId: 't1',
chunkKey: 'files/asset-1/tmp-embedding-chunks/frames-0-60.tar',
}),
).rejects.toThrow('No extracted frames found in chunk archive')
})
it('should throw non-retryable ApplicationFailure when chunk archive data is corrupted', async () => {
process.env.EMBEDDING_PROVIDER = 'gemma-embedding-2-openai'
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: Buffer.from('corrupted non-tar bytes'),
contentType: 'application/x-tar',
} as unknown as { buffer: Buffer; contentType: string })
await expect(
generateVideoChunkEmbeddingActivity({
teamId: 't1',
chunkKey: 'files/asset-1/tmp-embedding-chunks/frames-0-60.tar',
}),
).rejects.toThrow('Failed to read chunk archive')
})
it('should throw non-retryable ApplicationFailure when an extracted frame is empty (0 bytes)', async () => {
process.env.EMBEDDING_PROVIDER = 'gemma-embedding-2-openai'
const sampleArchive = new Bun.Archive({
'frame_0001.jpg': Buffer.alloc(0),
})
const archiveBytes = await sampleArchive.bytes()
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: Buffer.from(archiveBytes.buffer, archiveBytes.byteOffset, archiveBytes.byteLength),
contentType: 'application/x-tar',
} as unknown as { buffer: Buffer; contentType: string })
await expect(
generateVideoChunkEmbeddingActivity({
teamId: 't1',
chunkKey: 'files/asset-1/tmp-embedding-chunks/frames-0-60.tar',
}),
).rejects.toThrow('is empty (0 bytes)')
})
describe('extractAiMetadataActivity', () => {
it('should return existing images if they are already in S3 storage', async () => {
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- mock returned object cast as any due to partial implementation
@@ -331,6 +663,18 @@ describe('AI Database Activities Integration', () => {
expect(ctx.agent.id).toBe(user.id)
expect(ctx.asset.id).toBe(asset.id)
expect(ctx.asset.mediaType).toBe('video/mp4')
process.env.ENABLE_EMBEDDING_AGENT = 'false'
try {
await expect(
getEmbeddingContextActivity({
teamId: team.id,
assetId: asset.id,
}),
).rejects.toThrow('embedding feature is disabled or agent not found')
} finally {
delete process.env.ENABLE_EMBEDDING_AGENT
}
})
it('should parse valid positive chunk duration and fallback on invalid or negative ones', async () => {
@@ -389,6 +733,45 @@ describe('AI Database Activities Integration', () => {
delete process.env.EMBEDDING_CHUNK_DURATION
delete process.env.EMBEDDING_CHUNK_OVERLAP
})
it('should return existing embeddings for the asset', async () => {
// Ensure agent exists
await prisma.agent.upsert({
where: { id: user.id },
update: {},
create: {
id: user.id,
teamId: team.id,
type: 'embedding',
enabled: true,
config: { provider: 'openai', model: 'gpt-4' },
},
})
await saveAssetEmbeddingsActivity({
assetId: asset.id,
embeddings: [
{ embedding: Array(1536).fill(0.01), startTime: 0.0, endTime: 60.0 },
{ embedding: Array(768).fill(0.02), startTime: 55.0, endTime: 115.0 },
],
})
const ctx = await getEmbeddingContextActivity({
teamId: team.id,
assetId: asset.id,
})
expect(ctx.existingEmbeddings).toHaveLength(2)
const chunk1 = ctx.existingEmbeddings.find((e) => e.startTime === 0)
expect(chunk1).toBeDefined()
expect(chunk1?.has1536).toBe(true)
expect(chunk1?.has768).toBe(false)
const chunk2 = ctx.existingEmbeddings.find((e) => e.startTime === 55)
expect(chunk2).toBeDefined()
expect(chunk2?.has1536).toBe(false)
expect(chunk2?.has768).toBe(true)
})
})
describe('saveAssetEmbeddingsActivity', () => {
@@ -423,5 +806,85 @@ describe('AI Database Activities Integration', () => {
expect(dbEmbeddings[0].endTime).toBe(60.0)
expect(dbEmbeddings[1].startTime).toBeNull()
})
it('should save 768-dim embeddings to embedding_768 column in database', async () => {
const embeddings = [
{
embedding: Array(768).fill(0.01),
startTime: 0.0,
endTime: 60.0,
},
{
embedding: Array(768).fill(0.02),
},
]
await saveAssetEmbeddingsActivity({
assetId: asset.id,
embeddings,
})
// Query raw embeddings from DB
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- db query output is a raw postgres result array typed as any
const dbEmbeddings: any[] = await prisma.$queryRaw`
SELECT id, asset_id as "assetId", (embedding_768 IS NOT NULL) as "has768", (embedding IS NOT NULL) as "has1536"
FROM asset_embeddings
WHERE asset_id = ${asset.id}
`
expect(dbEmbeddings.length).toBe(2)
expect(dbEmbeddings[0].has768).toBe(true)
expect(dbEmbeddings[0].has1536).toBe(false)
expect(dbEmbeddings[1].has768).toBe(true)
expect(dbEmbeddings[1].has1536).toBe(false)
})
})
describe('deleteAssetEmbeddingsActivity', () => {
it('should delete specific embeddings by id', async () => {
await saveAssetEmbeddingsActivity({
assetId: asset.id,
embeddings: [
{ embedding: Array(1536).fill(0.01), startTime: 0, endTime: 60 },
{ embedding: Array(1536).fill(0.02), startTime: 55, endTime: 115 },
],
})
const before = await prisma.assetEmbedding.findMany({
where: { assetId: asset.id },
})
expect(before).toHaveLength(2)
const targetId = before[0].id
await deleteAssetEmbeddingsActivity({
assetId: asset.id,
ids: [targetId],
})
const after = await prisma.assetEmbedding.findMany({
where: { assetId: asset.id },
})
expect(after).toHaveLength(1)
expect(after[0].id).not.toBe(targetId)
})
it('should delete all embeddings for an asset when ids is not provided', async () => {
await saveAssetEmbeddingsActivity({
assetId: asset.id,
embeddings: [
{ embedding: Array(1536).fill(0.01), startTime: 0, endTime: 60 },
{ embedding: Array(1536).fill(0.02), startTime: 55, endTime: 115 },
],
})
await deleteAssetEmbeddingsActivity({
assetId: asset.id,
})
const after = await prisma.assetEmbedding.findMany({
where: { assetId: asset.id },
})
expect(after).toHaveLength(0)
})
})
})
+264 -167
View File
@@ -1,17 +1,11 @@
import { GoogleGenAI } from '@google/genai'
import { s3Service } from '@shumai/core/src/s3/s3'
import { transcodeService } from '@shumai/core'
import { execFile } from 'child_process'
import * as fs from 'fs'
import * as os from 'os'
import * as path from 'path'
import { promisify } from 'util'
import { ApplicationFailure } from '@temporalio/activity'
import { prisma } from '@shumai/db'
import { getProxyType } from '@shumai/core/src/utils/mime'
import { ulid } from 'ulid'
const execFileAsync = promisify(execFile)
import sharp from 'sharp'
export interface Usage {
inputTokens: number
@@ -25,15 +19,6 @@ export interface GeneratedEmbedding {
endTime?: number
}
export interface GenerateEmbeddingParams {
teamId: string
assetId: string
context: {
agent: unknown
asset: unknown
}
}
async function generateMultimodalEmbedding(
ai: GoogleGenAI,
buffer: Buffer,
@@ -62,130 +47,66 @@ async function generateMultimodalEmbedding(
return values
}
export async function generateEmbeddingActivity(params: GenerateEmbeddingParams): Promise<{
embeddings: GeneratedEmbedding[]
interface OpenAiEmbeddingItem {
type?: string
// eslint-disable-next-line @typescript-eslint/naming-convention
image_url?: {
url: string
}
}
async function callGemmaEmbeddingServer(input: string | OpenAiEmbeddingItem[]): Promise<{
embedding: number[]
usage: Usage
}> {
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- context properties are fetched as JSON from the DB activity
const { asset } = params.context as any
const rawBaseUrl = process.env.GEMMA_EMBEDDING_OPENAI_BASE_URL || 'http://localhost:9379/v1'
if (!asset.mediaType) {
throw ApplicationFailure.create({ message: 'asset has no media type', nonRetryable: true })
}
const normalized = rawBaseUrl.replace(/\/+$/, '')
const endpoint = normalized.endsWith('/embeddings')
? normalized
: normalized.endsWith('/v1')
? `${normalized}/embeddings`
: `${normalized}/v1/embeddings`
const apiKey = process.env.GEMINI_API_KEY
if (!apiKey) {
const response = await fetch(endpoint, {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
model: 'embeddinggemma-2',
input,
}),
})
if (!response.ok) {
const errorText = await response.text()
throw ApplicationFailure.create({
message: 'GEMINI_API_KEY environment variable is not configured',
nonRetryable: true,
message: `Embedding server returned status ${response.status}: ${errorText}`,
nonRetryable: false,
})
}
const resolvedApiKey = process.env[apiKey] || apiKey
const ai = new GoogleGenAI({ apiKey: resolvedApiKey })
const usage: Usage = {
model: 'gemini-embedding-2',
inputTokens: 0,
outputTokens: 0,
}
const proxyType =
(asset.media as PrismaJson.MediaInfo | null)?.proxyType ||
getProxyType(asset.mediaType, asset.name)
const isImage = proxyType === 'image'
const isVideo = proxyType === 'video'
if (!isImage && !isVideo) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- response json shape is dynamic from external endpoint
const json: any = await response.json()
const embedding = json.data?.[0]?.embedding
if (!embedding || !Array.isArray(embedding)) {
throw ApplicationFailure.create({
message: `unsupported media type for embeddings: ${asset.mediaType}`,
nonRetryable: true,
message: 'Invalid embedding returned from embedding server',
nonRetryable: false,
})
}
const key = asset.storageKey?.key
if (!key) {
throw ApplicationFailure.create({ message: 'asset has no key', nonRetryable: true })
}
const bucket = process.env.S3_BUCKET || 'shumai'
const results: GeneratedEmbedding[] = []
if (isImage) {
const { buffer: data } = await s3Service.getObject(bucket, key)
const embVec = await generateMultimodalEmbedding(ai, data, asset.mediaType)
results.push({ embedding: embVec })
} else if (isVideo) {
const inputSource = await s3Service.resolveInput(bucket, key)
const isRemote = inputSource.startsWith('http://') || inputSource.startsWith('https://')
const { stdout } = await execFileAsync('ffprobe', [
'-v',
'error',
'-show_entries',
'format=duration',
'-of',
'default=noprint_wrappers=1:nokey=1',
inputSource,
])
const duration = parseFloat(stdout.trim())
if (isNaN(duration)) {
throw ApplicationFailure.create({
message: 'failed to parse video duration',
nonRetryable: true,
})
}
const chunkSize = 60.0
for (let start = 0.0; start < duration; start += chunkSize) {
let end = start + chunkSize
if (end > duration) end = duration
const chunkTmp = path.join(os.tmpdir(), `video-chunk-${Date.now()}-${ulid()}.mp4`)
try {
await execFileAsync('ffmpeg', [
'-y',
'-loglevel',
'warning',
...(isRemote
? ['-reconnect', '1', '-reconnect_streamed', '1', '-reconnect_delay_max', '5']
: []),
'-ss',
start.toString(),
'-i',
inputSource,
'-t',
(end - start).toString(),
'-c',
'copy',
chunkTmp,
])
const chunkData = fs.readFileSync(chunkTmp)
try {
const embVec = await generateMultimodalEmbedding(ai, chunkData, 'video/mp4')
results.push({ embedding: embVec, startTime: start, endTime: end })
// eslint-disable-next-line @typescript-eslint/no-explicit-any
} catch (e: any) {
console.error(
`failed to generate video embedding for chunk ${start}-${end}: ${e.message}`,
)
}
} finally {
if (fs.existsSync(chunkTmp)) {
fs.unlinkSync(chunkTmp)
}
}
}
}
return {
embeddings: results,
usage,
embedding,
usage: {
model: json.model || 'gemma-embedding-2-openai',
inputTokens: json.usage?.prompt_tokens ?? 0,
outputTokens: 0,
},
}
}
import { textEmbeddingService } from '@shumai/core/src/search/text-embedding'
export interface GenerateTextEmbeddingParams {
text: string
teamId: string
@@ -195,44 +116,36 @@ export async function generateTextEmbeddingActivity(params: GenerateTextEmbeddin
embedding: number[]
usage: Usage
}> {
const apiKey = process.env.GEMINI_API_KEY
if (!apiKey) {
const provider = process.env.EMBEDDING_PROVIDER || 'gemini-embedding-2'
if (provider !== 'gemma-embedding-2-openai' && !process.env.GEMINI_API_KEY) {
throw ApplicationFailure.create({
message: 'GEMINI_API_KEY environment variable is not configured',
nonRetryable: true,
})
}
const resolvedApiKey = process.env[apiKey] || apiKey
const ai = new GoogleGenAI({ apiKey: resolvedApiKey })
try {
const result = await textEmbeddingService.generateEmbedding(params.text, {
teamId: params.teamId,
skipCache: true,
})
const usage: Usage = {
model: 'gemini-embedding-2',
inputTokens: 0,
outputTokens: 0,
}
const response = await ai.models.embedContent({
model: 'gemini-embedding-2',
contents: [
{
parts: [{ text: params.text }],
role: 'user',
return {
embedding: result.embedding,
usage: {
model: result.usage?.model || provider,
inputTokens: result.usage?.inputTokens ?? 0,
outputTokens: result.usage?.outputTokens ?? 0,
},
],
config: {
outputDimensionality: 1536,
},
})
const values = response.embeddings?.[0]?.values
if (!values) {
throw new Error('No embedding values returned from gemini-embedding-2')
}
return {
embedding: values,
usage,
}
} catch (err) {
if (err instanceof ApplicationFailure) {
throw err
}
throw ApplicationFailure.create({
message: err instanceof Error ? err.message : String(err),
nonRetryable: false,
})
}
}
@@ -311,7 +224,7 @@ export async function getEmbeddingContextActivity(params: { teamId: string; asse
user: { teamMembers: { some: { teamId: params.teamId } } },
},
})
if (!agent) {
if (!agent || process.env.ENABLE_EMBEDDING_AGENT === 'false') {
throw ApplicationFailure.create({
message: 'embedding feature is disabled or agent not found',
nonRetryable: true,
@@ -335,12 +248,51 @@ export async function getEmbeddingContextActivity(params: { teamId: string; asse
const parsedOverlap = parseFloat(process.env.EMBEDDING_CHUNK_OVERLAP || '')
const chunkOverlap = Number.isFinite(parsedOverlap) && parsedOverlap >= 0 ? parsedOverlap : 5.0
const embeddingProvider = process.env.EMBEDDING_PROVIDER || 'gemini-embedding-2'
const existingEmbeddings = await prisma.$queryRaw<ExistingAssetEmbeddingInfo[]>`
SELECT
id,
start_time as "startTime",
end_time as "endTime",
(embedding IS NOT NULL) as "has1536",
(embedding_768 IS NOT NULL) as "has768"
FROM asset_embeddings
WHERE asset_id = ${params.assetId}
`
return {
agent,
asset,
chunkDuration,
chunkOverlap,
embeddingProvider,
existingEmbeddings,
}
}
export interface ExistingAssetEmbeddingInfo {
id: string
startTime: number | null
endTime: number | null
has1536: boolean
has768: boolean
}
export async function deleteAssetEmbeddingsActivity(params: { assetId: string; ids?: string[] }) {
if (params.ids && params.ids.length > 0) {
await prisma.assetEmbedding.deleteMany({
where: {
assetId: params.assetId,
id: { in: params.ids },
},
})
} else {
await prisma.assetEmbedding.deleteMany({
where: {
assetId: params.assetId,
},
})
}
}
@@ -354,16 +306,32 @@ export async function saveAssetEmbeddingsActivity(params: {
}) {
for (const item of params.embeddings) {
const embVec = JSON.stringify(item.embedding)
const is768 = item.embedding.length === 768
if (item.startTime !== undefined && item.endTime !== undefined) {
await prisma.$executeRaw`
INSERT INTO asset_embeddings (id, asset_id, embedding, start_time, end_time, updated_at)
VALUES (gen_random_uuid()::text, ${params.assetId}, ${embVec}::vector, ${item.startTime}, ${item.endTime}, NOW())
`
if (is768) {
await prisma.$executeRaw`
INSERT INTO asset_embeddings (id, asset_id, embedding_768, start_time, end_time, updated_at)
VALUES (gen_random_uuid()::text, ${params.assetId}, ${embVec}::vector, ${item.startTime}, ${item.endTime}, NOW())
`
} else {
await prisma.$executeRaw`
INSERT INTO asset_embeddings (id, asset_id, embedding, start_time, end_time, updated_at)
VALUES (gen_random_uuid()::text, ${params.assetId}, ${embVec}::vector, ${item.startTime}, ${item.endTime}, NOW())
`
}
} else {
await prisma.$executeRaw`
INSERT INTO asset_embeddings (id, asset_id, embedding, updated_at)
VALUES (gen_random_uuid()::text, ${params.assetId}, ${embVec}::vector, NOW())
`
if (is768) {
await prisma.$executeRaw`
INSERT INTO asset_embeddings (id, asset_id, embedding_768, updated_at)
VALUES (gen_random_uuid()::text, ${params.assetId}, ${embVec}::vector, NOW())
`
} else {
await prisma.$executeRaw`
INSERT INTO asset_embeddings (id, asset_id, embedding, updated_at)
VALUES (gen_random_uuid()::text, ${params.assetId}, ${embVec}::vector, NOW())
`
}
}
}
}
@@ -380,6 +348,30 @@ export async function generateImageEmbeddingActivity(
embedding: number[]
usage: Usage
}> {
const provider = process.env.EMBEDDING_PROVIDER || 'gemini-embedding-2'
const { buffer: data } = await s3Service.getObject(
process.env.S3_BUCKET || 'shumai',
params.assetKey,
)
if (provider === 'gemma-embedding-2-openai') {
let imgData = data
let mimeType = params.mediaType || 'image/jpeg'
if (mimeType !== 'image/jpeg' && mimeType !== 'image/png') {
imgData = await sharp(data).jpeg().toBuffer()
mimeType = 'image/jpeg'
}
return callGemmaEmbeddingServer([
{
type: 'image_url',
// eslint-disable-next-line @typescript-eslint/naming-convention
image_url: {
url: `data:${mimeType};base64,${imgData.toString('base64')}`,
},
},
])
}
const apiKey = process.env.GEMINI_API_KEY
if (!apiKey) {
throw ApplicationFailure.create({
@@ -397,10 +389,6 @@ export async function generateImageEmbeddingActivity(
outputTokens: 0,
}
const { buffer: data } = await s3Service.getObject(
process.env.S3_BUCKET || 'shumai',
params.assetKey,
)
const embVec = await generateMultimodalEmbedding(ai, data, params.mediaType)
return {
@@ -409,6 +397,38 @@ export async function generateImageEmbeddingActivity(
}
}
/**
* Standard aspect ratios for video embedding:
* 1:1, 4:3, 3:4, 16:9, 9:16, 21:9 (including ultrawide 64:27), and 9:21.
*
* For these standard ratios, 15 frames guarantees that total input tokens stay strictly
* within the 2048 token signature bucket (<= 1982 tokens). For non-standard or custom
* aspect ratios (e.g. 5:3, 1.4:1), 14 frames is used to mathematically guarantee
* that sequence length never exceeds 2048 tokens (<= 1990 tokens), preventing OOM errors.
*/
export function getGemmaEmbeddingTargetFrameCount(width?: number, height?: number): number {
if (!width || !height || width <= 0 || height <= 0) {
return 14
}
const ratio = width / height
const standardRatios = [
1 / 1, // 1:1
4 / 3, // 4:3
3 / 4, // 3:4
16 / 9, // 16:9
9 / 16, // 9:16
21 / 9, // 21:9
64 / 27, // 21:9 ultrawide (2560x1080, 3440x1440)
9 / 21, // 9:21
27 / 64, // 9:21 ultrawide portrait
]
// Allow up to 3% relative tolerance for minor encoder rounding (e.g. 854x480 vs 16:9)
const isStandard = standardRatios.some((std) => Math.abs(ratio - std) / std <= 0.03)
return isStandard ? 15 : 14
}
export interface GenerateVideoChunkEmbeddingParams {
teamId: string
chunkKey: string
@@ -420,6 +440,83 @@ export async function generateVideoChunkEmbeddingActivity(
embedding: number[]
usage: Usage
}> {
const provider = process.env.EMBEDDING_PROVIDER || 'gemini-embedding-2'
const isGemma = provider === 'gemma-embedding-2-openai' || params.chunkKey.endsWith('.tar')
if (isGemma) {
const { buffer: data } = await s3Service.getObject(
process.env.S3_BUCKET || 'shumai',
params.chunkKey,
)
let filesMap: Map<string, File>
try {
const archive = new Bun.Archive(data)
filesMap = await archive.files()
} catch (err) {
throw ApplicationFailure.create({
message: `Failed to read chunk archive ${params.chunkKey}: ${err instanceof Error ? err.message : String(err)}`,
nonRetryable: true,
})
}
const frameFiles = Array.from(filesMap.keys())
.filter((f) => /\.(jpe?g|png|webp)$/i.test(f))
.sort()
if (frameFiles.length === 0) {
throw ApplicationFailure.create({
message: `No extracted frames found in chunk archive ${params.chunkKey}`,
nonRetryable: true,
})
}
let targetFrameCount = 14
try {
const firstFile = filesMap.get(frameFiles[0])
if (firstFile) {
const firstFileBytes = await firstFile.bytes()
if (firstFileBytes.byteLength > 0) {
const meta = await sharp(Buffer.from(firstFileBytes)).metadata()
targetFrameCount = getGemmaEmbeddingTargetFrameCount(meta.width, meta.height)
}
}
} catch {
targetFrameCount = 14
}
let selectedFrames = frameFiles
if (selectedFrames.length > targetFrameCount) {
const step = (selectedFrames.length - 1) / (targetFrameCount - 1)
selectedFrames = Array.from(
{ length: targetFrameCount },
(_, i) => selectedFrames[Math.round(i * step)],
)
}
const input: OpenAiEmbeddingItem[] = await Promise.all(
selectedFrames.map(async (f) => {
const file = filesMap.get(f)!
const fileBytes = await file.bytes()
if (fileBytes.byteLength === 0) {
throw ApplicationFailure.create({
message: `Extracted frame ${f} in chunk archive ${params.chunkKey} is empty (0 bytes)`,
nonRetryable: true,
})
}
return {
type: 'image_url',
// eslint-disable-next-line @typescript-eslint/naming-convention
image_url: {
url: `data:image/jpeg;base64,${Buffer.from(fileBytes).toString('base64')}`,
},
}
}),
)
return await callGemmaEmbeddingServer(input)
}
const apiKey = process.env.GEMINI_API_KEY
if (!apiKey) {
throw ApplicationFailure.create({
@@ -350,5 +350,24 @@ describe('serializeContextToXml', () => {
'<context>\n <annotation id="test&lt;id&gt;&amp;&quot;123&quot;" />\n</context>',
)
})
it('serializes position type="degree" with integer degree attribute without seconds', () => {
const context: ShumaiMessageContext = {
currentAsset: {
id: 'ast_3d',
name: 'robot.glb',
type: 'file',
mediaType: '3d',
path: 'robot.glb',
},
position: {
type: 'degree',
degree: 45,
},
}
expect(serializeContextToXml(context)).toBe(
'<context>\n <current_asset id="ast_3d" name="robot.glb" type="file" media_type="3d" path="robot.glb" />\n <position type="degree" degree="45" />\n</context>',
)
})
})
})
+3 -5
View File
@@ -610,11 +610,9 @@ import * as aiActivities from './activities/ai'
import { agentAutofillMedia } from './workflows/agent-autofill'
import { agentChat } from './workflows/agent-chat'
import { agentEmbeddingMedia } from './workflows/agent-embedding'
import { queryEmbeddingForSearch } from './workflows/query-embedding-for-search'
export function initAgentWorkflows() {
registerWorkflow(WorkflowTaskType.ai_embedding, agentEmbeddingMedia)
registerWorkflow(WorkflowTaskType.query_embedding_for_search, queryEmbeddingForSearch)
registerWorkflow(WorkflowTaskType.ai_metadata_autofill, agentAutofillMedia)
registerWorkflow(WorkflowTaskType.chat, agentChat)
@@ -625,21 +623,21 @@ export function initAgentWorkflows() {
export * from './activities/agent'
export {
extractAiMetadataActivity,
generateEmbeddingActivity,
generateImageEmbeddingActivity,
generateTextEmbeddingActivity,
generateVideoChunkEmbeddingActivity,
saveAssetEmbeddingsActivity,
deleteAssetEmbeddingsActivity,
type ExtractAiMetadataParams,
type GeneratedEmbedding,
type GenerateEmbeddingParams,
type GenerateTextEmbeddingParams,
type ExistingAssetEmbeddingInfo,
} from './activities/ai'
export * from './database-session-storage'
export * from './context/serialize-context'
export * from './workflows/agent-autofill'
export * from './workflows/agent-chat'
export * from './workflows/agent-embedding'
export * from './workflows/query-embedding-for-search'
export * from './tools/generate-image'
export * from './tools/generate-video'
export * from './tools/rename-asset'
@@ -4,6 +4,7 @@ import { setupTestDbHooks } from '@shumai/db/test'
import { resolveAnnotationsById } from './annotation-resolver'
import { DatabaseSessionStorage } from '../database-session-storage'
import type { SessionTreeEntry } from '@earendil-works/pi-agent-core'
import { degreeToSecond } from '@shumai/core/src/utils/model3d'
describe('resolveAnnotationsById', () => {
setupTestDbHooks()
@@ -210,6 +211,89 @@ describe('resolveAnnotationsById', () => {
])
})
it('resolves annotations and degree position timestamp from 1-on-1 chat AgentSessionEntry for 3D assets', async () => {
const user = await prisma.user.create({
data: { name: 'User 3D', email: 'user3d@test.com', password: 'pw' },
})
const team = await prisma.team.create({
data: { name: 'Team 3D' },
})
const project = await prisma.project.create({
data: { name: 'Project 3D', teamId: team.id },
})
const asset = await prisma.asset.create({
data: {
name: 'robot.glb',
type: 'file',
projectId: project.id,
status: 'uploaded',
},
})
const agentUser = await prisma.user.create({
data: { name: 'Agent 3D', email: 'agent3d@example.com', type: 'agent' },
})
const agent = await prisma.agent.create({
data: {
id: agentUser.id,
teamId: team.id,
type: 'chat',
config: { provider: 'test-provider', model: 'test-model' },
},
})
const storage = await DatabaseSessionStorage.create({
agentId: agent.id,
userId: user.id,
assetId: asset.id,
})
storage.currentMessageContext = {
currentAsset: {
id: asset.id,
name: asset.name,
type: 'file',
path: asset.name,
},
position: { type: 'degree', degree: 45 },
annotation: true,
annotations: [
{
type: 'rectangle',
x: 10,
y: 20,
width: 30,
height: 40,
},
],
}
const userEntry: SessionTreeEntry = {
type: 'message',
id: '01JTEST3DENTRY123456789',
parentId: null,
timestamp: new Date().toISOString(),
message: {
role: 'user',
content: 'Check markup at 45 degrees',
timestamp: Date.now(),
},
}
await storage.appendEntry(userEntry)
const result = await resolveAnnotationsById(asset.id, '01JTEST3DENTRY123456789')
expect(result.timestamp).toBeCloseTo(degreeToSecond(45))
expect(result.annotations).toEqual([
{
type: 'rectangle',
x: 10,
y: 20,
width: 30,
height: 40,
},
])
})
it('throws an error when AgentSessionEntry belongs to a different asset', async () => {
const user = await prisma.user.create({
data: { name: 'User', email: 'user3@test.com', password: 'pw' },
@@ -1,5 +1,6 @@
import { prisma } from '@shumai/db'
import type { ShumaiMessageContext } from '@shumai/dtos'
import { degreeToSecond } from '@shumai/core/src/utils/model3d'
export async function resolveAnnotationsById(
assetId: string,
@@ -44,7 +45,14 @@ export async function resolveAnnotationsById(
const list = details?.annotations
const annotations =
Array.isArray(list) && list.length > 0 ? (list as unknown as PrismaJson.AnnotationList) : null
const timestamp = details?.position?.type === 'time' ? details.position.seconds : null
let timestamp: number | null = null
if (details?.position?.type === 'time') {
timestamp = details.position.seconds
} else if (details?.position?.type === 'degree') {
timestamp = degreeToSecond(details.position.degree)
} else if (details?.position?.type === 'page') {
timestamp = details.position.page
}
return { annotations, timestamp }
}
+471
View File
@@ -11,6 +11,7 @@ import { s3Service } from '@shumai/core/src/s3/s3'
import { workflowService } from '@shumai/workflow-core'
import { authzService } from '@shumai/core/src/authz/authz'
import * as annotationResolver from './annotation-resolver'
import { degreeToSecond } from '@shumai/core/src/utils/model3d'
vi.mock('@shumai/db', async (importOriginal) => {
const actual = await importOriginal<typeof import('@shumai/db')>()
@@ -211,6 +212,476 @@ describe('readAssetTool', () => {
})
})
describe('3D Assets', () => {
it('should read 3D asset 4-views image and return image with quadrant layout description', async () => {
vi.mocked(authzService.hasPermission).mockResolvedValue()
vi.mocked(prisma.asset.findUnique).mockResolvedValue({
id: 'model-1',
name: 'robot.glb',
mediaType: 'model/gltf-binary',
storageKey: { key: 'raw/robot.glb' },
media: {
proxyType: '3d',
imageTranscodes: [{ key: 'proxy/views.webp' }],
},
} as unknown as Asset)
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: Buffer.from('fake-views-bytes'),
contentType: 'image/webp',
} as unknown as { buffer: Buffer; contentType: string })
const tool = createReadAssetTool('user-1')
const result = await tool.execute('call-3d', {
assetId: 'model-1',
annotationId: null,
imageConfig: null,
videoConfig: null,
docConfig: null,
})
expect(authzService.hasPermission).toHaveBeenCalledWith({
user: { id: 'user-1' },
permission: 'Read',
type: 'asset',
id: 'model-1',
})
expect(s3Service.getObject).toHaveBeenCalledWith('shumai', 'proxy/views.webp')
const textItem = result.content[0] as { type: 'text'; text: string }
const imageItem = result.content[1] as { type: 'image'; data: string; mimeType: string }
expect(textItem.text).toContain('Orthogonal 4-view composite image')
expect(textItem.text).toContain('Top-Left quadrant: Front view (0°)')
expect(textItem.text).toContain('Top-Right quadrant: Right side view (90°)')
expect(textItem.text).toContain('Bottom-Left quadrant: Back view (180°)')
expect(textItem.text).toContain('Bottom-Right quadrant: Left side view (270°)')
expect(imageItem.type).toBe('image')
expect(imageItem.data).toBe(Buffer.from('fake-views-bytes').toString('base64'))
expect(result.details.key).toBe('proxy/views.webp')
})
it('should support s3KeyOnly mode for 3D assets', async () => {
vi.mocked(authzService.hasPermission).mockResolvedValue()
vi.mocked(prisma.asset.findUnique).mockResolvedValue({
id: 'model-1',
name: 'robot.glb',
mediaType: 'model/gltf-binary',
media: {
proxyType: '3d',
imageTranscodes: [{ key: 'proxy/views.webp' }],
},
} as unknown as Asset)
const tool = createReadAssetTool('user-1')
const result = await tool.execute('call-3d-s3', {
assetId: 'model-1',
annotationId: null,
s3KeyOnly: true,
imageConfig: null,
videoConfig: null,
docConfig: null,
})
expect(result.content).toHaveLength(1)
const textItem = result.content[0] as { type: 'text'; text: string }
expect(textItem.text).toContain('proxy/views.webp')
expect(textItem.text).toContain('Front view (0°)')
expect(result.details.key).toBe('proxy/views.webp')
})
it('should throw error if videoConfig is passed for 3D assets', async () => {
vi.mocked(authzService.hasPermission).mockResolvedValue()
vi.mocked(prisma.asset.findUnique).mockResolvedValue({
id: 'model-1',
name: 'robot.glb',
mediaType: 'model/gltf-binary',
media: {
proxyType: '3d',
imageTranscodes: [{ key: 'proxy/views.webp' }],
},
} as unknown as Asset)
const tool = createReadAssetTool('user-1')
await expect(
tool.execute('call-3d-err', {
assetId: 'model-1',
annotationId: null,
imageConfig: null,
videoConfig: { start: 0, end: 4, count: 4 },
docConfig: null,
}),
).rejects.toThrow(
"To inspect specific angles pass 'config3d: { degree: ... }'; do not pass videoConfig",
)
})
it('should throw error if 3D asset imageTranscodes is missing when overview is requested', async () => {
vi.mocked(authzService.hasPermission).mockResolvedValue()
vi.mocked(prisma.asset.findUnique).mockResolvedValue({
id: 'model-1',
name: 'robot.glb',
mediaType: 'model/gltf-binary',
media: {
proxyType: '3d',
imageTranscodes: [],
},
} as unknown as Asset)
const tool = createReadAssetTool('user-1')
await expect(
tool.execute('call-3d-err2', {
assetId: 'model-1',
annotationId: null,
imageConfig: null,
videoConfig: null,
docConfig: null,
}),
).rejects.toThrow('4-views image (views.webp) is missing')
})
it('should extract specific turntable frame when config3d.degree is provided', async () => {
vi.mocked(authzService.hasPermission).mockResolvedValue()
vi.mocked(prisma.asset.findUnique).mockResolvedValue({
id: 'model-1',
name: 'robot.glb',
mediaType: 'model/gltf-binary',
projectId: 'proj-1',
media: {
proxyType: '3d',
videoTranscodes: [{ key: 'proxy/turntable.mp4' }],
imageTranscodes: [{ key: 'proxy/views.webp' }],
},
} as unknown as Asset)
vi.mocked(prisma.workflowTask.create).mockResolvedValue({
id: 'task-shot-3d',
} as unknown as WorkflowTask)
vi.mocked(workflowService.executeWait).mockResolvedValue({
id: 'task-shot-3d',
status: WorkflowTaskStatus.completed,
output: {
screenshots: [{ key: 'proxy/shot-45.webp', timestamp: 0.5833 }],
},
} as unknown as WorkflowTask)
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: Buffer.from('fake-shot-45-bytes'),
contentType: 'image/webp',
} as unknown as { buffer: Buffer; contentType: string })
const tool = createReadAssetTool('user-1')
const result = await tool.execute('call-3d-degree', {
assetId: 'model-1',
annotationId: null,
config3d: { degree: 45 },
imageConfig: null,
videoConfig: null,
docConfig: null,
})
expect(prisma.workflowTask.create).toHaveBeenCalledWith({
data: expect.objectContaining({
type: WorkflowTaskType.transcode_screenshot,
assetId: 'model-1',
payload: {
projectId: 'proj-1',
screenshot: {
start: expect.any(Number),
end: expect.any(Number),
count: 1,
commentTimestamp: expect.any(Number),
annotations: undefined,
},
},
}),
})
const textItem = result.content[0] as { type: 'text'; text: string }
const imageItem = result.content[1] as { type: 'image'; data: string; mimeType: string }
expect(textItem.text).toContain('view at 45°:')
expect(imageItem.data).toBe(Buffer.from('fake-shot-45-bytes').toString('base64'))
expect(result.details.key).toBe('proxy/shot-45.webp')
})
it('should throw error when config3d.degree is not a multiple of 15 or out of range', async () => {
vi.mocked(authzService.hasPermission).mockResolvedValue()
vi.mocked(prisma.asset.findUnique).mockResolvedValue({
id: 'model-1',
name: 'robot.glb',
mediaType: 'model/gltf-binary',
media: {
proxyType: '3d',
videoTranscodes: [{ key: 'proxy/turntable.mp4' }],
},
} as unknown as Asset)
const tool = createReadAssetTool('user-1')
await expect(
tool.execute('call-3d-invalid-deg', {
assetId: 'model-1',
annotationId: null,
config3d: { degree: 40 },
imageConfig: null,
videoConfig: null,
docConfig: null,
}),
).rejects.toThrow('Invalid turntable degree: 40')
await expect(
tool.execute('call-3d-out-of-range', {
assetId: 'model-1',
annotationId: null,
config3d: { degree: 360 },
imageConfig: null,
videoConfig: null,
docConfig: null,
}),
).rejects.toThrow('Invalid turntable degree: 360')
})
it('should extract frame with overlaid markups when annotationId is provided on 3D asset', async () => {
vi.mocked(authzService.hasPermission).mockResolvedValue()
vi.mocked(prisma.asset.findUnique).mockResolvedValue({
id: 'model-1',
name: 'robot.glb',
mediaType: 'model/gltf-binary',
projectId: 'proj-1',
media: {
proxyType: '3d',
videoTranscodes: [{ key: 'proxy/turntable.mp4' }],
},
} as unknown as Asset)
vi.spyOn(annotationResolver, 'resolveAnnotationsById').mockResolvedValue({
annotations: [
{
type: 'freehand',
color: '#ff0000',
points: [
[0, 0],
[10, 10],
],
},
],
timestamp: 0.5, // 0.5s = 45°
})
vi.mocked(prisma.workflowTask.create).mockResolvedValue({
id: 'task-shot-ann',
} as unknown as WorkflowTask)
vi.mocked(workflowService.executeWait).mockResolvedValue({
id: 'task-shot-ann',
status: WorkflowTaskStatus.completed,
output: {
screenshots: [{ key: 'proxy/shot-annotated.webp', timestamp: 0.5 }],
},
} as unknown as WorkflowTask)
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: Buffer.from('fake-shot-ann-bytes'),
contentType: 'image/webp',
} as unknown as { buffer: Buffer; contentType: string })
const tool = createReadAssetTool('user-1')
const result = await tool.execute('call-3d-ann', {
assetId: 'model-1',
annotationId: 'comm-1',
config3d: null,
imageConfig: null,
videoConfig: null,
docConfig: null,
})
expect(prisma.workflowTask.create).toHaveBeenCalledWith({
data: expect.objectContaining({
type: WorkflowTaskType.transcode_screenshot,
assetId: 'model-1',
payload: {
projectId: 'proj-1',
screenshot: {
start: 0.5,
end: 0.5,
count: 1,
commentTimestamp: 0.5,
annotations: [
{
type: 'freehand',
color: '#ff0000',
points: [
[0, 0],
[10, 10],
],
},
],
},
},
}),
})
const textItem = result.content[0] as { type: 'text'; text: string }
expect(textItem.text).toContain('view at 45° with visual markups overlaid:')
expect(result.details.key).toBe('proxy/shot-annotated.webp')
})
it('should extract turntable frame at resolved degree angle for 3D session annotations', async () => {
vi.mocked(authzService.hasPermission).mockResolvedValue()
vi.mocked(prisma.asset.findUnique).mockResolvedValue({
id: 'model-1',
name: 'robot.glb',
mediaType: 'model/gltf-binary',
projectId: 'proj-1',
media: {
proxyType: '3d',
videoTranscodes: [{ key: 'proxy/turntable.mp4' }],
},
} as unknown as Asset)
const expectedTime = degreeToSecond(90)
vi.spyOn(annotationResolver, 'resolveAnnotationsById').mockResolvedValue({
annotations: [
{
type: 'box',
color: '#00ff00',
points: [
[10, 20],
[40, 60],
],
},
],
timestamp: expectedTime,
})
vi.mocked(prisma.workflowTask.create).mockResolvedValue({
id: 'task-shot-3d-session',
} as unknown as WorkflowTask)
vi.mocked(workflowService.executeWait).mockResolvedValue({
id: 'task-shot-3d-session',
status: WorkflowTaskStatus.completed,
output: {
screenshots: [{ key: 'proxy/shot-annotated-90.webp', timestamp: expectedTime }],
},
} as unknown as WorkflowTask)
vi.mocked(s3Service.getObject).mockResolvedValue({
buffer: Buffer.from('fake-shot-90-bytes'),
contentType: 'image/webp',
} as unknown as { buffer: Buffer; contentType: string })
const tool = createReadAssetTool('user-1')
const result = await tool.execute('call-3d-session-ann', {
assetId: 'model-1',
annotationId: 'session-entry-1',
config3d: null,
imageConfig: null,
videoConfig: null,
docConfig: null,
})
expect(prisma.workflowTask.create).toHaveBeenCalledWith({
data: expect.objectContaining({
type: WorkflowTaskType.transcode_screenshot,
assetId: 'model-1',
payload: {
projectId: 'proj-1',
screenshot: {
start: expectedTime,
end: expectedTime,
count: 1,
commentTimestamp: expectedTime,
annotations: [
{
type: 'box',
color: '#00ff00',
points: [
[10, 20],
[40, 60],
],
},
],
},
},
}),
})
const textItem = result.content[0] as { type: 'text'; text: string }
expect(textItem.text).toContain('view at 90° with visual markups overlaid:')
expect(result.details.key).toBe('proxy/shot-annotated-90.webp')
})
it('should throw error if config3d.degree conflicts with annotation angle', async () => {
vi.mocked(authzService.hasPermission).mockResolvedValue()
vi.mocked(prisma.asset.findUnique).mockResolvedValue({
id: 'model-1',
name: 'robot.glb',
mediaType: 'model/gltf-binary',
projectId: 'proj-1',
media: {
proxyType: '3d',
videoTranscodes: [{ key: 'proxy/turntable.mp4' }],
},
} as unknown as Asset)
vi.spyOn(annotationResolver, 'resolveAnnotationsById').mockResolvedValue({
annotations: [
{
type: 'box',
color: '#ff0000',
points: [
[0, 0],
[10, 10],
],
},
],
timestamp: 0.5, // 45°
})
const tool = createReadAssetTool('user-1')
await expect(
tool.execute('call-3d-conflict', {
assetId: 'model-1',
annotationId: 'comm-1',
config3d: { degree: 90 }, // Conflicts with 45°
imageConfig: null,
videoConfig: null,
docConfig: null,
}),
).rejects.toThrow(
'Annotation "comm-1" was drawn at 45°, but degree 90° was specified. Do not pass a conflicting degree with annotationId.',
)
})
it('should throw error if 3D asset turntable video is missing when angle inspection is requested', async () => {
vi.mocked(authzService.hasPermission).mockResolvedValue()
vi.mocked(prisma.asset.findUnique).mockResolvedValue({
id: 'model-1',
name: 'robot.glb',
mediaType: 'model/gltf-binary',
media: {
proxyType: '3d',
videoTranscodes: [],
},
} as unknown as Asset)
const tool = createReadAssetTool('user-1')
await expect(
tool.execute('call-3d-missing-turntable', {
assetId: 'model-1',
annotationId: null,
config3d: { degree: 45 },
imageConfig: null,
videoConfig: null,
docConfig: null,
}),
).rejects.toThrow('turntable video (turntable.mp4) is missing')
})
})
describe('Video Assets', () => {
it('should throw error if videoConfig is missing for video assets', async () => {
vi.mocked(authzService.hasPermission).mockResolvedValue()
+237 -3
View File
@@ -7,6 +7,12 @@ import { workflowService } from '@shumai/workflow-core'
import { authzService, Permission, ResourceType } from '@shumai/core/src/authz/authz'
import { resolveAnnotationsById } from './annotation-resolver'
import { getFileMimeType } from '@shumai/core/src/utils/file-mime'
import { is3dModel } from '@shumai/core/src/utils/mime'
import {
isValidTurntableDegree,
secondToDegree,
degreeToSecond,
} from '@shumai/core/src/utils/model3d'
export const readAssetSchema = Type.Object(
{
@@ -37,6 +43,31 @@ export const readAssetSchema = Type.Object(
),
Type.Null(),
]),
config3d: Type.Optional(
Type.Union([
Type.Object(
{
degree: Type.Optional(
Type.Union([
Type.Integer({
minimum: 0,
maximum: 345,
description:
'Turntable rotation degree (integer: 0, 15, 30, ..., 345 in 15° increments) to inspect a specific angle of a 3D model. Leave null/omitted to view the 4-view composite overview.',
}),
Type.Null(),
]),
),
},
{
additionalProperties: false,
description:
'Configuration for 3D model assets. Pass null if inspecting another asset type. Optionally pass degree to inspect a specific turntable angle, or leave null/omitted to view the 4-view composite overview.',
},
),
Type.Null(),
]),
),
videoConfig: Type.Union([
Type.Object(
{
@@ -167,12 +198,12 @@ export function createReadAssetTool(
label: 'Read Asset',
description:
'Inspects and reads the content of an asset in the workspace. ' +
'Supports visual analysis for images, video frame extraction, PDF page rendering, and direct text reading for text/markdown files. ' +
'Pass assetId, and provide the specific config (imageConfig, videoConfig, or docConfig) while setting unused configs to null. ' +
'Supports visual analysis for images, 3D models (overview or specific degree angle), video frame extraction, PDF page rendering, and direct text reading for text/markdown files. ' +
'Pass assetId, and provide the specific config (imageConfig, config3d, videoConfig, or docConfig) while setting unused configs to null. ' +
'Optionally provide annotationId to view visual markups drawn on the asset.',
parameters: readAssetSchema,
execute: async (_toolCallId, params) => {
const { assetId, annotationId, s3KeyOnly, videoConfig, docConfig } = params
const { assetId, annotationId, s3KeyOnly, config3d, videoConfig, docConfig } = params
if (agentType === 'autofill' && !userId) {
if (!targetAssetId) {
@@ -243,6 +274,209 @@ export function createReadAssetTool(
const filename = asset.name?.toLowerCase() || ''
const bucket = process.env.S3_BUCKET || 'shumai'
// ----------------------------------------------------------------------
// Branch 0: 3D Assets (Orthogonal 4-View Composite Image or Degree Angle)
// ----------------------------------------------------------------------
const is3d = proxyType === '3d' || is3dModel(mediaType, filename)
if (is3d) {
if (videoConfig) {
throw new Error(
`Asset ${asset.id} is a 3D model. To inspect specific angles pass 'config3d: { degree: ... }'; do not pass videoConfig.`,
)
}
if (config3d?.degree !== undefined && config3d?.degree !== null) {
if (!isValidTurntableDegree(config3d.degree)) {
throw new Error(
`Invalid turntable degree: ${config3d.degree}. Degree must be an integer multiple of 15 between 0 and 345 (e.g. 0, 15, 30, 45, ..., 345).`,
)
}
}
let annotations: PrismaJson.AnnotationList | null = null
let commentTimestamp: number | null = null
if (annotationId) {
const resolved = await resolveAnnotationsById(asset.id, annotationId)
annotations = resolved.annotations
commentTimestamp = resolved.timestamp
if (commentTimestamp !== null && commentTimestamp !== undefined) {
const commentDegree = secondToDegree(commentTimestamp)
if (
config3d?.degree !== undefined &&
config3d?.degree !== null &&
config3d.degree !== commentDegree
) {
throw new Error(
`Annotation "${annotationId}" was drawn at ${commentDegree}°, but degree ${config3d.degree}° was specified. Do not pass a conflicting degree with annotationId.`,
)
}
}
}
if ((config3d?.degree !== undefined && config3d?.degree !== null) || annotationId) {
if (!mediaInfo?.videoTranscodes || mediaInfo.videoTranscodes.length === 0) {
throw new Error(
`Asset ${asset.id} is a 3D model, but its turntable video (turntable.mp4) is missing.`,
)
}
let targetDegree: number
let targetSecond: number
if (config3d?.degree !== undefined && config3d?.degree !== null) {
targetDegree = config3d.degree
targetSecond = degreeToSecond(targetDegree)
} else {
targetSecond =
commentTimestamp !== null && commentTimestamp !== undefined ? commentTimestamp : 0
targetDegree = secondToDegree(targetSecond)
}
const task = await prisma.workflowTask.create({
data: {
assetId: asset.id,
projectId: asset.projectId || 'none',
type: WorkflowTaskType.transcode_screenshot,
status: WorkflowTaskStatus.pending,
payload: {
projectId: asset.projectId || 'none',
screenshot: {
start: targetSecond,
end: targetSecond,
count: 1,
commentTimestamp: targetSecond,
annotations: annotations ?? undefined,
},
},
},
})
const completedTask = await workflowService.executeWait(task)
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- transcode screenshot output
const output = completedTask.output as any
const screenshots = output?.screenshots as
| Array<{ key: string; timestamp: number }>
| undefined
if (!screenshots || screenshots.length === 0) {
throw new Error(
`Failed to extract 3D turntable frame for asset ${asset.id} at ${targetDegree}°.`,
)
}
const shotKey = screenshots[0].key
const downloadUrl = await s3Service.presign(
bucket,
shotKey,
'GET',
true,
asset.name || undefined,
)
const content: Array<TextContent | ImageContent> = []
const hasMarkup = Boolean(annotations && annotations.length > 0)
const textDescription = hasMarkup
? `3D asset "${asset.name}" (ID: ${asset.id}, S3 Key: "${shotKey}") view at ${targetDegree}° with visual markups overlaid:`
: `3D asset "${asset.name}" (ID: ${asset.id}, S3 Key: "${shotKey}") view at ${targetDegree}°:`
if (s3KeyOnly) {
content.push({
type: 'text',
text: textDescription,
})
} else {
const { buffer, contentType } = await s3Service.getObject(bucket, shotKey)
const actualMimeType =
contentType && contentType !== 'application/octet-stream' ? contentType : 'image/webp'
content.push({
type: 'text',
text: textDescription,
})
content.push({
type: 'image',
data: buffer.toString('base64'),
mimeType: actualMimeType,
})
}
return {
content,
details: {
assetId: asset.id,
name: asset.name,
mediaType: asset.mediaType,
key: shotKey,
downloadUrl,
sourceKeys: [shotKey],
},
}
}
if (!mediaInfo?.imageTranscodes || mediaInfo.imageTranscodes.length === 0) {
throw new Error(
`Asset ${asset.id} is a 3D model, but its 4-views image (views.webp) is missing.`,
)
}
const mediaKey = mediaInfo.imageTranscodes[0].key
if (!mediaKey) {
throw new Error(`Asset ${asset.id} is a 3D model, but its 4-views image key is missing.`)
}
const keyToUse = mediaKey
const mimeType = 'image/webp'
const downloadUrl = await s3Service.presign(
bucket,
keyToUse,
'GET',
true,
asset.name || undefined,
)
const content: Array<TextContent | ImageContent> = []
const textDescription = [
`3D asset "${asset.name}" (ID: ${asset.id}, MIME: ${mimeType}, S3 Key: "${keyToUse}"):`,
'Orthogonal 4-view composite image (2x2 grid across 360° rotation):',
'- Top-Left quadrant: Front view (0°)',
'- Top-Right quadrant: Right side view (90°)',
'- Bottom-Left quadrant: Back view (180°)',
'- Bottom-Right quadrant: Left side view (270°)',
].join('\n')
if (s3KeyOnly) {
content.push({
type: 'text',
text: textDescription,
})
} else {
const { buffer, contentType } = await s3Service.getObject(bucket, keyToUse)
const actualMimeType =
contentType && contentType !== 'application/octet-stream' ? contentType : mimeType
content.push({
type: 'text',
text: textDescription,
})
content.push({
type: 'image',
data: buffer.toString('base64'),
mimeType: actualMimeType,
})
}
return {
content,
details: {
assetId: asset.id,
name: asset.name,
mediaType: asset.mediaType,
key: keyToUse,
downloadUrl,
sourceKeys: [keyToUse],
},
}
}
// ----------------------------------------------------------------------
// Branch 1: Image Assets
// ----------------------------------------------------------------------
@@ -129,6 +129,42 @@ describe('createReadThreadTool', () => {
expect(textContent).toContain('- [Ai Agent] (id: reply-1, has_markup: false): Agent answer')
})
it('should format 3d asset comment timestamps as degree: X°', async () => {
vi.mocked(prisma.assetComment.findUnique).mockResolvedValue({
id: 'root-3d',
message: 'Notice this angle',
creator: { name: 'Alice' },
second: 0.5,
annotation: [{ type: 'box' }],
asset: {
media: { proxyType: '3d' },
mediaType: 'model/gltf-binary',
name: 'robot.glb',
},
// eslint-disable-next-line @typescript-eslint/no-explicit-any
} as any)
vi.mocked(prisma.assetComment.findMany).mockResolvedValue([
{
id: 'reply-3d-1',
message: 'Looking at back side',
creator: { name: 'Bob' },
second: 2.0,
},
// eslint-disable-next-line @typescript-eslint/no-explicit-any
] as any)
const tool = createReadThreadTool()
const result = await tool.execute('call-3d', { threadId: 'root-3d' })
const textContent = (result.content[0] as { text: string }).text
expect(textContent).toContain(
'Thread Root [Alice] (id: root-3d, degree: 45°, has_markup: true): Notice this angle',
)
expect(textContent).toContain(
'- [Bob] (id: reply-3d-1, degree: 180°, has_markup: false): Looking at back side',
)
})
it('should describe threadId parameter as referencing <thread id="..." /> in <context>', () => {
const tool = createReadThreadTool()
// eslint-disable-next-line @typescript-eslint/no-explicit-any
+26 -9
View File
@@ -1,6 +1,8 @@
import { prisma } from '@shumai/db'
import { Type } from 'typebox'
import { type AgentTool } from '@earendil-works/pi-agent-core'
import { secondToDegree } from '@shumai/core/src/utils/model3d'
import { getProxyType } from '@shumai/core/src/utils/mime'
const readThreadSchema = Type.Object({
threadId: Type.String({
@@ -9,14 +11,21 @@ const readThreadSchema = Type.Object({
}),
})
function formatCommentMeta(c: {
id: string
second?: number | null
annotation?: unknown
}): string {
function formatCommentMeta(
c: {
id: string
second?: number | null
annotation?: unknown
},
proxyType?: string,
): string {
const parts = [`id: ${c.id}`]
if (c.second !== null && c.second !== undefined) {
parts.push(`time: ${c.second}s`)
if (proxyType === '3d') {
parts.push(`degree: ${secondToDegree(c.second)}°`)
} else {
parts.push(`time: ${c.second}s`)
}
}
const hasMarkup = Boolean(c.annotation && Array.isArray(c.annotation) && c.annotation.length > 0)
parts.push(`has_markup: ${hasMarkup}`)
@@ -34,7 +43,10 @@ export const createReadThreadTool = (): AgentTool<
execute: async (_toolCallId, params) => {
const rootComment = await prisma.assetComment.findUnique({
where: { id: params.threadId },
include: { creator: true },
include: {
creator: true,
asset: { select: { media: true, mediaType: true, name: true } },
},
})
if (!rootComment) {
@@ -61,9 +73,14 @@ export const createReadThreadTool = (): AgentTool<
include: { creator: true },
})
const proxyType =
(rootComment.asset?.media as PrismaJson.MediaInfo | null)?.proxyType ||
getProxyType(rootComment.asset?.mediaType, rootComment.asset?.name) ||
undefined
const isRootAgent = rootComment.creator?.type === 'agent' || rootComment.sessionId !== null
const rootAuthor = rootComment.creator?.name || (isRootAgent ? 'Ai Agent' : 'User')
const rootMeta = formatCommentMeta(rootComment)
const rootMeta = formatCommentMeta(rootComment, proxyType)
let output = `Thread Root [${rootAuthor}] ${rootMeta}: ${rootComment.message || ''}\n\nReplies (${replies.length}):`
if (replies.length === 0) {
@@ -72,7 +89,7 @@ export const createReadThreadTool = (): AgentTool<
for (const reply of replies) {
const isAgent = reply.creator?.type === 'agent' || reply.sessionId !== null
const author = reply.creator?.name || (isAgent ? 'Ai Agent' : 'User')
const replyMeta = formatCommentMeta(reply)
const replyMeta = formatCommentMeta(reply, proxyType)
output += `\n- [${author}] ${replyMeta}: ${reply.message || ''}`
}
}
@@ -806,4 +806,89 @@ describe('Agent Chat Workflow', () => {
}),
)
})
it('should format mediaPosition as degree for 3D assets in 1-on-1 chat', async () => {
mockActivities.getAssetActivity.mockResolvedValue({
id: 'a-3d',
project: { teamId: 't1' },
type: 'file',
name: 'model.glb',
mediaType: 'model/gltf-binary',
media: { proxyType: '3d' },
parentId: null,
})
const task = await prisma.workflowTask.create({
data: {
type: 'chat',
status: 'pending',
assetId: 'a-3d',
payload: {
projectId: 'p1',
agent: {
agentId: 'b1',
prompt: 'Look at this angle',
sessionId: 'session-3d-1',
second: 0.5, // 0.5s turntable time = 45°
},
},
},
})
await agentChat(task)
expect(mockActivities.agentChatActivity).toHaveBeenCalledWith(
expect.objectContaining({
messageContext: expect.objectContaining({
position: { type: 'degree', degree: 45 },
}),
}),
)
})
it('should format mediaPosition as degree for 3D assets in comment mode', async () => {
mockActivities.getAssetActivity.mockResolvedValue({
id: 'a-3d',
project: { teamId: 't1' },
type: 'file',
name: 'model.glb',
mediaType: 'model/gltf-binary',
media: { proxyType: '3d' },
parentId: null,
})
mockActivities.getCommentActivity.mockResolvedValue({
id: 'comment-3d-1',
creatorId: 'user-c1',
message: 'Look at the back',
replyToId: null,
second: 2.0, // 2.0s = 180°
attachments: [],
})
const task = await prisma.workflowTask.create({
data: {
type: 'chat',
status: 'pending',
assetId: 'a-3d',
payload: {
projectId: 'p1',
agent: {
agentId: 'b1',
userCommentId: 'comment-3d-1',
},
},
},
})
await agentChat(task)
expect(mockActivities.agentChatActivity).toHaveBeenCalledWith(
expect.objectContaining({
messageContext: expect.objectContaining({
position: { type: 'degree', degree: 180 },
}),
}),
)
})
})
+4 -1
View File
@@ -8,6 +8,7 @@ import type {
ShumaiMediaPosition,
} from '@shumai/dtos'
import { getProxyType } from '@shumai/core/src/utils/mime'
import { secondToDegree } from '@shumai/core/src/utils/model3d'
export async function agentChat(task: WorkflowTask): Promise<void> {
const {
@@ -273,7 +274,9 @@ export async function agentChat(task: WorkflowTask): Promise<void> {
// Build position
let mediaPosition: ShumaiMediaPosition | undefined
if (commentTimestamp !== undefined && commentTimestamp !== null) {
if (proxyType === 'pdf') {
if (proxyType === '3d') {
mediaPosition = { type: 'degree', degree: secondToDegree(commentTimestamp) }
} else if (proxyType === 'pdf') {
mediaPosition = { type: 'page', page: Math.round(commentTimestamp) }
} else {
mediaPosition = { type: 'time', seconds: commentTimestamp }
@@ -38,6 +38,9 @@ describe('Agent Embedding Workflow', () => {
saveAssetEmbeddingsActivity: Object.assign(vi.fn(), {
_activityName: 'saveAssetEmbeddingsActivity',
}),
deleteAssetEmbeddingsActivity: Object.assign(vi.fn(), {
_activityName: 'deleteAssetEmbeddingsActivity',
}),
updateTaskUsageActivity: Object.assign(vi.fn(), { _activityName: 'updateTaskUsageActivity' }),
createCommentActivity: Object.assign(vi.fn(), { _activityName: 'createCommentActivity' }),
updateCommentActivity: Object.assign(vi.fn(), { _activityName: 'updateCommentActivity' }),
@@ -211,6 +214,68 @@ describe('Agent Embedding Workflow', () => {
})
})
it('should run agent embedding workflow successfully for 3D asset using 4-views image', async () => {
mockActivities.getEmbeddingContextActivity.mockResolvedValue({
agent: { id: 'b1' },
asset: {
id: 'model-1',
mediaType: 'model/gltf-binary',
storageKey: { key: 'raw/model.glb' },
media: {
proxyType: '3d',
imageTranscodes: [{ key: 'files/views.webp', format: 'webp' }],
},
},
chunkDuration: 60.0,
chunkOverlap: 5.0,
})
const task = await prisma.workflowTask.create({
data: {
type: 'ai_embedding',
status: 'pending',
assetId: 'model-1',
teamId: 't1',
},
})
await agentEmbeddingMedia(task)
expect(mockActivities.generateImageEmbeddingActivity).toHaveBeenCalledWith({
teamId: 't1',
assetKey: 'files/views.webp',
mediaType: 'image/webp',
})
})
it('should reject 3D asset for embeddings if imageTranscodes is missing', async () => {
mockActivities.getEmbeddingContextActivity.mockResolvedValue({
agent: { id: 'b1' },
asset: {
id: 'model-1',
mediaType: 'model/gltf-binary',
storageKey: { key: 'raw/model.glb' },
media: {
proxyType: '3d',
imageTranscodes: [],
},
},
})
const task = await prisma.workflowTask.create({
data: {
type: 'ai_embedding',
status: 'pending',
assetId: 'model-1',
teamId: 't1',
},
})
await expect(agentEmbeddingMedia(task)).rejects.toThrow(
'unsupported media type for embeddings: model/gltf-binary',
)
})
it('should run agent embedding workflow successfully for video in chunks', async () => {
mockActivities.getEmbeddingContextActivity.mockResolvedValue({
agent: { id: 'b1' },
@@ -252,18 +317,21 @@ describe('Agent Embedding Workflow', () => {
assetKey: 'test-transcoded.mp4',
startTime: 0,
endTime: 60,
format: 'video',
})
expect(mockActivities.transcodeVideoChunkActivity).toHaveBeenNthCalledWith(2, {
assetId: 'a1',
assetKey: 'test-transcoded.mp4',
startTime: 55,
endTime: 115,
format: 'video',
})
expect(mockActivities.transcodeVideoChunkActivity).toHaveBeenNthCalledWith(3, {
assetId: 'a1',
assetKey: 'test-transcoded.mp4',
startTime: 110,
endTime: 150,
format: 'video',
})
// Verify video chunk embedding generation (3 chunks) on agent queue
@@ -295,14 +363,19 @@ describe('Agent Embedding Workflow', () => {
model: 'gemini-embedding-2',
})
// Verify embeddings saved
expect(mockActivities.saveAssetEmbeddingsActivity).toHaveBeenCalledWith({
// Verify embeddings saved per chunk immediately
expect(mockActivities.saveAssetEmbeddingsActivity).toHaveBeenCalledTimes(3)
expect(mockActivities.saveAssetEmbeddingsActivity).toHaveBeenNthCalledWith(1, {
assetId: 'a1',
embeddings: [
{ embedding: [0.3, 0.4], startTime: 0, endTime: 60 },
{ embedding: [0.3, 0.4], startTime: 55, endTime: 115 },
{ embedding: [0.3, 0.4], startTime: 110, endTime: 150 },
],
embeddings: [{ embedding: [0.3, 0.4], startTime: 0, endTime: 60 }],
})
expect(mockActivities.saveAssetEmbeddingsActivity).toHaveBeenNthCalledWith(2, {
assetId: 'a1',
embeddings: [{ embedding: [0.3, 0.4], startTime: 55, endTime: 115 }],
})
expect(mockActivities.saveAssetEmbeddingsActivity).toHaveBeenNthCalledWith(3, {
assetId: 'a1',
embeddings: [{ embedding: [0.3, 0.4], startTime: 110, endTime: 150 }],
})
})
@@ -388,6 +461,49 @@ describe('Agent Embedding Workflow', () => {
})
})
it('should save completed chunks to DB immediately before subsequent chunk failure', async () => {
mockActivities.getEmbeddingContextActivity.mockResolvedValue({
agent: { id: 'b1' },
asset: {
id: 'a1',
mediaType: 'video/mp4',
storageKey: { key: 'test.mp4' },
media: {
duration: 150.0,
videoTranscodes: [{ key: 'test-transcoded.mp4' }],
},
},
chunkDuration: 60.0,
chunkOverlap: 5.0,
})
// Chunk 1 succeeds, Chunk 2 fails
mockActivities.generateVideoChunkEmbeddingActivity
.mockResolvedValueOnce({
embedding: [0.1, 0.2],
usage: { inputTokens: 3, outputTokens: 3, model: 'gemini-embedding-2' },
})
.mockRejectedValueOnce(new Error('Rate limit exceeded at chunk 2'))
const task = await prisma.workflowTask.create({
data: {
type: 'ai_embedding',
status: 'pending',
assetId: 'a1',
teamId: 't1',
},
})
await expect(agentEmbeddingMedia(task)).rejects.toThrow('Rate limit exceeded at chunk 2')
// Chunk 1 MUST have been saved to the database before chunk 2 failed
expect(mockActivities.saveAssetEmbeddingsActivity).toHaveBeenCalledTimes(1)
expect(mockActivities.saveAssetEmbeddingsActivity).toHaveBeenCalledWith({
assetId: 'a1',
embeddings: [{ embedding: [0.1, 0.2], startTime: 0, endTime: 60 }],
})
})
it('should not create redundant chunks when video exactly matches chunk duration', async () => {
mockActivities.getEmbeddingContextActivity.mockResolvedValue({
agent: { id: 'b1' },
@@ -423,4 +539,137 @@ describe('Agent Embedding Workflow', () => {
}),
)
})
it('should run agent embedding workflow for video with gemma-embedding-2-openai requesting frames format', async () => {
mockActivities.getEmbeddingContextActivity.mockResolvedValue({
agent: { id: 'b1' },
asset: {
id: 'a1',
mediaType: 'video/mp4',
storageKey: { key: 'test.mp4' },
media: {
duration: 30.0,
},
},
chunkDuration: 60.0,
chunkOverlap: 5.0,
embeddingProvider: 'gemma-embedding-2-openai',
})
const task = await prisma.workflowTask.create({
data: {
type: 'ai_embedding',
status: 'pending',
assetId: 'a1',
teamId: 't1',
},
})
await agentEmbeddingMedia(task)
expect(mockActivities.transcodeVideoChunkActivity).toHaveBeenCalledWith(
expect.objectContaining({
startTime: 0,
endTime: 30,
format: 'frames',
}),
)
expect(mockActivities.updateTaskUsageActivity).toHaveBeenCalledWith(
expect.objectContaining({
model: 'gemma-embedding-2-openai',
}),
)
})
it('should skip already completed chunks when resuming a video embedding task', async () => {
mockActivities.getEmbeddingContextActivity.mockResolvedValue({
agent: { id: 'b1' },
asset: {
id: 'a1',
mediaType: 'video/mp4',
storageKey: { key: 'test.mp4' },
media: {
duration: 150.0,
videoTranscodes: [{ key: 'test-transcoded.mp4' }],
},
},
chunkDuration: 60.0,
chunkOverlap: 5.0,
existingEmbeddings: [
{
id: 'emb-1',
startTime: 0,
endTime: 60,
has1536: true,
has768: false,
},
],
})
const task = await prisma.workflowTask.create({
data: {
type: 'ai_embedding',
status: 'pending',
assetId: 'a1',
teamId: 't1',
},
})
await agentEmbeddingMedia(task)
// Chunk 1 (0-60) should be skipped; only chunk 2 (55-115) and chunk 3 (110-150) should be transcoded and embedded
expect(mockActivities.transcodeVideoChunkActivity).toHaveBeenCalledTimes(2)
expect(mockActivities.transcodeVideoChunkActivity).toHaveBeenNthCalledWith(
1,
expect.objectContaining({ startTime: 55, endTime: 115 }),
)
expect(mockActivities.transcodeVideoChunkActivity).toHaveBeenNthCalledWith(
2,
expect.objectContaining({ startTime: 110, endTime: 150 }),
)
expect(mockActivities.saveAssetEmbeddingsActivity).toHaveBeenCalledTimes(2)
})
it('should purge incompatible embeddings before processing', async () => {
mockActivities.getEmbeddingContextActivity.mockResolvedValue({
agent: { id: 'b1' },
asset: {
id: 'a1',
mediaType: 'image/png',
storageKey: { key: 'test.png' },
},
embeddingProvider: 'gemma-embedding-2-openai',
existingEmbeddings: [
{
id: 'old-gemini-emb',
startTime: null,
endTime: null,
has1536: true,
has768: false,
},
],
})
const task = await prisma.workflowTask.create({
data: {
type: 'ai_embedding',
status: 'pending',
assetId: 'a1',
teamId: 't1',
},
})
await agentEmbeddingMedia(task)
// Should purge incompatible old-gemini-emb
expect(mockActivities.deleteAssetEmbeddingsActivity).toHaveBeenCalledWith({
assetId: 'a1',
ids: ['old-gemini-emb'],
})
// Should generate and save new embedding
expect(mockActivities.generateImageEmbeddingActivity).toHaveBeenCalled()
expect(mockActivities.saveAssetEmbeddingsActivity).toHaveBeenCalled()
})
})
+91 -29
View File
@@ -13,6 +13,14 @@ export interface GeneratedEmbedding {
endTime?: number
}
export interface ExistingAssetEmbeddingInfo {
id: string
startTime: number | null
endTime: number | null
has1536: boolean
has768: boolean
}
export async function agentEmbeddingMedia(task: WorkflowTask): Promise<void> {
const {
updateTaskStatusActivity,
@@ -20,6 +28,7 @@ export async function agentEmbeddingMedia(task: WorkflowTask): Promise<void> {
generateImageEmbeddingActivity,
generateVideoChunkEmbeddingActivity,
saveAssetEmbeddingsActivity,
deleteAssetEmbeddingsActivity,
updateTaskUsageActivity,
createCommentActivity,
updateCommentActivity,
@@ -62,15 +71,45 @@ export async function agentEmbeddingMedia(task: WorkflowTask): Promise<void> {
}
// 1. Fetch Agent Context
const context = await executeActivity(agentWorkerQueue, getEmbeddingContextActivity, {
const context = (await executeActivity(agentWorkerQueue, getEmbeddingContextActivity, {
teamId: task.teamId,
assetId: task.assetId,
})
})) as unknown as {
asset: {
id: string
mediaType: string | null
storageKey?: { key: string } | null
media?: unknown
}
chunkDuration: number
chunkOverlap: number
embeddingProvider: string
existingEmbeddings?: ExistingAssetEmbeddingInfo[]
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
const { asset, chunkDuration, chunkOverlap } = context as any
const { asset, chunkDuration, chunkOverlap, embeddingProvider, existingEmbeddings } = context
const isGemmaProvider = embeddingProvider === 'gemma-embedding-2-openai'
const isImage = asset.mediaType?.startsWith('image/')
const existing = existingEmbeddings || []
// Purge incompatible embeddings (e.g. 1536d when Gemma is used, or 768d when Gemini is used)
const incompatibleIds = existing
.filter((e) => (isGemmaProvider ? e.has1536 || !e.has768 : e.has768 || !e.has1536))
.map((e) => e.id)
if (incompatibleIds.length > 0) {
await executeActivity(agentWorkerQueue, deleteAssetEmbeddingsActivity, {
assetId: task.assetId,
ids: incompatibleIds,
})
}
const validExistingChunks = existing.filter((e) => !incompatibleIds.includes(e.id))
const mediaInfo = asset.media as PrismaJson.MediaInfo | null
const is3dWithViews =
mediaInfo?.proxyType === '3d' && (mediaInfo.imageTranscodes?.length ?? 0) > 0
const isImage = asset.mediaType?.startsWith('image/') || is3dWithViews
const isVideo = asset.mediaType?.startsWith('video/')
if (!isImage && !isVideo) {
@@ -110,23 +149,30 @@ export async function agentEmbeddingMedia(task: WorkflowTask): Promise<void> {
throw ApplicationFailure.create({ message: 'asset has no key', nonRetryable: true })
}
const results: GeneratedEmbedding[] = []
const usage = {
model: 'gemini-embedding-2',
model: isGemmaProvider ? 'gemma-embedding-2-openai' : 'gemini-embedding-2',
inputTokens: 0,
outputTokens: 0,
}
if (isImage) {
const result = await executeActivity(agentWorkerQueue, generateImageEmbeddingActivity, {
teamId: task.teamId,
assetKey: embeddingKey,
mediaType: embeddingMediaType,
})
results.push({ embedding: result.embedding })
if (result.usage) {
usage.inputTokens += result.usage.inputTokens || 0
usage.outputTokens += result.usage.outputTokens || 0
const alreadyEmbedded = validExistingChunks.some(
(e) => e.startTime === null && e.endTime === null,
)
if (!alreadyEmbedded) {
const result = await executeActivity(agentWorkerQueue, generateImageEmbeddingActivity, {
teamId: task.teamId,
assetKey: embeddingKey,
mediaType: embeddingMediaType,
})
await executeActivity(agentWorkerQueue, saveAssetEmbeddingsActivity, {
assetId: task.assetId,
embeddings: [{ embedding: result.embedding }],
})
if (result.usage) {
usage.inputTokens += result.usage.inputTokens || 0
usage.outputTokens += result.usage.outputTokens || 0
}
}
} else if (isVideo) {
// Get video duration
@@ -142,12 +188,30 @@ export async function agentEmbeddingMedia(task: WorkflowTask): Promise<void> {
end = duration
}
// Check if this chunk is already completed
const chunkExists = validExistingChunks.some(
(c) =>
c.startTime !== null &&
c.endTime !== null &&
Math.abs(c.startTime - start) < 0.1 &&
Math.abs(c.endTime - end) < 0.1,
)
if (chunkExists) {
// If this chunk reached the end of the video, stop
if (end >= duration) {
break
}
continue
}
// Step A: Slice chunk on transcode worker and upload to S3
const chunkRes = await executeActivity(transcodeWorkerQueue, transcodeVideoChunkActivity, {
assetId: asset.id,
assetKey: embeddingKey,
startTime: start,
endTime: end,
format: isGemmaProvider ? 'frames' : 'video',
})
const { chunkKey } = chunkRes
@@ -162,10 +226,16 @@ export async function agentEmbeddingMedia(task: WorkflowTask): Promise<void> {
},
)
results.push({
embedding: chunkResult.embedding,
startTime: start,
endTime: end,
// Step C: Save chunk embedding to database immediately
await executeActivity(agentWorkerQueue, saveAssetEmbeddingsActivity, {
assetId: task.assetId,
embeddings: [
{
embedding: chunkResult.embedding,
startTime: start,
endTime: end,
},
],
})
if (chunkResult.usage) {
@@ -173,7 +243,7 @@ export async function agentEmbeddingMedia(task: WorkflowTask): Promise<void> {
usage.outputTokens += chunkResult.usage.outputTokens || 0
}
} finally {
// Step C: Delete temporary chunk from S3
// Step D: Delete temporary chunk from S3
await executeActivity(transcodeWorkerQueue, deleteS3ObjectActivity, {
key: chunkKey,
})
@@ -186,14 +256,6 @@ export async function agentEmbeddingMedia(task: WorkflowTask): Promise<void> {
}
}
// Save computed embeddings
if (results.length > 0) {
await executeActivity(agentWorkerQueue, saveAssetEmbeddingsActivity, {
assetId: task.assetId,
embeddings: results,
})
}
// Update Usage
await executeActivity(agentWorkerQueue, updateTaskUsageActivity, {
taskId: task.id,
@@ -1,145 +0,0 @@
import { describe, it, expect, vi, beforeEach } from 'vitest'
import { queryEmbeddingForSearch } from './query-embedding-for-search'
import { prisma } from '@shumai/db'
import { setupTestDbHooks } from '@shumai/db/test'
import * as workflowUtils from '@shumai/workflow-core'
vi.mock('@shumai/workflow-core', async () => {
const actual = await vi.importActual('@shumai/workflow-core')
return {
...actual,
getActivities: vi.fn(),
executeActivity: vi.fn(),
}
})
describe('Query Embedding For Search Workflow', () => {
setupTestDbHooks()
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- mockActivities holds mock functions cast to expected types
let mockActivities: any
beforeEach(() => {
vi.clearAllMocks()
mockActivities = {
updateWorkflowTaskActivity: Object.assign(vi.fn(), {
_activityName: 'updateWorkflowTaskActivity',
}),
generateTextEmbeddingActivity: Object.assign(vi.fn(), {
_activityName: 'generateTextEmbeddingActivity',
}),
getAgentWorkerQueueActivity: Object.assign(vi.fn(), {
_activityName: 'getAgentWorkerQueueActivity',
}),
}
mockActivities.getAgentWorkerQueueActivity.mockResolvedValue('agent_queue')
mockActivities.updateWorkflowTaskActivity.mockResolvedValue({})
mockActivities.generateTextEmbeddingActivity.mockResolvedValue({
embedding: [0.1, 0.2, 0.3],
usage: { inputTokens: 4, outputTokens: 0, model: 'gemini-embedding-2' },
})
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- mockActivities contains vi.fn mock functions which are cast to expected activity proxy types
vi.mocked(workflowUtils.getActivities).mockReturnValue(mockActivities as any)
vi.mocked(workflowUtils.executeActivity).mockImplementation(async (_queue, act, ...args) => {
// eslint-disable-next-line @typescript-eslint/no-explicit-any -- act is one of the mocked activities
return (act as any)(...args)
})
})
it('should run query embedding workflow successfully', async () => {
const task = await prisma.workflowTask.create({
data: {
type: 'query_embedding_for_search',
status: 'pending',
teamId: 't1',
assetId: 'a1',
payload: {
projectId: 'p1',
queryEmbeddingForSearch: {
text: 'search query',
},
},
},
})
await queryEmbeddingForSearch(task)
// Verify queue discovery
expect(mockActivities.getAgentWorkerQueueActivity).toHaveBeenCalled()
// Verify task processing status
expect(mockActivities.updateWorkflowTaskActivity).toHaveBeenCalledWith({
taskId: task.id,
status: 'processing',
heartbeat: true,
})
// Verify embedding generation
expect(mockActivities.generateTextEmbeddingActivity).toHaveBeenCalledWith({
text: 'search query',
teamId: 't1',
})
// Verify completed status and details
expect(mockActivities.updateWorkflowTaskActivity).toHaveBeenCalledWith({
taskId: task.id,
status: 'completed',
output: { embedding: [0.1, 0.2, 0.3] },
inputTokens: 4,
outputTokens: 0,
model: 'gemini-embedding-2',
})
})
it('should throw immediately if text is missing', async () => {
const task = await prisma.workflowTask.create({
data: {
type: 'query_embedding_for_search',
status: 'pending',
teamId: 't1',
assetId: 'a1',
},
})
await expect(queryEmbeddingForSearch(task)).rejects.toThrow(
'Missing text in query_embedding_for_search task payload',
)
// Should not call any activities because it throws before queue discovery
expect(mockActivities.getAgentWorkerQueueActivity).not.toHaveBeenCalled()
expect(mockActivities.updateWorkflowTaskActivity).not.toHaveBeenCalled()
})
it('should handle activity failure and update task status to failed', async () => {
mockActivities.generateTextEmbeddingActivity.mockRejectedValue(
new Error('Embedding API failed'),
)
const task = await prisma.workflowTask.create({
data: {
type: 'query_embedding_for_search',
status: 'pending',
teamId: 't1',
assetId: 'a1',
payload: {
projectId: 'p1',
queryEmbeddingForSearch: {
text: 'search query',
},
},
},
})
await expect(queryEmbeddingForSearch(task)).rejects.toThrow('Embedding API failed')
// Verify it updates task to failed
expect(mockActivities.updateWorkflowTaskActivity).toHaveBeenCalledWith({
taskId: task.id,
status: 'failed',
output: { error: 'Embedding API failed' },
})
})
})
@@ -1,56 +0,0 @@
import type { WorkflowTask } from '@shumai/db'
import { getActivities, executeActivity, TaskQueueAgent } from '@shumai/workflow-core'
export async function queryEmbeddingForSearch(task: WorkflowTask): Promise<void> {
const { generateTextEmbeddingActivity, updateWorkflowTaskActivity, getAgentWorkerQueueActivity } =
getActivities()
let agentWorkerQueue = ''
const payload = task.payload as PrismaJson.WorkflowTaskPayload | null
const text = payload?.queryEmbeddingForSearch?.text
if (!text) {
throw new Error('Missing text in query_embedding_for_search task payload')
}
try {
// 0. Discover queue
agentWorkerQueue = await executeActivity(TaskQueueAgent, getAgentWorkerQueueActivity)
// 1. Mark as processing
await executeActivity(agentWorkerQueue, updateWorkflowTaskActivity, {
taskId: task.id,
status: 'processing',
heartbeat: true,
})
// 2. Generate embedding
const result = await executeActivity(agentWorkerQueue, generateTextEmbeddingActivity, {
text,
teamId: task.teamId!,
})
// 3. Save output and complete
await executeActivity(agentWorkerQueue, updateWorkflowTaskActivity, {
taskId: task.id,
status: 'completed',
output: { embedding: result.embedding },
inputTokens: result.usage.inputTokens,
outputTokens: result.usage.outputTokens,
model: result.usage.model,
})
} catch (err) {
console.error(`queryEmbeddingForSearch failed for task ${task.id}:`, err)
if (agentWorkerQueue) {
await executeActivity(agentWorkerQueue, updateWorkflowTaskActivity, {
taskId: task.id,
status: 'failed',
output: {
error: err instanceof Error ? err.message : String(err),
},
})
}
throw err
}
}
+1 -1
View File
@@ -1,6 +1,6 @@
{
"name": "@shumai/api",
"version": "0.4.6",
"version": "0.5.3",
"private": true,
"type": "module",
"main": "./src/index.ts",
+199 -1
View File
@@ -1,4 +1,4 @@
import { describe, it, expect, vi, beforeEach } from 'vitest'
import { describe, it, expect, vi, beforeEach, afterEach } from 'vitest'
import { HTTPException } from 'hono/http-exception'
import { agentService } from '@shumai/core/src/agent/agent'
import { authzService } from '@shumai/core/src/authz/authz'
@@ -210,6 +210,30 @@ describe('Agent API', () => {
itemId: 'agent1',
})
})
it('returns 403 when updating embedding agent and ENABLE_EMBEDDING_AGENT is true', async () => {
process.env.ENABLE_EMBEDDING_AGENT = 'true'
try {
vi.mocked(agentService.getAgent).mockResolvedValueOnce({
id: 'agent1',
teamId: 'team1',
type: 'embedding',
} as unknown as Awaited<ReturnType<typeof agentService.getAgent>>)
const res = await app.request('/agents/agent1', {
method: 'PUT',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
name: 'Updated Embedding Agent',
type: 'embedding',
}),
})
expect(res.status).toBe(403)
} finally {
delete process.env.ENABLE_EMBEDDING_AGENT
}
})
})
describe('DELETE /agents/:agentId', () => {
@@ -227,6 +251,25 @@ describe('Agent API', () => {
itemId: 'agent1',
})
})
it('returns 403 when deleting embedding agent and ENABLE_EMBEDDING_AGENT is true', async () => {
process.env.ENABLE_EMBEDDING_AGENT = 'true'
try {
vi.mocked(agentService.getAgent).mockResolvedValueOnce({
id: 'agent1',
teamId: 'team1',
type: 'embedding',
} as unknown as Awaited<ReturnType<typeof agentService.getAgent>>)
const res = await app.request('/agents/agent1', {
method: 'DELETE',
})
expect(res.status).toBe(403)
} finally {
delete process.env.ENABLE_EMBEDDING_AGENT
}
})
})
describe('PATCH /agents/:agentId/permission', () => {
@@ -271,6 +314,27 @@ describe('Agent API', () => {
})
})
it('returns 403 when modifying embedding agent permission and ENABLE_EMBEDDING_AGENT is true', async () => {
process.env.ENABLE_EMBEDDING_AGENT = 'true'
try {
vi.mocked(agentService.getAgent).mockResolvedValueOnce({
id: 'agent1',
teamId: 'team1',
type: 'embedding',
} as unknown as Awaited<ReturnType<typeof agentService.getAgent>>)
const res = await app.request('/agents/agent1/permission', {
method: 'PATCH',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({ permission: 'owner' }),
})
expect(res.status).toBe(403)
} finally {
delete process.env.ENABLE_EMBEDDING_AGENT
}
})
it('denies access when user is not admin', async () => {
vi.mocked(authzService.hasPermission).mockRejectedValue(
new HTTPException(403, { message: 'Forbidden' }),
@@ -410,4 +474,138 @@ describe('Agent API', () => {
expect(agentService.listProjectChatAgents).not.toHaveBeenCalled()
})
})
describe('read-only embedding agent when ENABLE_EMBEDDING_AGENT is true', () => {
beforeEach(() => {
process.env.ENABLE_EMBEDDING_AGENT = 'true'
vi.mocked(agentService.getAgent).mockResolvedValue({
id: 'emb1',
teamId: 'team1',
type: 'embedding',
} as unknown as Awaited<ReturnType<typeof agentService.getAgent>>)
})
afterEach(() => {
delete process.env.ENABLE_EMBEDDING_AGENT
})
it('rejects PUT /agents/:agentId with 403', async () => {
const res = await app.request('/agents/emb1', {
method: 'PUT',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
name: 'Updated Embedding',
type: 'embedding',
enabled: false,
}),
})
expect(res.status).toBe(403)
expect(agentService.updateAgent).not.toHaveBeenCalled()
})
it('rejects DELETE /agents/:agentId with 403', async () => {
const res = await app.request('/agents/emb1', {
method: 'DELETE',
headers: { 'Content-Type': 'application/json' },
})
expect(res.status).toBe(403)
expect(agentService.deleteAgent).not.toHaveBeenCalled()
})
it('rejects PATCH /agents/:agentId/permission with 403', async () => {
const res = await app.request('/agents/emb1/permission', {
method: 'PATCH',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
permission: 'owner',
}),
})
expect(res.status).toBe(403)
expect(agentService.updateAgentPermission).not.toHaveBeenCalled()
})
it('rejects POST /teams/:teamId/agents with 403 for embedding agent', async () => {
const res = await app.request('/teams/team1/agents', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
name: 'Manual Embedding',
type: 'embedding',
}),
})
expect(res.status).toBe(403)
expect(agentService.createAgent).not.toHaveBeenCalled()
})
})
describe('read-only embedding agent when ENABLE_EMBEDDING_AGENT is false', () => {
beforeEach(() => {
process.env.ENABLE_EMBEDDING_AGENT = 'false'
vi.mocked(agentService.getAgent).mockResolvedValue({
id: 'emb1',
teamId: 'team1',
type: 'embedding',
} as unknown as Awaited<ReturnType<typeof agentService.getAgent>>)
})
afterEach(() => {
delete process.env.ENABLE_EMBEDDING_AGENT
})
it('rejects PUT /agents/:agentId with 403', async () => {
const res = await app.request('/agents/emb1', {
method: 'PUT',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
name: 'Updated Embedding',
type: 'embedding',
enabled: false,
}),
})
expect(res.status).toBe(403)
expect(agentService.updateAgent).not.toHaveBeenCalled()
})
it('rejects DELETE /agents/:agentId with 403', async () => {
const res = await app.request('/agents/emb1', {
method: 'DELETE',
headers: { 'Content-Type': 'application/json' },
})
expect(res.status).toBe(403)
expect(agentService.deleteAgent).not.toHaveBeenCalled()
})
it('rejects PATCH /agents/:agentId/permission with 403', async () => {
const res = await app.request('/agents/emb1/permission', {
method: 'PATCH',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
permission: 'owner',
}),
})
expect(res.status).toBe(403)
expect(agentService.updateAgentPermission).not.toHaveBeenCalled()
})
it('rejects POST /teams/:teamId/agents with 403 for embedding agent', async () => {
const res = await app.request('/teams/team1/agents', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
name: 'Manual Embedding',
type: 'embedding',
}),
})
expect(res.status).toBe(403)
expect(agentService.createAgent).not.toHaveBeenCalled()
})
})
})
+42
View File
@@ -1,4 +1,5 @@
import { Hono } from 'hono'
import { HTTPException } from 'hono/http-exception'
import { zValidator } from '@hono/zod-validator'
import { agentService } from '@shumai/core/src/agent/agent'
import { authzService, Permission, ResourceType } from '@shumai/core/src/authz/authz'
@@ -62,6 +63,16 @@ const route = new Hono<{ Variables: { user: User } }>()
id: teamId,
})
const isEmbeddingManagedByEnv =
process.env.ENABLE_EMBEDDING_AGENT === 'true' ||
process.env.ENABLE_EMBEDDING_AGENT === 'false'
if (req.type === 'embedding' && isEmbeddingManagedByEnv) {
throw new HTTPException(403, {
message: 'Embedding agent is managed by environment configuration and cannot be created',
})
}
const agent = await agentService.createAgent({
teamId,
...req,
@@ -92,6 +103,17 @@ const route = new Hono<{ Variables: { user: User } }>()
id: agentId,
})
const isEmbeddingManagedByEnv =
process.env.ENABLE_EMBEDDING_AGENT === 'true' ||
process.env.ENABLE_EMBEDDING_AGENT === 'false'
const existingAgent = await agentService.getAgent({ agentId })
if (existingAgent?.type === 'embedding' && isEmbeddingManagedByEnv) {
throw new HTTPException(403, {
message: 'Embedding agent is managed by environment configuration and cannot be modified',
})
}
const agent = await agentService.updateAgent({
agentId,
...req,
@@ -123,6 +145,17 @@ const route = new Hono<{ Variables: { user: User } }>()
id: agentId,
})
const isEmbeddingManagedByEnv =
process.env.ENABLE_EMBEDDING_AGENT === 'true' ||
process.env.ENABLE_EMBEDDING_AGENT === 'false'
const existingAgent = await agentService.getAgent({ agentId })
if (existingAgent?.type === 'embedding' && isEmbeddingManagedByEnv) {
throw new HTTPException(403, {
message: 'Embedding agent is managed by environment configuration and cannot be modified',
})
}
const agent = await agentService.updateAgentPermission(agentId, req.permission)
await auditLogService.logAction({
@@ -148,7 +181,16 @@ const route = new Hono<{ Variables: { user: User } }>()
id: agentId,
})
const isEmbeddingManagedByEnv =
process.env.ENABLE_EMBEDDING_AGENT === 'true' ||
process.env.ENABLE_EMBEDDING_AGENT === 'false'
const existingAgent = await agentService.getAgent({ agentId })
if (existingAgent?.type === 'embedding' && isEmbeddingManagedByEnv) {
throw new HTTPException(403, {
message: 'Embedding agent is managed by environment configuration and cannot be deleted',
})
}
await agentService.deleteAgent({
agentId,
+43
View File
@@ -774,4 +774,47 @@ describe('file api', () => {
mockGetProjectIds.mockRestore()
})
describe('GET /files/:fileId/m3u8 routes', () => {
it('returns master playlist with correct content-type', async () => {
const mockMaster = vi
.spyOn(assetService, 'getHlsMasterPlaylist')
.mockResolvedValue('#EXTM3U\n#EXT-X-STREAM-INF:BANDWIDTH=4628000\n1080p/index.m3u8\n')
const app = new Hono().use('*', authMiddleware).route('/', fileRoute)
const res = await app.request('/files/asset-1/m3u8/master.m3u8')
expect(res.status).toBe(200)
expect(res.headers.get('content-type')).toContain('application/vnd.apple.mpegurl')
const text = await res.text()
expect(text).toContain('#EXTM3U')
expect(text).toContain('1080p/index.m3u8')
expect(mockMaster).toHaveBeenCalledWith({ assetId: 'asset-1' })
expect(authzService.hasPermission).toHaveBeenCalledWith({
user: { id: 'user1', name: 'Test User' },
permission: Permission.Read,
type: 'asset',
id: 'asset-1',
})
mockMaster.mockRestore()
})
it('returns variant playlist with rewritten URLs', async () => {
const mockVariant = vi
.spyOn(assetService, 'getHlsVariantPlaylist')
.mockResolvedValue(
'#EXTM3U\n#EXT-X-MAP:URI="http://presigned/init.mp4"\nhttp://presigned/seg0.m4s\n',
)
const app = new Hono().use('*', authMiddleware).route('/', fileRoute)
const res = await app.request('/files/asset-1/m3u8/1080p/index.m3u8')
expect(res.status).toBe(200)
expect(res.headers.get('content-type')).toContain('application/vnd.apple.mpegurl')
const text = await res.text()
expect(text).toContain('http://presigned/init.mp4')
expect(mockVariant).toHaveBeenCalledWith({ assetId: 'asset-1', resolution: '1080p' })
mockVariant.mockRestore()
})
})
})
+65
View File
@@ -480,5 +480,70 @@ const route = new Hono<{ Variables: { user: User } }>()
return c.json({ url })
},
)
.get('/files/:fileId/m3u8', async (c) => {
const fileId = c.req.param('fileId')
const user = c.get('user')
await authzService.hasPermission({
user,
permission: Permission.Read,
type: ResourceType.Asset,
id: fileId,
})
const playlist = await assetService.getHlsMasterPlaylist({ assetId: fileId })
c.header('Content-Type', 'application/vnd.apple.mpegurl')
return c.body(playlist)
})
.get('/files/:fileId/m3u8/master.m3u8', async (c) => {
const fileId = c.req.param('fileId')
const user = c.get('user')
await authzService.hasPermission({
user,
permission: Permission.Read,
type: ResourceType.Asset,
id: fileId,
})
const playlist = await assetService.getHlsMasterPlaylist({ assetId: fileId })
c.header('Content-Type', 'application/vnd.apple.mpegurl')
return c.body(playlist)
})
.get('/files/:fileId/m3u8/:resolution', async (c) => {
const fileId = c.req.param('fileId')
let resolution = c.req.param('resolution')
if (resolution.endsWith('.m3u8')) {
resolution = resolution.replace(/\.m3u8$/, '')
}
const user = c.get('user')
await authzService.hasPermission({
user,
permission: Permission.Read,
type: ResourceType.Asset,
id: fileId,
})
const playlist = await assetService.getHlsVariantPlaylist({ assetId: fileId, resolution })
c.header('Content-Type', 'application/vnd.apple.mpegurl')
return c.body(playlist)
})
.get('/files/:fileId/m3u8/:resolution/index.m3u8', async (c) => {
const fileId = c.req.param('fileId')
const resolution = c.req.param('resolution')
const user = c.get('user')
await authzService.hasPermission({
user,
permission: Permission.Read,
type: ResourceType.Asset,
id: fileId,
})
const playlist = await assetService.getHlsVariantPlaylist({ assetId: fileId, resolution })
c.header('Content-Type', 'application/vnd.apple.mpegurl')
return c.body(playlist)
})
export default route
+110 -7
View File
@@ -2,27 +2,34 @@ import { describe, expect, it, vi, beforeEach } from 'vitest'
import { Hono, type Context, type Next } from 'hono'
import notificationRoute from './notification'
import { notificationService } from '@shumai/core/src/notification/notification'
import { emailService } from '@shumai/core/src/notification/email'
import { authzService, Permission, ResourceType } from '@shumai/core/src/authz/authz'
import { userMetadataService } from '@shumai/core/src/user-metadata/user-metadata'
const mockUser = { id: 'user1', name: 'Test User', email: 'test@example.com' }
vi.mock('./middleware/auth', () => ({
authMiddleware: async (
c: Context<{ Variables: { user: { id: string; name: string } } }>,
next: Next,
) => {
c.set('user', { id: 'user1', name: 'Test User' })
authMiddleware: async (c: Context<{ Variables: { user: typeof mockUser } }>, next: Next) => {
c.set('user', mockUser)
await next()
},
}))
vi.mock('@shumai/core/src/authz/authz')
vi.mock('@shumai/core/src/notification/notification')
vi.mock('@shumai/core/src/notification/email', () => ({
emailService: {
getEmailSettings: vi.fn(),
updateEmailSettings: vi.fn(),
sendTestEmail: vi.fn(),
},
}))
vi.mock('@shumai/core/src/user-metadata/user-metadata')
describe('notification api', () => {
const app = new Hono<{ Variables: { user: { id: string; name: string } } }>()
const app = new Hono<{ Variables: { user: typeof mockUser } }>()
.use('*', async (c, next) => {
c.set('user', { id: 'user1', name: 'Test User' })
c.set('user', mockUser)
await next()
})
.route('/', notificationRoute)
@@ -155,4 +162,100 @@ describe('notification api', () => {
},
)
})
it('GET /teams/:teamId/notifications/email - requires admin and returns settings', async () => {
const mockSettings = {
enabled: true,
host: 'smtp.example.com',
port: 587,
username: 'user',
password: 'secret-password',
smtps: true,
ignoreCert: false,
from: 'noreply@example.com',
replyTo: '',
}
vi.mocked(emailService.getEmailSettings).mockResolvedValue(mockSettings)
const res = await app.request('/teams/t1/notifications/email')
expect(res.status).toBe(200)
const json = await res.json()
expect(json).toEqual(mockSettings)
expect(authzService.hasPermission).toHaveBeenCalledWith({
user: expect.objectContaining({ id: 'user1' }),
permission: Permission.Admin,
type: ResourceType.Team,
id: 't1',
})
expect(emailService.getEmailSettings).toHaveBeenCalledWith('t1')
})
it('PUT /teams/:teamId/notifications/email - requires admin and updates settings', async () => {
const updateInput = {
enabled: true,
host: 'mail.example.com',
port: 465,
username: 'user',
password: 'new-password',
smtps: true,
ignoreCert: false,
from: 'shumai@example.com',
replyTo: 'support@example.com',
}
vi.mocked(emailService.updateEmailSettings).mockResolvedValue(updateInput)
const res = await app.request('/teams/t1/notifications/email', {
method: 'PUT',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify(updateInput),
})
expect(res.status).toBe(200)
const json = await res.json()
expect(json).toEqual(updateInput)
expect(authzService.hasPermission).toHaveBeenCalledWith({
user: expect.objectContaining({ id: 'user1' }),
permission: Permission.Admin,
type: ResourceType.Team,
id: 't1',
})
expect(emailService.updateEmailSettings).toHaveBeenCalledWith('t1', updateInput)
})
it('POST /teams/:teamId/notifications/email/test - requires admin and sends test email', async () => {
const testReq = {
host: 'smtp.example.com',
port: 587,
from: 'noreply@example.com',
}
vi.mocked(emailService.sendTestEmail).mockResolvedValue({
success: true,
messageId: '<test-msg-id@example.com>',
})
const res = await app.request('/teams/t1/notifications/email/test', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify(testReq),
})
expect(res.status).toBe(200)
const json = await res.json()
expect(json).toEqual({
success: true,
messageId: '<test-msg-id@example.com>',
})
expect(authzService.hasPermission).toHaveBeenCalledWith({
user: expect.objectContaining({ id: 'user1' }),
permission: Permission.Admin,
type: ResourceType.Team,
id: 't1',
})
expect(emailService.sendTestEmail).toHaveBeenCalledWith(
't1',
expect.objectContaining({ id: 'user1', email: 'test@example.com' }),
testReq,
)
})
})
+55
View File
@@ -1,11 +1,14 @@
import { zValidator } from '@hono/zod-validator'
import { Hono } from 'hono'
import { notificationService } from '@shumai/core/src/notification/notification'
import { emailService } from '@shumai/core/src/notification/email'
import { authzService, Permission, ResourceType } from '@shumai/core/src/authz/authz'
import {
listNotificationsRequestSchema,
markNotificationReadRequestSchema,
notificationSettingsSchema,
updateEmailNotificationSettingsSchema,
testEmailRequestSchema,
} from '@shumai/dtos'
import { userMetadataService } from '@shumai/core/src/user-metadata/user-metadata'
import type { Prisma } from '@shumai/db'
@@ -109,5 +112,57 @@ const route = new Hono<{ Variables: { user: User } }>()
return c.json({ success: true })
},
)
.get('/teams/:teamId/notifications/email', async (c) => {
const teamId = c.req.param('teamId')
const user = c.get('user')
await authzService.hasPermission({
user,
permission: Permission.Admin,
type: ResourceType.Team,
id: teamId,
})
const settings = await emailService.getEmailSettings(teamId)
return c.json(settings)
})
.put(
'/teams/:teamId/notifications/email',
zValidator('json', updateEmailNotificationSettingsSchema),
async (c) => {
const teamId = c.req.param('teamId')
const user = c.get('user')
const req = c.req.valid('json')
await authzService.hasPermission({
user,
permission: Permission.Admin,
type: ResourceType.Team,
id: teamId,
})
const settings = await emailService.updateEmailSettings(teamId, req)
return c.json(settings)
},
)
.post(
'/teams/:teamId/notifications/email/test',
zValidator('json', testEmailRequestSchema),
async (c) => {
const teamId = c.req.param('teamId')
const user = c.get('user')
const req = c.req.valid('json')
await authzService.hasPermission({
user,
permission: Permission.Admin,
type: ResourceType.Team,
id: teamId,
})
const result = await emailService.sendTestEmail(teamId, user, req)
return c.json(result)
},
)
export default route
+151 -9
View File
@@ -38,9 +38,12 @@ import type { AssetInfo } from '@shumai/dtos'
async function applyWatermarkToAssetMedia(
asset: AssetInfo,
watermarkMedia: PrismaJson.MediaInfo | null,
shareId?: string,
password?: string,
): Promise<AssetInfo> {
if (!asset.media) return asset
if (asset.proxyType !== 'video' && asset.proxyType !== 'image') return asset
if (asset.proxyType !== 'video' && asset.proxyType !== 'image' && asset.proxyType !== '3d')
return asset
const updatedMedia = { ...asset.media }
const bucket = process.env.S3_BUCKET || 'shumai'
@@ -55,10 +58,28 @@ async function applyWatermarkToAssetMedia(
width: vt.width ?? 0,
height: vt.height ?? 0,
size: 0,
hdr: vt.hdr,
resolution: vt.resolution,
})),
)
: []
if (watermarkMedia?.videoTranscodes && watermarkMedia.videoTranscodes.length > 0) {
if (updatedMedia.videoPreview) {
const watermarkedKey =
watermarkMedia.videoPreview?.key || watermarkMedia.videoTranscodes[0]?.key
if (watermarkedKey) {
updatedMedia.videoPreview = {
...updatedMedia.videoPreview,
key: watermarkedKey,
url: await s3Service.presign(bucket, watermarkedKey, 'GET'),
}
}
}
} else {
delete updatedMedia.videoPreview
}
updatedMedia.imageTranscodes =
watermarkMedia?.imageTranscodes && watermarkMedia.imageTranscodes.length > 0
? await Promise.all(
@@ -73,6 +94,19 @@ async function applyWatermarkToAssetMedia(
)
: []
if (watermarkMedia?.isHls && watermarkMedia.hls?.key && shareId) {
const pwQuery = password ? `?p=${encodeURIComponent(password)}` : ''
updatedMedia.isHls = true
updatedMedia.hls = {
key: watermarkMedia.hls.key,
url: `/api/shares/${shareId}/files/${asset.id}/m3u8/master.m3u8${pwQuery}`,
resolutions: watermarkMedia.hls.resolutions,
}
} else {
updatedMedia.isHls = false
delete updatedMedia.hls
}
return {
...asset,
media: updatedMedia,
@@ -169,7 +203,8 @@ const route = app
zValidator('query', sharedChildrenRequestSchema),
async (c) => {
const folderId = c.req.param('folderId')
const password = c.req.header('x-share-password')
const password =
c.req.header('x-share-password') || c.req.query('password') || c.req.query('p')
const req = c.req.valid('query')
try {
@@ -196,7 +231,7 @@ const route = app
// Resolve symlink targets once and reuse for the media lookup, so each
// item costs a single query instead of two.
const targetIds = await Promise.all(
res.data.map((item) => assetService.resolveTargetAssetId(item.id)),
res.data.map((item) => assetService.resolveLatestVersionId(item.id)),
)
const wfMap = await watermarkService.getCompletedWatermarkMediaMap(
targetIds,
@@ -205,9 +240,21 @@ const route = app
res.data = await Promise.all(
res.data.map(async (item, index) =>
applyWatermarkToAssetMedia(item, wfMap.get(targetIds[index]) ?? null),
applyWatermarkToAssetMedia(
item,
wfMap.get(targetIds[index]) ?? null,
shareLink.id,
password,
),
),
)
} else if (res.data.length > 0) {
const pwQuery = password ? `?p=${encodeURIComponent(password)}` : ''
for (const item of res.data) {
if (item.media?.isHls && item.media.hls?.key) {
item.media.hls.url = `/api/shares/${shareLink.id}/files/${item.id}/m3u8/master.m3u8${pwQuery}`
}
}
}
return c.json(res)
@@ -218,25 +265,33 @@ const route = app
)
.get('/shares/:shareId/files/:fileId', async (c) => {
const fileId = c.req.param('fileId')
const password = c.req.header('x-share-password')
const password = c.req.header('x-share-password') || c.req.query('password') || c.req.query('p')
try {
const shareLink = await shareService.verifyPublicAccess(fileId, password)
let asset = await assetService.getAsset({ assetId: fileId })
// Only video/image assets are subject to watermark protection; skip the
// Only video/image/3d assets are subject to watermark protection; skip the
// watermark lookup entirely for other media types (pdf, audio, ...).
if (
shareLink.watermarkConfigId &&
(asset.proxyType === 'video' || asset.proxyType === 'image')
(asset.proxyType === 'video' || asset.proxyType === 'image' || asset.proxyType === '3d')
) {
const targetAssetId = await assetService.resolveTargetAssetId(fileId)
const targetAssetId = await assetService.resolveLatestVersionId(fileId)
const wfMap = await watermarkService.getCompletedWatermarkMediaMap(
[targetAssetId],
shareLink.watermarkConfigId,
)
asset = await applyWatermarkToAssetMedia(asset, wfMap.get(targetAssetId) ?? null)
asset = await applyWatermarkToAssetMedia(
asset,
wfMap.get(targetAssetId) ?? null,
shareLink.id,
password,
)
} else if (asset.media?.isHls && asset.media.hls?.key) {
const pwQuery = password ? `?p=${encodeURIComponent(password)}` : ''
asset.media.hls.url = `/api/shares/${shareLink.id}/files/${asset.id}/m3u8/master.m3u8${pwQuery}`
}
return c.json(asset)
@@ -359,5 +414,92 @@ const route = app
}
},
)
.get('/shares/:shareId/files/:fileId/m3u8', async (c) => {
const fileId = c.req.param('fileId')
const password = c.req.header('x-share-password') || c.req.query('password') || c.req.query('p')
try {
const shareLink = await shareService.verifyPublicAccess(fileId, password)
const targetAssetId = await assetService.resolveLatestVersionId(fileId)
let playlist = await assetService.getHlsMasterPlaylist({
assetId: targetAssetId,
watermarkConfigId: shareLink.watermarkConfigId,
})
if (password) {
playlist = playlist.replace(
/([0-9]+p\/index\.m3u8)/g,
`$1?p=${encodeURIComponent(password)}`,
)
}
c.header('Content-Type', 'application/vnd.apple.mpegurl')
return c.body(playlist)
} catch (err) {
return handlePublicShareError(c, err)
}
})
.get('/shares/:shareId/files/:fileId/m3u8/master.m3u8', async (c) => {
const fileId = c.req.param('fileId')
const password = c.req.header('x-share-password') || c.req.query('password') || c.req.query('p')
try {
const shareLink = await shareService.verifyPublicAccess(fileId, password)
const targetAssetId = await assetService.resolveLatestVersionId(fileId)
let playlist = await assetService.getHlsMasterPlaylist({
assetId: targetAssetId,
watermarkConfigId: shareLink.watermarkConfigId,
})
if (password) {
playlist = playlist.replace(
/([0-9]+p\/index\.m3u8)/g,
`$1?p=${encodeURIComponent(password)}`,
)
}
c.header('Content-Type', 'application/vnd.apple.mpegurl')
return c.body(playlist)
} catch (err) {
return handlePublicShareError(c, err)
}
})
.get('/shares/:shareId/files/:fileId/m3u8/:resolution', async (c) => {
const fileId = c.req.param('fileId')
let resolution = c.req.param('resolution')
if (resolution.endsWith('.m3u8')) {
resolution = resolution.replace(/\.m3u8$/, '')
}
const password = c.req.header('x-share-password') || c.req.query('password') || c.req.query('p')
try {
const shareLink = await shareService.verifyPublicAccess(fileId, password)
const targetAssetId = await assetService.resolveLatestVersionId(fileId)
const playlist = await assetService.getHlsVariantPlaylist({
assetId: targetAssetId,
resolution,
watermarkConfigId: shareLink.watermarkConfigId,
})
c.header('Content-Type', 'application/vnd.apple.mpegurl')
return c.body(playlist)
} catch (err) {
return handlePublicShareError(c, err)
}
})
.get('/shares/:shareId/files/:fileId/m3u8/:resolution/index.m3u8', async (c) => {
const fileId = c.req.param('fileId')
const resolution = c.req.param('resolution')
const password = c.req.header('x-share-password') || c.req.query('password') || c.req.query('p')
try {
const shareLink = await shareService.verifyPublicAccess(fileId, password)
const targetAssetId = await assetService.resolveLatestVersionId(fileId)
const playlist = await assetService.getHlsVariantPlaylist({
assetId: targetAssetId,
resolution,
watermarkConfigId: shareLink.watermarkConfigId,
})
c.header('Content-Type', 'application/vnd.apple.mpegurl')
return c.body(playlist)
} catch (err) {
return handlePublicShareError(c, err)
}
})
export default route
+273 -1
View File
@@ -277,6 +277,12 @@ describe('Share API', () => {
},
],
imageTranscodes: [],
isHls: true,
hls: {
key: 'files/orig-hls/master.m3u8',
url: '/api/files/video-pending/m3u8/master.m3u8',
resolutions: [{ resolution: '1080p', width: 1920, height: 1080 }],
},
original: { key: 'files/orig.mp4' },
},
},
@@ -331,9 +337,11 @@ describe('Share API', () => {
size: 0,
},
])
// Pending item → original transcodes emptied
// Pending item → original transcodes emptied and HLS cleared
expect(body.data[1].media.videoTranscodes).toEqual([])
expect(body.data[1].media.imageTranscodes).toEqual([])
expect(body.data[1].media.isHls).toBe(false)
expect(body.data[1].media.hls).toBeUndefined()
// Non-media item → untouched
expect(body.data[2].media.pdfTranscode).toEqual({ url: 'u', key: 'files/doc.pdf' })
})
@@ -418,6 +426,51 @@ describe('Share API', () => {
expect(body.media.imageTranscodes).toEqual([])
})
test('serves watermarked transcodes with hdr flags preserved when watermark enabled and completed', async () => {
vi.spyOn(shareService, 'verifyPublicAccess').mockResolvedValue(shareLinkWithWatermark)
vi.spyOn(watermarkService, 'getCompletedWatermarkMediaMap').mockResolvedValue(
new Map([
[
'file1',
watermarkMediaInfo({
videoTranscodes: [
{ key: 'files/watermarked-1080p.mp4', width: 1920, height: 1080, hdr: false },
{ key: 'files/watermarked-1080p-hdr.mp4', width: 1920, height: 1080, hdr: true },
],
}),
],
]),
)
const res = await app.request('/shares/share1/files/file1', {
method: 'GET',
headers: { 'x-share-password': 'pass' },
})
expect(res.status).toBe(200)
const body = await res.json()
expect(body.media.videoTranscodes).toEqual([
{
id: 'files/watermarked-1080p.mp4',
url: 'http://s3/presigned',
key: 'files/watermarked-1080p.mp4',
width: 1920,
height: 1080,
size: 0,
hdr: false,
},
{
id: 'files/watermarked-1080p-hdr.mp4',
url: 'http://s3/presigned',
key: 'files/watermarked-1080p-hdr.mp4',
width: 1920,
height: 1080,
size: 0,
hdr: true,
},
])
})
test('serves empty transcode arrays while watermark transcoding is in flight', async () => {
vi.spyOn(shareService, 'verifyPublicAccess').mockResolvedValue(shareLinkWithWatermark)
vi.spyOn(watermarkService, 'getCompletedWatermarkMediaMap').mockResolvedValue(new Map())
@@ -433,6 +486,148 @@ describe('Share API', () => {
expect(body.media.imageTranscodes).toEqual([])
})
test('clears HLS fields while watermark transcoding is in flight', async () => {
vi.spyOn(shareService, 'verifyPublicAccess').mockResolvedValue(shareLinkWithWatermark)
vi.spyOn(watermarkService, 'getCompletedWatermarkMediaMap').mockResolvedValue(new Map())
vi.spyOn(assetService, 'getAsset').mockResolvedValue({
...videoAsset,
media: {
...videoAsset.media,
isHls: true,
hls: {
key: 'files/hls/master.m3u8',
url: '/api/files/file1/m3u8/master.m3u8',
resolutions: [{ resolution: '1080p', width: 1920, height: 1080 }],
},
},
})
const res = await app.request('/shares/share1/files/file1', {
method: 'GET',
headers: { 'x-share-password': 'pass' },
})
expect(res.status).toBe(200)
const body = await res.json()
expect(body.media.isHls).toBe(false)
expect(body.media.hls).toBeUndefined()
})
test('serves watermarked turntable video for a 3d asset when watermark enabled and completed', async () => {
vi.spyOn(shareService, 'verifyPublicAccess').mockResolvedValue(shareLinkWithWatermark)
const asset3d = {
id: 'file1',
name: 'model.glb',
sizeByte: 1024,
fileCount: 0,
type: 'file',
status: 'processed',
proxyType: '3d',
media: {
proxyType: '3d',
videoPreview: {
url: 'http://s3/unwatermarked-turntable.mp4',
key: 'files/turntable.mp4',
},
videoTranscodes: [
{
id: 'orig-turntable',
url: 'http://s3/unwatermarked-turntable.mp4',
key: 'files/turntable.mp4',
width: 1080,
height: 1080,
},
],
original: { key: 'files/model.glb' },
},
// eslint-disable-next-line @typescript-eslint/no-explicit-any
} as any
vi.spyOn(assetService, 'getAsset').mockResolvedValue(asset3d)
vi.spyOn(watermarkService, 'getCompletedWatermarkMediaMap').mockResolvedValue(
new Map([
[
'file1',
watermarkMediaInfo({
proxyType: '3d',
videoTranscodes: [
{
key: 'files/watermarked-turntable.mp4',
width: 1080,
height: 1080,
resolution: '1080p',
},
],
videoPreview: { key: 'files/watermarked-turntable.mp4', width: 1080, height: 1080 },
}),
],
]),
)
const res = await app.request('/shares/share1/files/file1', {
method: 'GET',
headers: { 'x-share-password': 'pass' },
})
expect(res.status).toBe(200)
const body = await res.json()
expect(body.media.videoTranscodes).toEqual([
{
id: 'files/watermarked-turntable.mp4',
url: 'http://s3/presigned',
key: 'files/watermarked-turntable.mp4',
width: 1080,
height: 1080,
size: 0,
resolution: '1080p',
},
])
expect(body.media.videoPreview?.url).toBe('http://s3/presigned')
expect(body.media.videoPreview?.key).toBe('files/watermarked-turntable.mp4')
})
test('strips videoPreview url and serves empty videoTranscodes for a 3d asset while watermark is in flight', async () => {
vi.spyOn(shareService, 'verifyPublicAccess').mockResolvedValue(shareLinkWithWatermark)
const asset3d = {
id: 'file1',
name: 'model.glb',
sizeByte: 1024,
fileCount: 0,
type: 'file',
status: 'processed',
proxyType: '3d',
media: {
proxyType: '3d',
videoPreview: {
url: 'http://s3/unwatermarked-turntable.mp4',
key: 'files/turntable.mp4',
},
videoTranscodes: [
{
id: 'orig-turntable',
url: 'http://s3/unwatermarked-turntable.mp4',
key: 'files/turntable.mp4',
width: 1080,
height: 1080,
},
],
original: { key: 'files/model.glb' },
},
// eslint-disable-next-line @typescript-eslint/no-explicit-any
} as any
vi.spyOn(assetService, 'getAsset').mockResolvedValue(asset3d)
vi.spyOn(watermarkService, 'getCompletedWatermarkMediaMap').mockResolvedValue(new Map())
const res = await app.request('/shares/share1/files/file1', {
method: 'GET',
headers: { 'x-share-password': 'pass' },
})
expect(res.status).toBe(200)
const body = await res.json()
expect(body.media.videoTranscodes).toEqual([])
expect(body.media.videoPreview?.url).toBeUndefined()
})
test('leaves non-video/image media untouched when watermark enabled', async () => {
vi.spyOn(shareService, 'verifyPublicAccess').mockResolvedValue(shareLinkWithWatermark)
vi.spyOn(assetService, 'getAsset').mockResolvedValue({
@@ -821,4 +1016,81 @@ describe('Share API', () => {
expect(getDownloadUrlSpy).not.toHaveBeenCalled()
})
})
describe('GET /shares/:shareId/files/:fileId/m3u8 routes', () => {
test('returns master playlist with query password preserved in variant URLs', async () => {
vi.spyOn(shareService, 'verifyPublicAccess').mockResolvedValue({
id: 'share1',
watermarkConfigId: null,
// eslint-disable-next-line @typescript-eslint/no-explicit-any
} as any)
vi.spyOn(assetService, 'getHlsMasterPlaylist').mockResolvedValue(
'#EXTM3U\n#EXT-X-STREAM-INF:BANDWIDTH=4628000\n1080p/index.m3u8\n',
)
const res = await app.request('/shares/share1/files/file1/m3u8/master.m3u8?p=secret123')
expect(res.status).toBe(200)
expect(res.headers.get('content-type')).toContain('application/vnd.apple.mpegurl')
const text = await res.text()
expect(text).toContain('1080p/index.m3u8?p=secret123')
expect(shareService.verifyPublicAccess).toHaveBeenCalledWith('file1', 'secret123')
expect(assetService.getHlsMasterPlaylist).toHaveBeenCalledWith({
assetId: 'file1',
watermarkConfigId: null,
})
})
test('returns variant playlist with rewritten segment URLs', async () => {
vi.spyOn(shareService, 'verifyPublicAccess').mockResolvedValue({
id: 'share1',
watermarkConfigId: 'wm1',
// eslint-disable-next-line @typescript-eslint/no-explicit-any
} as any)
vi.spyOn(assetService, 'getHlsVariantPlaylist').mockResolvedValue(
'#EXTM3U\n#EXT-X-MAP:URI="http://presigned/init.mp4"\nhttp://presigned/seg0.m4s\n',
)
const res = await app.request('/shares/share1/files/file1/m3u8/1080p/index.m3u8')
expect(res.status).toBe(200)
expect(res.headers.get('content-type')).toContain('application/vnd.apple.mpegurl')
const text = await res.text()
expect(text).toContain('http://presigned/init.mp4')
expect(assetService.getHlsVariantPlaylist).toHaveBeenCalledWith({
assetId: 'file1',
resolution: '1080p',
watermarkConfigId: 'wm1',
})
})
test('returns 401 when password is invalid', async () => {
vi.spyOn(shareService, 'verifyPublicAccess').mockRejectedValue(
new ShareLinkPasswordInvalidError('Invalid password'),
)
const res = await app.request('/shares/share1/files/file1/m3u8/master.m3u8?p=wrong')
expect(res.status).toBe(401)
const body = await res.json()
expect(body.error).toBe('Unauthorized')
})
test('resolves latest version id when fileId is a symlink to a version stack', async () => {
vi.spyOn(shareService, 'verifyPublicAccess').mockResolvedValue({
id: 'share1',
watermarkConfigId: null,
// eslint-disable-next-line @typescript-eslint/no-explicit-any
} as any)
vi.spyOn(assetService, 'resolveLatestVersionId').mockResolvedValue('child-version-file-id')
vi.spyOn(assetService, 'getHlsMasterPlaylist').mockResolvedValue(
'#EXTM3U\n#EXT-X-STREAM-INF:BANDWIDTH=4628000\n1080p/index.m3u8\n',
)
const res = await app.request('/shares/share1/files/symlink1/m3u8/master.m3u8')
expect(res.status).toBe(200)
expect(assetService.resolveLatestVersionId).toHaveBeenCalledWith('symlink1')
expect(assetService.getHlsMasterPlaylist).toHaveBeenCalledWith({
assetId: 'child-version-file-id',
watermarkConfigId: null,
})
})
})
})
+305 -4
View File
@@ -52,6 +52,22 @@ vi.mock('@shumai/core/src/auditLog/auditLog', () => ({
},
}))
vi.mock('@shumai/core/src/turntable/turntable', () => ({
turntableService: {
getSettings: vi.fn(),
updateSettings: vi.fn(),
testConnection: vi.fn(),
},
}))
vi.mock('@shumai/core/src/gotenberg/gotenberg', () => ({
gotenbergService: {
getSettings: vi.fn(),
updateSettings: vi.fn(),
testConnection: vi.fn(),
},
}))
describe('team api', () => {
const app = new Hono().use('*', authMiddleware).route('/', teamRoute)
let mockCreateTeam: any // eslint-disable-line @typescript-eslint/no-explicit-any
@@ -223,18 +239,39 @@ describe('team api', () => {
expect(mockGetSettings).toHaveBeenCalledWith('t1')
})
it('GET /teams/:teamId/settings does not return emailNotification to readers', async () => {
mockGetSettings.mockResolvedValue({
semanticSearchEnabled: false,
transcode: { videoStrategy: 'best_match', hardwareAcceleration: 'off', threads: 0 },
appearance: { hideAgent: false },
})
const res = await app.request('/teams/t1/settings')
expect(res.status).toBe(200)
const data = await res.json()
expect(data.emailNotification).toBeUndefined()
expect(authzService.hasPermission).toHaveBeenCalledWith(
expect.objectContaining({
type: ResourceType.Team,
id: 't1',
permission: Permission.Read,
}),
)
})
it('PATCH /teams/:teamId/settings updates settings', async () => {
mockUpdateSettings.mockResolvedValue({ transcode: { videoStrategy: 'all' } })
mockUpdateSettings.mockResolvedValue({ transcode: { videoStrategy: 'multi' } })
const res = await app.request('/teams/t1/settings', {
method: 'PATCH',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({ key: 'transcode.videoStrategy', value: 'all' }),
body: JSON.stringify({ key: 'transcode.videoStrategy', value: 'multi' }),
})
expect(res.status).toBe(200)
const data = await res.json()
expect(data.transcode.videoStrategy).toBe('all')
expect(data.transcode.videoStrategy).toBe('multi')
expect(authzService.hasPermission).toHaveBeenCalledWith(
expect.objectContaining({
type: ResourceType.Team,
@@ -242,7 +279,30 @@ describe('team api', () => {
permission: Permission.Admin,
}),
)
expect(mockUpdateSettings).toHaveBeenCalledWith('t1', 'transcode.videoStrategy', 'all')
expect(mockUpdateSettings).toHaveBeenCalledWith('t1', 'transcode.videoStrategy', 'multi')
})
it('PATCH /teams/:teamId/settings updates transcode.videoResolutions', async () => {
mockUpdateSettings.mockResolvedValue({
transcode: { videoResolutions: ['720p', '1080p'] },
})
const res = await app.request('/teams/t1/settings', {
method: 'PATCH',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
key: 'transcode.videoResolutions',
value: ['720p', '1080p'],
}),
})
expect(res.status).toBe(200)
const data = await res.json()
expect(data.transcode.videoResolutions).toEqual(['720p', '1080p'])
expect(mockUpdateSettings).toHaveBeenCalledWith('t1', 'transcode.videoResolutions', [
'720p',
'1080p',
])
})
it('PATCH /teams/:teamId/settings updates transcode.hardwareAcceleration', async () => {
@@ -260,6 +320,44 @@ describe('team api', () => {
expect(mockUpdateSettings).toHaveBeenCalledWith('t1', 'transcode.hardwareAcceleration', 'auto')
})
it('PATCH /teams/:teamId/settings updates transcode.threads', async () => {
mockUpdateSettings.mockResolvedValue({ transcode: { threads: 8 } })
const res = await app.request('/teams/t1/settings', {
method: 'PATCH',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({ key: 'transcode.threads', value: 8 }),
})
expect(res.status).toBe(200)
const data = await res.json()
expect(data.transcode.threads).toBe(8)
expect(mockUpdateSettings).toHaveBeenCalledWith('t1', 'transcode.threads', 8)
})
it('PATCH /teams/:teamId/settings rejects invalid transcode.threads', async () => {
const resNegative = await app.request('/teams/t1/settings', {
method: 'PATCH',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({ key: 'transcode.threads', value: -1 }),
})
expect(resNegative.status).toBe(400)
const resTooLarge = await app.request('/teams/t1/settings', {
method: 'PATCH',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({ key: 'transcode.threads', value: 33 }),
})
expect(resTooLarge.status).toBe(400)
const resFloat = await app.request('/teams/t1/settings', {
method: 'PATCH',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({ key: 'transcode.threads', value: 4.5 }),
})
expect(resFloat.status).toBe(400)
})
it('PATCH /teams/:teamId/settings updates appearance.hideAgent', async () => {
mockUpdateSettings.mockResolvedValue({ appearance: { hideAgent: true } })
@@ -556,5 +654,208 @@ describe('team api', () => {
}),
)
})
it('GET /teams/:teamId/turntable-settings returns turntable settings', async () => {
const { turntableService } = await import('@shumai/core/src/turntable/turntable')
vi.mocked(turntableService.getSettings).mockResolvedValue({
url: 'http://localhost:3001',
username: 'user',
hasPassword: true,
isEnvConfigured: false,
})
const res = await app.request('/teams/t1/turntable-settings')
expect(res.status).toBe(200)
const data = await res.json()
expect(data).toEqual({
url: 'http://localhost:3001',
username: 'user',
hasPassword: true,
isEnvConfigured: false,
})
expect(authzService.hasPermission).toHaveBeenCalledWith(
expect.objectContaining({
permission: Permission.Read,
type: ResourceType.Team,
id: 't1',
}),
)
})
it('PUT /teams/:teamId/turntable-settings updates settings when not env configured', async () => {
const { turntableService } = await import('@shumai/core/src/turntable/turntable')
vi.mocked(turntableService.updateSettings).mockResolvedValue({
url: 'http://custom:3001',
username: 'newuser',
hasPassword: true,
isEnvConfigured: false,
})
const res = await app.request('/teams/t1/turntable-settings', {
method: 'PUT',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
url: 'http://custom:3001',
username: 'newuser',
password: 'secretpassword',
}),
})
expect(res.status).toBe(200)
const data = await res.json()
expect(data.url).toBe('http://custom:3001')
expect(authzService.hasPermission).toHaveBeenCalledWith(
expect.objectContaining({
permission: Permission.Admin,
type: ResourceType.Team,
id: 't1',
}),
)
})
it('PUT /teams/:teamId/turntable-settings returns 400 when env configured', async () => {
const { turntableService } = await import('@shumai/core/src/turntable/turntable')
vi.mocked(turntableService.updateSettings).mockRejectedValue(
new Error(
'Turntable settings are configured via environment variables and cannot be edited',
),
)
const res = await app.request('/teams/t1/turntable-settings', {
method: 'PUT',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
url: 'http://custom:3001',
}),
})
expect(res.status).toBe(400)
const data = await res.json()
expect(data.message).toContain('configured via environment variables')
})
it('POST /teams/:teamId/turntable-settings/test tests connection', async () => {
const { turntableService } = await import('@shumai/core/src/turntable/turntable')
vi.mocked(turntableService.testConnection).mockResolvedValue({
ok: true,
version: '1.0.0',
blender: '5.2.0',
})
const res = await app.request('/teams/t1/turntable-settings/test', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
url: 'http://custom:3001',
}),
})
expect(res.status).toBe(200)
const data = await res.json()
expect(data.ok).toBe(true)
expect(data.blender).toBe('5.2.0')
})
it('GET /teams/:teamId/gotenberg-settings returns gotenberg settings', async () => {
const { gotenbergService } = await import('@shumai/core/src/gotenberg/gotenberg')
vi.mocked(gotenbergService.getSettings).mockResolvedValue({
url: 'http://localhost:3000',
username: 'user',
hasPassword: true,
isEnvConfigured: false,
})
const res = await app.request('/teams/t1/gotenberg-settings')
expect(res.status).toBe(200)
const data = await res.json()
expect(data).toEqual({
url: 'http://localhost:3000',
username: 'user',
hasPassword: true,
isEnvConfigured: false,
})
expect(authzService.hasPermission).toHaveBeenCalledWith(
expect.objectContaining({
permission: Permission.Read,
type: ResourceType.Team,
id: 't1',
}),
)
})
it('PUT /teams/:teamId/gotenberg-settings updates settings when not env configured', async () => {
const { gotenbergService } = await import('@shumai/core/src/gotenberg/gotenberg')
vi.mocked(gotenbergService.updateSettings).mockResolvedValue({
url: 'http://custom:3000',
username: 'newuser',
hasPassword: true,
isEnvConfigured: false,
})
const res = await app.request('/teams/t1/gotenberg-settings', {
method: 'PUT',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
url: 'http://custom:3000',
username: 'newuser',
password: 'secretpassword',
}),
})
expect(res.status).toBe(200)
const data = await res.json()
expect(data.url).toBe('http://custom:3000')
expect(authzService.hasPermission).toHaveBeenCalledWith(
expect.objectContaining({
permission: Permission.Admin,
type: ResourceType.Team,
id: 't1',
}),
)
})
it('PUT /teams/:teamId/gotenberg-settings returns 400 when env configured', async () => {
const { gotenbergService } = await import('@shumai/core/src/gotenberg/gotenberg')
vi.mocked(gotenbergService.updateSettings).mockRejectedValue(
new Error(
'Gotenberg settings are configured via environment variables and cannot be edited',
),
)
const res = await app.request('/teams/t1/gotenberg-settings', {
method: 'PUT',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
url: 'http://custom:3000',
}),
})
expect(res.status).toBe(400)
const data = await res.json()
expect(data.message).toContain('configured via environment variables')
})
it('POST /teams/:teamId/gotenberg-settings/test tests connection', async () => {
const { gotenbergService } = await import('@shumai/core/src/gotenberg/gotenberg')
vi.mocked(gotenbergService.testConnection).mockResolvedValue({
ok: true,
version: '8.3.0',
})
const res = await app.request('/teams/t1/gotenberg-settings/test', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
url: 'http://custom:3000',
}),
})
expect(res.status).toBe(200)
const data = await res.json()
expect(data.ok).toBe(true)
expect(data.version).toBe('8.3.0')
})
})
})
+118
View File
@@ -10,6 +10,8 @@ import {
updateTeamSettingsRequestSchema,
listMembersQuerySchema,
updateSandboxSettingsRequestSchema,
updateTurntableSettingsRequestSchema,
updateGotenbergSettingsRequestSchema,
updateMeRequestSchema,
updateTeamMemberRoleRequestSchema,
createApiTokenRequestSchema,
@@ -23,6 +25,8 @@ import {
import { apiTokenService } from '@shumai/core/src/user/api-token'
import { aiUsageService } from '@shumai/core/src/ai-usage/ai-usage'
import { auditLogService } from '@shumai/core/src/auditLog/auditLog'
import { turntableService } from '@shumai/core/src/turntable/turntable'
import { gotenbergService } from '@shumai/core/src/gotenberg/gotenberg'
import { NotificationType } from '@shumai/db'
import type { Prisma } from '@shumai/db'
@@ -265,6 +269,120 @@ const route = new Hono<{ Variables: { user: User } }>()
return c.json(sandbox)
},
)
.get('/teams/:teamId/turntable-settings', async (c) => {
const user = c.get('user')
const teamId = c.req.param('teamId')
await authzService.hasPermission({
user,
permission: Permission.Read,
type: ResourceType.Team,
id: teamId,
})
const settings = await turntableService.getSettings(teamId)
return c.json(settings)
})
.put(
'/teams/:teamId/turntable-settings',
zValidator('json', updateTurntableSettingsRequestSchema),
async (c) => {
const user = c.get('user')
const teamId = c.req.param('teamId')
const req = c.req.valid('json')
await authzService.hasPermission({
user,
permission: Permission.Admin,
type: ResourceType.Team,
id: teamId,
})
try {
const settings = await turntableService.updateSettings(teamId, req)
return c.json(settings)
} catch (err) {
const message = err instanceof Error ? err.message : String(err)
return c.json({ message }, 400)
}
},
)
.post(
'/teams/:teamId/turntable-settings/test',
zValidator('json', updateTurntableSettingsRequestSchema.partial()),
async (c) => {
const user = c.get('user')
const teamId = c.req.param('teamId')
const req = c.req.valid('json')
await authzService.hasPermission({
user,
permission: Permission.Admin,
type: ResourceType.Team,
id: teamId,
})
const res = await turntableService.testConnection(req, teamId)
return c.json(res)
},
)
.get('/teams/:teamId/gotenberg-settings', async (c) => {
const user = c.get('user')
const teamId = c.req.param('teamId')
await authzService.hasPermission({
user,
permission: Permission.Read,
type: ResourceType.Team,
id: teamId,
})
const settings = await gotenbergService.getSettings(teamId)
return c.json(settings)
})
.put(
'/teams/:teamId/gotenberg-settings',
zValidator('json', updateGotenbergSettingsRequestSchema),
async (c) => {
const user = c.get('user')
const teamId = c.req.param('teamId')
const req = c.req.valid('json')
await authzService.hasPermission({
user,
permission: Permission.Admin,
type: ResourceType.Team,
id: teamId,
})
try {
const settings = await gotenbergService.updateSettings(teamId, req)
return c.json(settings)
} catch (err) {
const message = err instanceof Error ? err.message : String(err)
return c.json({ message }, 400)
}
},
)
.post(
'/teams/:teamId/gotenberg-settings/test',
zValidator('json', updateGotenbergSettingsRequestSchema.partial()),
async (c) => {
const user = c.get('user')
const teamId = c.req.param('teamId')
const req = c.req.valid('json')
await authzService.hasPermission({
user,
permission: Permission.Admin,
type: ResourceType.Team,
id: teamId,
})
const res = await gotenbergService.testConnection(req, teamId)
return c.json(res)
},
)
.get('/teams/:teamId/user-metadata', async (c) => {
const user = c.get('user')
const teamId = c.req.param('teamId')
+3 -1
View File
@@ -105,15 +105,17 @@ const route = new Hono<{ Variables: { user: User } }>()
await uploadService.confirmFileUpload(user.id, taskId, req)
if (!req.errorMessage) {
const context = await assetService.getAssetContext(req.fileId)
await notificationService.create({
type: NotificationType.successful_file_uploaded,
teamId: teamId,
projectId: context.projectId,
creatorId: user.id,
assetId: req.fileId,
taskId: taskId,
})
const context = await assetService.getAssetContext(req.fileId)
await auditLogService.logAction({
action: AuditAction.file_create,
teamId: teamId,
+7 -1
View File
@@ -1,6 +1,6 @@
{
"name": "@shumai/core",
"version": "0.4.6",
"version": "0.5.3",
"private": true,
"type": "module",
"main": "./src/index.ts",
@@ -29,10 +29,13 @@
"@aws-sdk/s3-request-presigner": "3.1121.0",
"@earendil-works/pi-agent-core": "0.82.1",
"@earendil-works/pi-ai": "0.82.1",
"@google/genai": "2.28.0",
"@modelcontextprotocol/client": "2.0.0",
"@shumai/db": "workspace:*",
"@shumai/dtos": "workspace:*",
"@sinclair/typebox": "0.34.52",
"@temporalio/activity": "1.24.0",
"@temporalio/workflow": "1.24.0",
"adm-zip": "0.6.0",
"ai": "7.0.84",
"better-auth": "1.6.11",
@@ -40,9 +43,11 @@
"date-fns": "4.4.0",
"dotenv": "17.4.2",
"dotenv-expand": "13.0.0",
"exiftool-vendored": "38.1.0",
"extract-zip": "2.0.1",
"hono": "4.13.5",
"jittered-fractional-indexing": "1.0.1",
"nodemailer": "10.0.10",
"pdfkit": "0.20.1",
"pino": "10.3.1",
"sharp": "0.35.4",
@@ -56,6 +61,7 @@
"@modelcontextprotocol/server": "2.0.0",
"@types/adm-zip": "0.5.8",
"@types/extract-zip": "2.0.3",
"@types/nodemailer": "8.0.2",
"@types/pdfkit": "0.17.6"
}
}
+167
View File
@@ -976,4 +976,171 @@ describe('AgentService', () => {
expect(ids).toHaveLength(0)
})
})
describe('ensureEmbeddingAgent', () => {
test('creates embedding agent when ENABLE_EMBEDDING_AGENT is true', async () => {
const db = prisma
const svc = new AgentService()
const { team } = await setupTestData(db)
process.env.ENABLE_EMBEDDING_AGENT = 'true'
try {
await svc.ensureEmbeddingAgentForTeam(team.id)
const agent = await db.agent.findFirst({
where: { teamId: team.id, type: 'embedding' },
include: { user: true, skills: { include: { skill: true } }, mcpServers: true },
})
expect(agent).toBeDefined()
const info = await svc.toAgentInfo(agent!)
expect(info.name).toBe('embedding-agent')
expect(info.type).toBe('embedding')
expect(info.enabled).toBe(true)
expect(info.avatarPreset).toBe('avatar-1')
expect(info.avatar).toContain('avatar-1.webp')
expect(info.readOnly).toBe(true)
// Ensure idempotency
await svc.ensureEmbeddingAgentForTeam(team.id)
const count = await db.agent.count({
where: { teamId: team.id, type: 'embedding' },
})
expect(count).toBe(1)
} finally {
delete process.env.ENABLE_EMBEDDING_AGENT
}
})
test('enables existing embedding agent when ENABLE_EMBEDDING_AGENT is true', async () => {
const db = prisma
const svc = new AgentService()
const { team } = await setupTestData(db)
const bot = await db.user.create({
data: { name: 'Existing Embedding', email: 'existing-emb@shumai.ai', type: 'agent' },
})
await db.agent.create({
data: {
id: bot.id,
teamId: team.id,
type: 'embedding',
enabled: false,
config: { provider: '', model: '' },
},
})
process.env.ENABLE_EMBEDDING_AGENT = 'true'
try {
await svc.ensureEmbeddingAgentForTeam(team.id)
const agent = await db.agent.findFirst({
where: { teamId: team.id, type: 'embedding' },
include: { user: true, skills: { include: { skill: true } }, mcpServers: true },
})
expect(agent?.id).toBe(bot.id)
expect(agent?.enabled).toBe(true)
const info = await svc.toAgentInfo(agent!)
expect(info.readOnly).toBe(true)
} finally {
delete process.env.ENABLE_EMBEDDING_AGENT
}
})
test('does nothing when ENABLE_EMBEDDING_AGENT is not set', async () => {
const db = prisma
const svc = new AgentService()
const team = await db.team.create({ data: { name: 'Team No Agent' } })
delete process.env.ENABLE_EMBEDDING_AGENT
await svc.ensureEmbeddingAgentForTeam(team.id)
const agent = await db.agent.findFirst({
where: { teamId: team.id, type: 'embedding' },
})
expect(agent).toBeNull()
})
test('does not create embedding agent when ENABLE_EMBEDDING_AGENT is false and agent does not exist', async () => {
const db = prisma
const svc = new AgentService()
const { team } = await setupTestData(db)
process.env.ENABLE_EMBEDDING_AGENT = 'false'
try {
await svc.ensureEmbeddingAgentForTeam(team.id)
const agent = await db.agent.findFirst({
where: { teamId: team.id, type: 'embedding' },
})
expect(agent).toBeNull()
} finally {
delete process.env.ENABLE_EMBEDDING_AGENT
}
})
test('disables existing embedding agent and marks readOnly: true when ENABLE_EMBEDDING_AGENT is false', async () => {
const db = prisma
const svc = new AgentService()
const { team } = await setupTestData(db)
const bot = await db.user.create({
data: { name: 'Existing Embedding', email: 'existing-emb-false@shumai.ai', type: 'agent' },
})
await db.agent.create({
data: {
id: bot.id,
teamId: team.id,
type: 'embedding',
enabled: true,
config: { provider: '', model: '' },
},
})
process.env.ENABLE_EMBEDDING_AGENT = 'false'
try {
await svc.ensureEmbeddingAgentForTeam(team.id)
const agent = await db.agent.findFirst({
where: { teamId: team.id, type: 'embedding' },
include: { user: true, skills: { include: { skill: true } }, mcpServers: true },
})
expect(agent?.id).toBe(bot.id)
expect(agent?.enabled).toBe(false)
const info = await svc.toAgentInfo(agent!)
expect(info.readOnly).toBe(true)
expect(info.enabled).toBe(false)
} finally {
delete process.env.ENABLE_EMBEDDING_AGENT
}
})
test('ensureSystemEmbeddingAgents updates all teams when ENABLE_EMBEDDING_AGENT is false', async () => {
const db = prisma
const svc = new AgentService()
const { team: team1 } = await setupTestData(db)
const team2 = await db.team.create({ data: { name: 'Team 2' } })
// Create enabled embedding agent on team1
const bot1 = await db.user.create({
data: { name: 'Bot 1', email: 'b1-false@shumai.ai', type: 'agent' },
})
await db.agent.create({
data: {
id: bot1.id,
teamId: team1.id,
type: 'embedding',
enabled: true,
config: { provider: '', model: '' },
},
})
process.env.ENABLE_EMBEDDING_AGENT = 'false'
try {
await svc.ensureSystemEmbeddingAgents()
const a1 = await db.agent.findFirst({ where: { teamId: team1.id, type: 'embedding' } })
const a2 = await db.agent.findFirst({ where: { teamId: team2.id, type: 'embedding' } })
expect(a1?.enabled).toBe(false)
expect(a2).toBeNull()
} finally {
delete process.env.ENABLE_EMBEDDING_AGENT
}
})
})
})
+63 -3
View File
@@ -43,8 +43,8 @@ interface AgentInfoSource {
config: unknown
soul: string | null
user: { name: string; image: string | null }
skills: Array<{ id: string; skillId: string; skill: { id: string; name: string } }>
mcpServers: Array<{ mcpServerId: string }>
skills?: Array<{ id: string; skillId: string; skill: { id: string; name: string } }>
mcpServers?: Array<{ mcpServerId: string }>
}
function extractS3Key(urlOrKey?: string | null): string | null | undefined {
@@ -462,6 +462,11 @@ export class AgentService {
*/
async toAgentInfo(agent: AgentInfoSource): Promise<AgentInfo> {
const config = agent.config as unknown as PrismaJson.AgentConfig
const isManagedByEnv =
process.env.ENABLE_EMBEDDING_AGENT === 'true' ||
process.env.ENABLE_EMBEDDING_AGENT === 'false'
const isReadOnly = agent.type === 'embedding' && isManagedByEnv
return {
id: agent.id,
name: agent.user.name,
@@ -475,13 +480,14 @@ export class AgentService {
thinkingLevel: config.thinkingLevel || 'off',
systemPrompt: config.systemPrompt,
soul: agent.soul || undefined,
skills: agent.skills.map((s) => ({
skills: (agent.skills || []).map((s) => ({
id: s.id,
skillId: s.skillId,
skill: s.skill,
})),
mcpServerIds: (agent.mcpServers || []).map((m) => m.mcpServerId),
deniedTools: config.deniedTools || [],
readOnly: isReadOnly,
}
}
@@ -718,6 +724,60 @@ export class AgentService {
params,
)
}
async ensureSystemEmbeddingAgents(): Promise<void> {
const envVal = process.env.ENABLE_EMBEDDING_AGENT
if (envVal !== 'true' && envVal !== 'false') {
return
}
const teams = await this.prismaClient.team.findMany({
select: { id: true },
})
for (const team of teams) {
await this.ensureEmbeddingAgentForTeam(team.id)
}
}
async ensureEmbeddingAgentForTeam(teamId: string): Promise<void> {
const envVal = process.env.ENABLE_EMBEDDING_AGENT
if (envVal !== 'true' && envVal !== 'false') {
return
}
const existing = await this.prismaClient.agent.findFirst({
where: {
teamId,
type: 'embedding',
},
})
if (envVal === 'true') {
if (!existing) {
await this.createAgent({
teamId,
name: 'embedding-agent',
type: 'embedding',
avatar: 'avatar-1',
thinkingLevel: 'off',
enabled: true,
})
} else if (!existing.enabled) {
await this.prismaClient.agent.update({
where: { id: existing.id },
data: { enabled: true },
})
}
} else if (envVal === 'false') {
if (existing && existing.enabled) {
await this.prismaClient.agent.update({
where: { id: existing.id },
data: { enabled: false },
})
}
}
}
}
export const agentService = new AgentService()
+533 -2
View File
@@ -3,18 +3,20 @@ import { prisma } from '@shumai/db'
import { setupTestDbHooks } from '@shumai/db/test'
import { HTTPException } from 'hono/http-exception'
import { AssetType, AssetStatus, Prisma } from '@shumai/db'
import { AssetType, AssetStatus, Prisma, WorkflowTaskType, WorkflowTaskStatus } from '@shumai/db'
import { s3Service } from '@shumai/core/src/s3/s3'
import { uploadService } from '@shumai/core/src/upload/upload'
import { metadataService } from '@shumai/core/src/metadata/metadata'
import { AssetService } from './asset'
import { ensureJpegInStorage } from '@shumai/core/src/s3/preview-converter'
import { workflowService } from '@shumai/workflow-core'
vi.mock('@shumai/core/src/s3/s3', () => ({
s3Service: {
presign: vi.fn().mockResolvedValue('http://mock-s3-url'),
putObject: vi.fn().mockResolvedValue(undefined),
getObject: vi.fn().mockResolvedValue({ buffer: Buffer.from('') }),
deleteObject: vi.fn().mockResolvedValue(1),
deletePrefix: vi.fn().mockResolvedValue(1),
copyObject: vi.fn().mockResolvedValue(undefined),
@@ -1929,6 +1931,36 @@ describe('AssetService', () => {
expect(info.preview?.duration).toBeUndefined()
})
it('toPreviewInfo handles 3d asset with proxyType 3d and omits duration', async () => {
const { project, user } = await setupBasicAssets()
const model = await prisma.asset.create({
data: {
name: 'model.glb',
type: AssetType.file,
projectId: project.id,
creatorId: user.id,
sizeByte: 8000,
mediaType: 'model/gltf-binary',
status: 'processed',
media: {
proxyType: '3d',
poster: { key: 'poster-key' },
sprite: { key: 'sprite-key', frames: 100, tileX: 10, tileY: 10 },
metadata: { originalWidth: 1080, originalHeight: 1080, duration: 4 },
// eslint-disable-next-line @typescript-eslint/no-explicit-any
} as any,
},
})
const info = await assetService.getAsset({ assetId: model.id })
expect(info.preview).toBeDefined()
expect(info.preview?.proxyType).toBe('3d')
expect(info.preview?.duration).toBeUndefined()
expect(info.preview?.spriteUrl).toBeDefined()
})
it('converts webp thumbnail to jpeg on demand when hasJpegPreview is false', async () => {
vi.mocked(ensureJpegInStorage).mockClear()
const { project, user } = await setupBasicAssets()
@@ -2529,6 +2561,209 @@ describe('AssetService', () => {
expect(updatedEntry?.sessionId).toBe(session.id)
})
})
describe('listChildren with showDeleted', () => {
const createImageFile = async (params: {
name: string
projectId: string
parentId: string
creatorId: string
thumbnailKey: string
duration?: number
sortIndex?: string
}) => {
return prisma.asset.create({
data: {
name: params.name,
type: AssetType.file,
status: AssetStatus.processed,
isDeleted: true,
projectId: params.projectId,
parentId: params.parentId,
creatorId: params.creatorId,
sizeByte: 100,
sortIndex: params.sortIndex,
mediaType: 'image/png',
media: {
duration: 0,
filesize: 0,
frames: 0,
imageTranscodes: [],
videoTranscodes: [],
finishedAt: new Date(0).toISOString(),
metadata:
params.duration === undefined
? null
: {
originalHeight: 0,
originalWidth: 0,
hasAudio: false,
duration: params.duration,
bitRate: 0,
frameRate: 0,
totalFrames: 0,
startTimecode: '00:00:00:00',
format: null,
},
original: null,
proxyType: 'image',
thumbnail: {
key: params.thumbnailKey,
width: 0,
height: 0,
quality: 0,
format: 'png',
},
},
},
})
}
it('returns previews for children of a folder trashed via deleteAssets', async () => {
const { project, user, assets } = await setupBasicAssets()
const folder = await prisma.asset.create({
data: {
name: 'Trashed Folder',
type: AssetType.folder,
status: AssetStatus.uploaded,
projectId: project.id,
parentId: assets.root.id,
creatorId: user.id,
},
})
await createImageFile({
name: 'inside-1.png',
projectId: project.id,
parentId: folder.id,
creatorId: user.id,
thumbnailKey: 'thumb-1',
})
await createImageFile({
name: 'inside-2.png',
projectId: project.id,
parentId: folder.id,
creatorId: user.id,
thumbnailKey: 'thumb-2',
})
await assetService.deleteAssets([folder.id])
const result = await assetService.listChildren({
projectId: project.id,
showDeleted: true,
assetType: 'folder',
})
expect(result.data).toHaveLength(1)
expect(result.data[0].id).toBe(folder.id)
expect(result.data[0].deletedAt).not.toBeNull()
expect(result.data[0].latestChildren).toHaveLength(2)
expect(result.data[0].latestChildren?.every((c) => c.preview?.thumbnailUrl)).toBe(true)
})
it('resolves the latest deleted version for a trashed version stack child', async () => {
const { project, user, assets } = await setupBasicAssets()
const folder = await prisma.asset.create({
data: {
name: 'Trashed Folder With Stack',
type: AssetType.folder,
status: AssetStatus.uploaded,
projectId: project.id,
parentId: assets.root.id,
creatorId: user.id,
},
})
const stack = await prisma.asset.create({
data: {
name: 'clip',
type: AssetType.version_stack,
status: AssetStatus.processed,
projectId: project.id,
parentId: folder.id,
creatorId: user.id,
},
})
await createImageFile({
name: 'clip-v1.png',
projectId: project.id,
parentId: stack.id,
creatorId: user.id,
thumbnailKey: 'v1-thumb',
duration: 11,
sortIndex: 'a1',
})
await createImageFile({
name: 'clip-v2.png',
projectId: project.id,
parentId: stack.id,
creatorId: user.id,
thumbnailKey: 'v2-thumb',
duration: 42,
sortIndex: 'a0',
})
await assetService.deleteAssets([folder.id])
const result = await assetService.listChildren({
projectId: project.id,
showDeleted: true,
assetType: 'folder',
})
const previews = result.data[0].latestChildren ?? []
expect(previews).toHaveLength(1)
// The latest version (lowest sortIndex) must be the one used for the preview
expect(previews[0].preview?.duration).toBe(42)
expect(previews[0].preview?.thumbnailUrl).toBeDefined()
})
it('still hides soft-deleted children when browsing a folder normally', async () => {
const { project, user, assets } = await setupBasicAssets()
const folder = await prisma.asset.create({
data: {
name: 'Active Folder',
type: AssetType.folder,
status: AssetStatus.uploaded,
projectId: project.id,
parentId: assets.root.id,
creatorId: user.id,
},
})
await prisma.asset.create({
data: {
name: 'active.png',
type: AssetType.file,
status: AssetStatus.processed,
isDeleted: false,
projectId: project.id,
parentId: folder.id,
creatorId: user.id,
},
})
await prisma.asset.create({
data: {
name: 'deleted.png',
type: AssetType.file,
status: AssetStatus.processed,
isDeleted: true,
deletedAt: new Date(),
projectId: project.id,
parentId: folder.id,
creatorId: user.id,
},
})
const folderInfo = await assetService.getAsset({ assetId: folder.id })
expect(folderInfo.latestChildren).toHaveLength(1)
})
})
})
describe('AssetService — natural sort by name', () => {
@@ -3004,6 +3239,65 @@ describe('AssetService — natural sort by name', () => {
await expect(assetService.emptyTrash(project.id)).resolves.not.toThrow()
})
it('should purge trashed assets whose status was processed, cancel their in-flight tasks, and delete WorkflowTasks', async () => {
const cancelSpy = vi.spyOn(workflowService, 'cancel').mockResolvedValue(undefined)
const team = await prisma.team.create({ data: { name: 'Purge Task Team' } })
const project = await prisma.project.create({
data: { name: 'Purge Task Project', teamId: team.id },
})
const rootFolder = await prisma.asset.create({
data: {
name: 'root',
type: AssetType.folder,
projectId: project.id,
status: AssetStatus.uploaded,
},
})
const user = await prisma.user.create({
data: { name: 'Test User Purge', email: `test-purge-${Date.now()}@example.com` },
})
// Asset was deleted, but status was set to 'processed' by background activity
const file = await prisma.asset.create({
data: {
name: 'skewed-status.mp4',
type: AssetType.file,
status: AssetStatus.processed,
isDeleted: true,
deletedAt: new Date(),
project: { connect: { id: project.id } },
parent: { connect: { id: rootFolder.id } },
creator: { connect: { id: user.id } },
storageKey: { create: { key: 'files/skewed-status/raw.mp4' } },
},
})
// In-flight task for this asset
const task = await prisma.workflowTask.create({
data: {
assetId: file.id,
type: WorkflowTaskType.transcode_video,
status: WorkflowTaskStatus.processing,
},
})
await assetService.emptyTrash(project.id)
// Asset should be deleted from DB
const assetInDb = await prisma.asset.findUnique({ where: { id: file.id } })
expect(assetInDb).toBeNull()
// Task should be cancelled
expect(cancelSpy).toHaveBeenCalledWith(task.id)
// WorkflowTask should be deleted from DB
const taskInDb = await prisma.workflowTask.findUnique({ where: { id: task.id } })
expect(taskInDb).toBeNull()
cancelSpy.mockRestore()
})
})
describe('expireTrashedAssets', () => {
@@ -3116,7 +3410,53 @@ describe('AssetService — natural sort by name', () => {
'files/01TESTULID000000000000000/test-720p.mp4',
'GET',
true,
'test.mp4',
'test-720p.mp4',
)
})
it('downloading video proxy for test.mov should use proxy filename with mp4 extension instead of test.mov', async () => {
const asset = await prisma.asset.create({
data: {
name: 'test.mov',
type: AssetType.file,
status: AssetStatus.uploaded,
storageKey: {
create: { key: 'files/01TESTULID000000000000000/test.mov' },
},
},
})
await assetService.getDownloadUrl(asset.id, 'files/01TESTULID000000000000000/test-1080p.mp4')
expect(s3Service.presign).toHaveBeenCalledWith(
expect.any(String),
'files/01TESTULID000000000000000/test-1080p.mp4',
'GET',
true,
'test-1080p.mp4',
)
})
it('downloading proxy for a renamed asset should preserve new stem with proxy resolution and extension', async () => {
const asset = await prisma.asset.create({
data: {
name: 'awesome.mov',
type: AssetType.file,
status: AssetStatus.uploaded,
storageKey: {
create: { key: 'files/01TESTULID000000000000000/test.mov' },
},
},
})
await assetService.getDownloadUrl(asset.id, 'files/01TESTULID000000000000000/test-1080p.mp4')
expect(s3Service.presign).toHaveBeenCalledWith(
expect.any(String),
'files/01TESTULID000000000000000/test-1080p.mp4',
'GET',
true,
'awesome-1080p.mp4',
)
})
@@ -4858,4 +5198,195 @@ describe('AssetService — natural sort by name', () => {
expect(stackInfo?.commentsCount).toBe(1)
})
})
describe('restoreAssets with in-flight transcoding', () => {
it('should restore asset with status "processing" if a transcode task is still active', async () => {
const team = await prisma.team.create({ data: { name: 'Restore Transcode Team' } })
const project = await prisma.project.create({
data: { name: 'Restore Transcode Project', teamId: team.id },
})
const rootFolder = await prisma.asset.create({
data: {
name: 'root',
type: AssetType.folder,
projectId: project.id,
status: AssetStatus.uploaded,
},
})
const file = await prisma.asset.create({
data: {
name: 'transcoding-video.mp4',
type: AssetType.file,
projectId: project.id,
parentId: rootFolder.id,
status: AssetStatus.trashed,
isDeleted: true,
deletedAt: new Date(),
},
})
// Create an active transcode task
await prisma.workflowTask.create({
data: {
assetId: file.id,
type: WorkflowTaskType.transcode_video,
status: WorkflowTaskStatus.processing,
},
})
await assetService.restoreAssets([file.id])
const restored = await prisma.asset.findUnique({ where: { id: file.id } })
expect(restored?.isDeleted).toBe(false)
expect(restored?.status).toBe(AssetStatus.processing)
})
it('should restore asset with status "processed" if no active transcode task exists', async () => {
const team = await prisma.team.create({ data: { name: 'Restore Completed Team' } })
const project = await prisma.project.create({
data: { name: 'Restore Completed Project', teamId: team.id },
})
const rootFolder = await prisma.asset.create({
data: {
name: 'root',
type: AssetType.folder,
projectId: project.id,
status: AssetStatus.uploaded,
},
})
const file = await prisma.asset.create({
data: {
name: 'completed-video.mp4',
type: AssetType.file,
projectId: project.id,
parentId: rootFolder.id,
status: AssetStatus.trashed,
isDeleted: true,
deletedAt: new Date(),
},
})
// Completed transcode task
await prisma.workflowTask.create({
data: {
assetId: file.id,
type: WorkflowTaskType.transcode_video,
status: WorkflowTaskStatus.completed,
},
})
await assetService.restoreAssets([file.id])
const restored = await prisma.asset.findUnique({ where: { id: file.id } })
expect(restored?.isDeleted).toBe(false)
expect(restored?.status).toBe(AssetStatus.processed)
})
})
describe('HLS playlists', () => {
const setupHlsAssets = async () => {
const team = await prisma.team.create({ data: { name: 'hls-team-' + Date.now() } })
const project = await prisma.project.create({
data: { name: 'hls-proj-' + Date.now(), teamId: team.id },
})
const rootFolder = await prisma.asset.create({
data: {
name: 'root',
type: AssetType.folder,
projectId: project.id,
status: AssetStatus.processed,
},
})
return { team, project, rootFolder }
}
it('getHlsMasterPlaylist throws 404 if asset is not HLS', async () => {
const { project, rootFolder } = await setupHlsAssets()
const file = await prisma.asset.create({
data: {
name: 'regular.mp4',
type: AssetType.file,
projectId: project.id,
parentId: rootFolder.id,
status: AssetStatus.processed,
media: {
isHls: false,
} as PrismaJson.MediaInfo,
},
})
await expect(assetService.getHlsMasterPlaylist({ assetId: file.id })).rejects.toThrow(
'HLS playlist not found',
)
})
it('getHlsMasterPlaylist returns playlist string when asset is HLS', async () => {
const { project, rootFolder } = await setupHlsAssets()
const file = await prisma.asset.create({
data: {
name: 'hls-video.mp4',
type: AssetType.file,
projectId: project.id,
parentId: rootFolder.id,
status: AssetStatus.processed,
media: {
isHls: true,
hls: {
key: 'files/asset/hls/master.m3u8',
resolutions: [
{ resolution: '1080p', width: 1920, height: 1080 },
{ resolution: '720p', width: 1280, height: 720 },
],
},
} as PrismaJson.MediaInfo,
},
})
const masterContent = '#EXTM3U\n#EXT-X-STREAM-INF:BANDWIDTH=4628000\n1080p/index.m3u8\n'
vi.mocked(s3Service.getObject).mockResolvedValueOnce({
buffer: Buffer.from(masterContent),
} as unknown as Awaited<ReturnType<typeof s3Service.getObject>>)
const res = await assetService.getHlsMasterPlaylist({ assetId: file.id })
expect(res).toBe(masterContent)
expect(s3Service.getObject).toHaveBeenCalledWith(
expect.any(String),
'files/asset/hls/master.m3u8',
)
})
it('getHlsVariantPlaylist fetches and rewrites segment URLs', async () => {
const { project, rootFolder } = await setupHlsAssets()
const file = await prisma.asset.create({
data: {
name: 'hls-video.mp4',
type: AssetType.file,
projectId: project.id,
parentId: rootFolder.id,
status: AssetStatus.processed,
media: {
isHls: true,
hls: {
key: 'files/asset/hls/master.m3u8',
resolutions: [{ resolution: '1080p', width: 1920, height: 1080 }],
},
} as PrismaJson.MediaInfo,
},
})
const sampleVariant = '#EXTM3U\n#EXT-X-MAP:URI="init.mp4"\nsegment_000.m4s\n'
vi.mocked(s3Service.getObject).mockResolvedValueOnce({
buffer: Buffer.from(sampleVariant),
} as unknown as Awaited<ReturnType<typeof s3Service.getObject>>)
vi.mocked(s3Service.presign).mockResolvedValue('http://mock-s3-url')
const variant = await assetService.getHlsVariantPlaylist({
assetId: file.id,
resolution: '1080p',
})
expect(variant).toBeDefined()
expect(variant).toContain('#EXT-X-MAP:URI="http://mock-s3-url"')
expect(variant).toContain('http://mock-s3-url')
})
})
})
+269 -25
View File
@@ -1,4 +1,4 @@
import { prisma } from '@shumai/db'
import { cancelWorkflowTask, prisma } from '@shumai/db'
import {
AncestorFolder,
AssetInfo,
@@ -20,6 +20,7 @@ import {
ListRecentsRequest,
CommentExportFormat,
PreviewFormat,
TRASH_RETENTION_DAYS,
} from '@shumai/dtos'
import {
exportCommentsToFormat,
@@ -28,7 +29,15 @@ import {
type ExportOptions,
type ExportResult,
} from './comments-export'
import { type Asset, AssetStatus, AssetType, Prisma, type StorageKey } from '@shumai/db'
import {
type Asset,
AssetStatus,
AssetType,
Prisma,
type StorageKey,
WorkflowTaskStatus,
WorkflowTaskType,
} from '@shumai/db'
import { HTTPException } from 'hono/http-exception'
import { logger } from '@shumai/core/src/logger'
import { PaginatedData, paginateQuery, PaginationParams } from '@shumai/core/src/pagination'
@@ -46,6 +55,9 @@ import { getAgentRequiredLevel, getRoleLevel } from '@shumai/core/src/agent/perm
import { resolveEffectiveRole } from '@shumai/core/src/authz/authz'
import { metadataService } from '@shumai/core/src/metadata/metadata'
import { uploadService } from '@shumai/core/src/upload/upload'
import path from 'path'
import { stemFromKey } from '@shumai/core/src/utils/filename'
import { rewriteM3u8WithPresignedUrls } from '@shumai/core/src/transcode/transcode'
export const assetInclude = {
creator: true,
@@ -88,6 +100,31 @@ export type AssetWithIncludes = Prisma.AssetGetPayload<{
include: typeof assetInclude
}>
/**
* Same shape as {@link assetInclude}, but keeps soft-deleted children.
*
* The "Recently Deleted" view needs this because deleting a folder soft-deletes
* all of its descendants by cascade, so filtering out deleted children leaves
* the folder card with no previews of what was inside it.
*/
export const assetIncludeIncludingDeletedChildren = {
...assetInclude,
target: {
...assetInclude.target,
include: {
...assetInclude.target.include,
children: {
...assetInclude.target.include.children,
where: {},
},
},
},
children: {
...assetInclude.children,
where: {},
},
} as const
type CommentWithIncludes = Prisma.AssetCommentGetPayload<{
include: {
creator: true
@@ -688,7 +725,7 @@ export class AssetService {
throw new Error('ProjectID is required when ShowDeleted is true')
}
const thirtyDaysAgo = new Date()
thirtyDaysAgo.setDate(thirtyDaysAgo.getDate() - 30)
thirtyDaysAgo.setDate(thirtyDaysAgo.getDate() - TRASH_RETENTION_DAYS)
where = {
projectId: req.projectId,
@@ -741,11 +778,13 @@ export class AssetService {
break
}
const include = req.showDeleted ? assetIncludeIncludingDeletedChildren : assetInclude
const { data: assets, pageInfo } = await paginateQuery(
async (skip, take) => {
return this.prismaClient.asset.findMany({
where,
include: assetInclude,
include,
orderBy,
skip,
take,
@@ -755,7 +794,9 @@ export class AssetService {
req,
)
const infos = await this.toAssetInfos(assets)
const infos = await this.toAssetInfos(assets, undefined, {
includeDeletedChildren: req.showDeleted,
})
return { data: infos, pageInfo }
}
@@ -1341,7 +1382,7 @@ export class AssetService {
private async cascadeStatusToPendingPurge(rootIds: string[]): Promise<void> {
if (rootIds.length === 0) return
await this.prismaClient.$executeRaw`
const rows = await this.prismaClient.$queryRaw<{ id: string }[]>`
WITH RECURSIVE descendant AS (
SELECT id FROM assets WHERE id = ANY(${rootIds})
UNION ALL
@@ -1356,15 +1397,35 @@ export class AssetService {
WHERE s.type = 'symlink' AND s.target_id IN (SELECT id FROM all_purging_targets)
)
UPDATE assets SET status = 'pending_purge', updated_at = NOW()
WHERE id IN (SELECT id FROM all_purging_targets) OR id IN (SELECT id FROM all_affected_symlinks);
WHERE id IN (SELECT id FROM all_purging_targets) OR id IN (SELECT id FROM all_affected_symlinks)
RETURNING id;
`
const affectedIds = rows.map((r) => r.id)
if (affectedIds.length > 0) {
const inFlightTasks = await this.prismaClient.workflowTask.findMany({
where: {
assetId: { in: affectedIds },
status: { in: [WorkflowTaskStatus.pending, WorkflowTaskStatus.processing] },
},
select: { id: true },
})
for (const task of inFlightTasks) {
try {
logger.info({ taskId: task.id }, 'Cancelling in-flight workflow task for purging asset')
await cancelWorkflowTask(task.id)
} catch (err) {
logger.warn({ taskId: task.id, err }, 'Failed to cancel workflow task during purge')
}
}
}
}
async emptyTrash(projectId: string): Promise<void> {
const trashedRoots = await this.prismaClient.asset.findMany({
where: {
projectId,
status: 'trashed',
isDeleted: true,
},
select: { id: true },
@@ -1411,14 +1472,14 @@ export class AssetService {
* Cascade the 'pending_purge' status to all their descendants.
*/
private async expireTrashedAssets() {
const thirtyDaysAgo = new Date(Date.now() - 30 * 24 * 60 * 60 * 1000)
const thirtyDaysAgo = new Date(Date.now() - TRASH_RETENTION_DAYS * 24 * 60 * 60 * 1000)
// Find up to 100 expired roots, ordered by oldest first to avoid starvation
const expiredRoots = await this.prismaClient.asset.findMany({
where: {
status: 'trashed',
deletedAt: { lt: thirtyDaysAgo },
isDeleted: true,
status: { not: AssetStatus.pending_purge },
deletedAt: { lt: thirtyDaysAgo },
},
orderBy: { deletedAt: 'asc' },
select: { id: true },
@@ -1433,6 +1494,35 @@ export class AssetService {
* Stage 2: Delete DB records for assets marked as 'pending_purge'.
*/
private async purgePendingAssets() {
const pendingAssets = await this.prismaClient.asset.findMany({
where: { status: AssetStatus.pending_purge },
select: { id: true },
})
const pendingAssetIds = pendingAssets.map((a) => a.id)
if (pendingAssetIds.length > 0) {
const inFlightTasks = await this.prismaClient.workflowTask.findMany({
where: {
assetId: { in: pendingAssetIds },
status: { in: [WorkflowTaskStatus.pending, WorkflowTaskStatus.processing] },
},
select: { id: true },
})
for (const task of inFlightTasks) {
try {
logger.info({ taskId: task.id }, 'Cancelling in-flight workflow task before purge')
await cancelWorkflowTask(task.id)
} catch (err) {
logger.warn({ taskId: task.id, err }, 'Failed to cancel workflow task during purge')
}
}
await this.prismaClient.workflowTask.deleteMany({
where: { assetId: { in: pendingAssetIds } },
})
}
const { count } = await this.prismaClient.asset.deleteMany({
where: { status: AssetStatus.pending_purge },
})
@@ -1575,11 +1665,30 @@ export class AssetService {
}
}
const activeTask = await tx.workflowTask.findFirst({
where: {
assetId: a.id,
status: { in: [WorkflowTaskStatus.pending, WorkflowTaskStatus.processing] },
type: {
in: [
WorkflowTaskType.transcode,
WorkflowTaskType.transcode_video,
WorkflowTaskType.transcode_image,
WorkflowTaskType.transcode_pdf,
WorkflowTaskType.transcode_pdf_pages,
WorkflowTaskType.transcode_screenshot,
WorkflowTaskType.transcode_image_annotation,
WorkflowTaskType.transcode_watermark,
],
},
},
})
await tx.asset.update({
where: { id: a.id },
data: {
isDeleted: false,
status: 'processed',
status: activeTask ? AssetStatus.processing : AssetStatus.processed,
deletedAt: null,
},
})
@@ -2045,7 +2154,9 @@ export class AssetService {
async toAssetInfos(
assets: AssetWithIncludes[],
previewFormat?: PreviewFormat,
options?: { includeDeletedChildren?: boolean },
): Promise<AssetInfo[]> {
const includeDeletedChildren = options?.includeDeletedChildren ?? false
const stackIds = new Set<string>()
for (const a of assets) {
@@ -2076,7 +2187,10 @@ export class AssetService {
const versionsMap = new Map<string, AssetWithIncludes['children'][0][]>()
if (stackIds.size > 0) {
const allVersions = await this.prismaClient.asset.findMany({
where: { parentId: { in: Array.from(stackIds) }, isDeleted: false },
where: {
parentId: { in: Array.from(stackIds) },
...(includeDeletedChildren ? {} : { isDeleted: false }),
},
include: {
creator: true,
agent: { include: { user: true } },
@@ -2202,7 +2316,7 @@ export class AssetService {
folderForPreview.children
) {
for (const child of folderForPreview.children) {
if (child.isDeleted) continue
if (!includeDeletedChildren && child.isDeleted) continue
let previewAsset = child as Asset
let mediaType = child.mediaType
@@ -2273,6 +2387,9 @@ export class AssetService {
'GET',
)
}
if (media && media.isHls && media.hls?.key) {
media.hls.url = `/api/files/${latestVersion.id}/m3u8/master.m3u8`
}
const key = latestVersion.storageKey?.key
if (key) {
@@ -2290,7 +2407,13 @@ export class AssetService {
}
}
const proxyType = (media?.proxyType || null) as 'image' | 'video' | 'audio' | 'pdf' | null
const proxyType = (media?.proxyType || null) as
| 'image'
| 'video'
| 'audio'
| 'pdf'
| '3d'
| null
result.push({
id: a.id,
@@ -2594,6 +2717,7 @@ export class AssetService {
let storageKey = asset.storageKey.key
let assetDir = storageKey.substring(0, storageKey.lastIndexOf('/') + 1)
let targetName = asset.name
if (!key.startsWith(assetDir) && asset.parentId) {
// Check if requested key belongs to another version in the same version stack
@@ -2603,11 +2727,14 @@ export class AssetService {
isDeleted: false,
storageKey: { key: { startsWith: key.substring(0, key.lastIndexOf('/') + 1) } },
},
select: { storageKey: { select: { key: true } } },
select: { name: true, storageKey: { select: { key: true } } },
})
if (sibling?.storageKey?.key) {
storageKey = sibling.storageKey.key
assetDir = storageKey.substring(0, storageKey.lastIndexOf('/') + 1)
if (sibling.name) {
targetName = sibling.name
}
}
}
@@ -2615,13 +2742,35 @@ export class AssetService {
throw new Error('Key does not belong to this asset')
}
return s3Service.presign(
process.env.S3_BUCKET || 'shumai',
key,
'GET',
true,
asset.name ?? undefined,
)
let downloadFilename: string | undefined
if (key === storageKey) {
downloadFilename = targetName ?? undefined
} else {
const keyBasename = path.posix.basename(key)
const originalStem = stemFromKey(storageKey)
const assetStem = targetName ? stemFromKey(targetName) : originalStem
if (
originalStem &&
assetStem &&
originalStem !== assetStem &&
keyBasename.startsWith(originalStem + '-')
) {
downloadFilename = assetStem + keyBasename.slice(originalStem.length)
} else if (
originalStem &&
assetStem &&
originalStem !== assetStem &&
keyBasename.startsWith('watermark-' + originalStem + '-')
) {
downloadFilename =
'watermark-' + assetStem + keyBasename.slice(('watermark-' + originalStem).length)
} else {
downloadFilename = keyBasename
}
}
return s3Service.presign(process.env.S3_BUCKET || 'shumai', key, 'GET', true, downloadFilename)
}
async toPreviewInfo(
@@ -2630,7 +2779,13 @@ export class AssetService {
): Promise<PreviewInfo | null> {
if (!asset.media) return null
const proxyType = (asset.media?.proxyType || null) as 'image' | 'video' | 'audio' | 'pdf' | null
const proxyType = (asset.media?.proxyType || null) as
| 'image'
| 'video'
| 'audio'
| 'pdf'
| '3d'
| null
let rawThumbKey: string | undefined
if (proxyType === 'image' && asset.media.thumbnail?.key) {
@@ -2683,7 +2838,7 @@ export class AssetService {
originalHeight: asset.media.metadata?.originalHeight,
originalWidth: asset.media.metadata?.originalWidth,
spriteUrl,
duration: asset.media.metadata?.duration,
duration: proxyType === '3d' ? undefined : asset.media.metadata?.duration,
pageCount: asset.media.metadata?.totalFrames,
}
}
@@ -3057,6 +3212,95 @@ export class AssetService {
req,
)
}
async getHlsMasterPlaylist(params: {
assetId: string
watermarkConfigId?: string | null
}): Promise<string> {
const bucket = process.env.S3_BUCKET || 'shumai'
const targetAssetId = await this.resolveLatestVersionId(params.assetId)
if (params.watermarkConfigId) {
const wf = await this.prismaClient.watermarkFile.findUnique({
where: {
// eslint-disable-next-line @typescript-eslint/naming-convention
assetId_watermarkConfigId: {
assetId: targetAssetId,
watermarkConfigId: params.watermarkConfigId,
},
},
})
const media = wf?.media as PrismaJson.MediaInfo | null
if (!media?.isHls || !media.hls?.key) {
throw new HTTPException(404, { message: 'HLS playlist not found' })
}
const obj = await s3Service.getObject(bucket, media.hls.key)
return obj.buffer.toString('utf-8')
}
const asset = await this.prismaClient.asset.findUnique({
where: { id: targetAssetId },
})
if (!asset) {
throw new HTTPException(404, { message: 'Asset not found' })
}
const media = asset.media as PrismaJson.MediaInfo | null
if (!media?.isHls || !media.hls?.key) {
throw new HTTPException(404, { message: 'HLS playlist not found' })
}
const obj = await s3Service.getObject(bucket, media.hls.key)
return obj.buffer.toString('utf-8')
}
async getHlsVariantPlaylist(params: {
assetId: string
resolution: string
watermarkConfigId?: string | null
}): Promise<string> {
const bucket = process.env.S3_BUCKET || 'shumai'
if (!/^\d+p$/.test(params.resolution)) {
throw new HTTPException(400, { message: 'Invalid resolution format' })
}
const targetAssetId = await this.resolveLatestVersionId(params.assetId)
let media: PrismaJson.MediaInfo | null
if (params.watermarkConfigId) {
const wf = await this.prismaClient.watermarkFile.findUnique({
where: {
// eslint-disable-next-line @typescript-eslint/naming-convention
assetId_watermarkConfigId: {
assetId: targetAssetId,
watermarkConfigId: params.watermarkConfigId,
},
},
})
media = (wf?.media as PrismaJson.MediaInfo | null) ?? null
} else {
const asset = await this.prismaClient.asset.findUnique({
where: { id: targetAssetId },
})
if (!asset) {
throw new HTTPException(404, { message: 'Asset not found' })
}
media = (asset.media as PrismaJson.MediaInfo | null) ?? null
}
if (!media?.isHls || !media.hls?.key) {
throw new HTTPException(404, { message: 'HLS playlist not found' })
}
const matchedRes = media.hls.resolutions?.find((r) => r.resolution === params.resolution)
const variantKey =
matchedRes?.key ||
path.posix.join(path.posix.dirname(media.hls.key), params.resolution, 'index.m3u8')
const obj = await s3Service.getObject(bucket, variantKey)
const rawM3u8 = obj.buffer.toString('utf-8')
const s3Prefix = path.posix.dirname(variantKey)
return await rewriteM3u8WithPresignedUrls(rawM3u8, s3Prefix, bucket)
}
}
export const assetService = new AssetService()
@@ -2,9 +2,12 @@ import { describe, it, expect, beforeEach, afterEach, vi } from 'vitest'
import * as fs from 'fs'
import * as path from 'path'
import * as os from 'os'
import { prisma } from '@shumai/db'
import { setupTestDbHooks } from '@shumai/db/test'
import { GotenbergService } from './gotenberg'
describe('GotenbergService', () => {
setupTestDbHooks()
const originalEnv = process.env
let tmpDir: string
let testFilePath: string
@@ -216,4 +219,148 @@ describe('GotenbergService', () => {
expect(indexHtmlText).toContain('{{ toHTML "file.md" }}')
})
})
describe('team settings & fallback', () => {
it('resolves config from team settings when env is not set', async () => {
delete process.env.GOTENBERG_URL
const team = await prisma.team.create({
data: {
name: 'Gotenberg Team',
settings: {
gotenberg: {
url: 'http://custom-gotenberg:4000',
username: 'team-user',
password: 'team-password',
},
},
},
})
const service = new GotenbergService()
const config = await service.resolveConfig(team.id)
expect(config).toEqual({
url: 'http://custom-gotenberg:4000',
username: 'team-user',
password: 'team-password',
isEnv: false,
})
})
it('prioritizes environment variables over team settings', async () => {
process.env.GOTENBERG_URL = 'http://env-gotenberg:3000'
const team = await prisma.team.create({
data: {
name: 'Gotenberg Team',
settings: {
gotenberg: {
url: 'http://custom-gotenberg:4000',
},
},
},
})
const service = new GotenbergService()
const config = await service.resolveConfig(team.id)
expect(config.url).toBe('http://env-gotenberg:3000')
expect(config.isEnv).toBe(true)
})
it('getSettings returns masked settings response', async () => {
delete process.env.GOTENBERG_URL
const team = await prisma.team.create({
data: {
name: 'Gotenberg Team',
settings: {
gotenberg: {
url: 'http://custom-gotenberg:4000',
username: 'team-user',
password: 'team-password',
},
},
},
})
const service = new GotenbergService()
const settings = await service.getSettings(team.id)
expect(settings).toEqual({
url: 'http://custom-gotenberg:4000',
username: 'team-user',
hasPassword: true,
isEnvConfigured: false,
})
})
it('updateSettings updates team settings and rejects edits when env is set', async () => {
delete process.env.GOTENBERG_URL
const team = await prisma.team.create({
data: { name: 'Gotenberg Team' },
})
const service = new GotenbergService()
const updated = await service.updateSettings(team.id, {
url: 'http://new-gotenberg:3000',
username: 'admin',
password: 'secret-password',
})
expect(updated.url).toBe('http://new-gotenberg:3000')
expect(updated.username).toBe('admin')
expect(updated.hasPassword).toBe(true)
// When env is set, reject update
process.env.GOTENBERG_URL = 'http://env-gotenberg:3000'
await expect(service.updateSettings(team.id, { url: 'http://another:3000' })).rejects.toThrow(
'Gotenberg settings are configured via environment variables and cannot be edited',
)
})
})
describe('testConnection', () => {
it('returns error when url is empty and no team settings exist', async () => {
delete process.env.GOTENBERG_URL
const service = new GotenbergService()
const result = await service.testConnection({})
expect(result.ok).toBe(false)
expect(result.error).toBe('URL is required')
})
it('successfully connects to /health and enriches with /version', async () => {
const fetchSpy = vi.spyOn(globalThis, 'fetch').mockImplementation(async (url) => {
const urlStr = String(url)
if (urlStr.endsWith('/health')) {
return { ok: true, status: 200 } as Response
}
if (urlStr.endsWith('/version')) {
return {
ok: true,
status: 200,
json: async () => ({ version: '8.3.0' }),
} as Response
}
return { ok: false, status: 404 } as Response
})
const service = new GotenbergService()
const result = await service.testConnection({
url: 'http://gotenberg.local:3000',
username: 'test-user',
password: 'test-pass',
})
expect(result.ok).toBe(true)
expect(result.version).toBe('8.3.0')
expect(fetchSpy).toHaveBeenCalledWith(
'http://gotenberg.local:3000/health',
expect.objectContaining({
method: 'GET',
headers: expect.objectContaining({
Authorization: expect.stringMatching(/^Basic /),
}),
}),
)
})
})
})
+126 -29
View File
@@ -1,40 +1,135 @@
import * as fs from 'fs'
import type { GotenbergSettingsResponse, UpdateGotenbergSettingsRequest } from '@shumai/dtos'
import {
IntegrationDefinition,
resolveIntegrationConfig,
getIntegrationSettings,
updateIntegrationSettings,
resolveTestConnectionConfig,
getIntegrationBasicAuthHeader,
ResolvedIntegrationConfig,
} from '@shumai/core/src/integrations/helper'
export const GOTENBERG_INTEGRATION_DEF: IntegrationDefinition<'gotenberg'> = {
name: 'Gotenberg',
settingsKey: 'gotenberg',
envKeys: {
url: ['GOTENBERG_URL'],
username: ['GOTENBERG_BASIC_AUTH_USERNAME'],
password: ['GOTENBERG_BASIC_AUTH_PASSWORD'],
},
}
export interface GotenbergConvertOptions {
landscape?: boolean
teamId?: string
}
export class GotenbergService {
private getUrl(): string | undefined {
const url = process.env.GOTENBERG_URL?.trim()
return url ? url.replace(/\/+$/, '') : undefined
async resolveConfig(teamId?: string): Promise<ResolvedIntegrationConfig> {
return resolveIntegrationConfig(GOTENBERG_INTEGRATION_DEF, teamId)
}
private getAuthHeader(): string | undefined {
const username = process.env.GOTENBERG_BASIC_AUTH_USERNAME
const password = process.env.GOTENBERG_BASIC_AUTH_PASSWORD
private getAuthHeader(config: { username?: string; password?: string }): string | undefined {
return getIntegrationBasicAuthHeader(config)
}
if (username && password) {
const credentials = Buffer.from(`${username}:${password}`).toString('base64')
return `Basic ${credentials}`
async getSettings(teamId: string): Promise<GotenbergSettingsResponse> {
return getIntegrationSettings(GOTENBERG_INTEGRATION_DEF, teamId)
}
async updateSettings(
teamId: string,
req: UpdateGotenbergSettingsRequest,
): Promise<GotenbergSettingsResponse> {
return updateIntegrationSettings(GOTENBERG_INTEGRATION_DEF, teamId, req)
}
async testConnection(
config?: UpdateGotenbergSettingsRequest,
teamId?: string,
): Promise<{ ok: boolean; version?: string; error?: string }> {
const target = await resolveTestConnectionConfig(GOTENBERG_INTEGRATION_DEF, config, teamId)
if (!target.resolvedUrl) {
return { ok: false, error: 'URL is required' }
}
const cleanUrl = target.resolvedUrl.replace(/\/+$/, '')
try {
const headers: Record<string, string> = {}
const authHeader = this.getAuthHeader({
username: target.resolvedUsername,
password: target.resolvedPassword,
})
if (authHeader) {
headers['Authorization'] = authHeader
}
const response = await fetch(`${cleanUrl}/health`, {
method: 'GET',
headers,
signal: AbortSignal.timeout(5000),
})
if (!response.ok) {
const errText = await response.text().catch(() => '')
return {
ok: false,
error: `Health check returned status ${response.status}: ${errText || response.statusText}`,
}
}
// Optionally try /version to obtain version string
let version: string | undefined
try {
const versionRes = await fetch(`${cleanUrl}/version`, {
method: 'GET',
headers,
signal: AbortSignal.timeout(2000),
})
if (versionRes.ok) {
const vData = (await versionRes.json().catch(() => null)) as
| { version?: string }
| string
| null
if (vData && typeof vData === 'object' && 'version' in vData) {
version = String(vData.version)
} else if (typeof vData === 'string') {
version = vData
}
}
} catch {
// Version endpoint is optional in some Gotenberg versions
}
return {
ok: true,
version,
}
} catch (err) {
const msg = err instanceof Error ? err.message : String(err)
return {
ok: false,
error: msg,
}
}
return undefined
}
async isAvailable(): Promise<boolean> {
const baseUrl = this.getUrl()
if (!baseUrl) {
async isAvailable(teamId?: string): Promise<boolean> {
const config = await this.resolveConfig(teamId)
if (!config.url) {
return false
}
try {
const headers: Record<string, string> = {}
const authHeader = this.getAuthHeader()
const authHeader = this.getAuthHeader(config)
if (authHeader) {
headers['Authorization'] = authHeader
}
const response = await fetch(`${baseUrl}/health`, {
const response = await fetch(`${config.url}/health`, {
method: 'GET',
headers,
signal: AbortSignal.timeout(3000),
@@ -50,9 +145,11 @@ export class GotenbergService {
filePath: string,
filename: string,
options: GotenbergConvertOptions = {},
teamId?: string,
): Promise<Buffer> {
const baseUrl = this.getUrl()
if (!baseUrl) {
const effectiveTeamId = teamId ?? options.teamId
const config = await this.resolveConfig(effectiveTeamId)
if (!config.url) {
throw new Error('Gotenberg URL is not configured.')
}
@@ -65,12 +162,12 @@ export class GotenbergService {
}
const headers: Record<string, string> = {}
const authHeader = this.getAuthHeader()
const authHeader = this.getAuthHeader(config)
if (authHeader) {
headers['Authorization'] = authHeader
}
const response = await fetch(`${baseUrl}/forms/libreoffice/convert`, {
const response = await fetch(`${config.url}/forms/libreoffice/convert`, {
method: 'POST',
headers,
body: formData,
@@ -88,9 +185,9 @@ export class GotenbergService {
return Buffer.from(arrayBuffer)
}
async convertHtmlToPdf(filePath: string, filename: string): Promise<Buffer> {
const baseUrl = this.getUrl()
if (!baseUrl) {
async convertHtmlToPdf(filePath: string, filename: string, teamId?: string): Promise<Buffer> {
const config = await this.resolveConfig(teamId)
if (!config.url) {
throw new Error('Gotenberg URL is not configured.')
}
@@ -103,12 +200,12 @@ export class GotenbergService {
formData.append('files', new Blob([fileBytes]), targetFilename)
const headers: Record<string, string> = {}
const authHeader = this.getAuthHeader()
const authHeader = this.getAuthHeader(config)
if (authHeader) {
headers['Authorization'] = authHeader
}
const response = await fetch(`${baseUrl}/forms/chromium/convert/html`, {
const response = await fetch(`${config.url}/forms/chromium/convert/html`, {
method: 'POST',
headers,
body: formData,
@@ -126,9 +223,9 @@ export class GotenbergService {
return Buffer.from(arrayBuffer)
}
async convertMarkdownToPdf(filePath: string, filename: string): Promise<Buffer> {
const baseUrl = this.getUrl()
if (!baseUrl) {
async convertMarkdownToPdf(filePath: string, filename: string, teamId?: string): Promise<Buffer> {
const config = await this.resolveConfig(teamId)
if (!config.url) {
throw new Error('Gotenberg URL is not configured.')
}
@@ -151,12 +248,12 @@ export class GotenbergService {
formData.append('files', new Blob([fileBytes]), mdFilename)
const headers: Record<string, string> = {}
const authHeader = this.getAuthHeader()
const authHeader = this.getAuthHeader(config)
if (authHeader) {
headers['Authorization'] = authHeader
}
const response = await fetch(`${baseUrl}/forms/chromium/convert/markdown`, {
const response = await fetch(`${config.url}/forms/chromium/convert/markdown`, {
method: 'POST',
headers,
body: formData,
+6 -1
View File
@@ -5,11 +5,13 @@ export * from './authz/authz'
export * from './collection/collection'
export * from './invite/invite'
export * from './metadata/metadata'
export * from './notification/notification'
export * from './notification/index'
export * from './project/project'
export * from './provider/provider'
export * from './s3/s3'
export * from './search/search'
export * from './search/text-embedding'
export * from './search/embedding-service'
export * from './share/share'
export * from './skill/skill'
export * from './team/team'
@@ -31,3 +33,6 @@ export * from './kanban/kanban'
export * from './media-generation/media-generation'
export * from './sandbox'
export * from './asset/comments-export'
export * from './turntable/turntable'
export * from './utils/model3d'
export * from './integrations/helper'
@@ -0,0 +1,271 @@
import { describe, it, expect, beforeEach, afterEach } from 'vitest'
import { prisma } from '@shumai/db'
import { setupTestDbHooks } from '@shumai/db/test'
import {
IntegrationDefinition,
resolveIntegrationConfig,
getIntegrationSettings,
updateIntegrationSettings,
resolveTestConnectionConfig,
getIntegrationBasicAuthHeader,
} from './helper'
const TEST_INTEGRATION_DEF: IntegrationDefinition<'turntable'> = {
name: 'TestIntegration',
settingsKey: 'turntable',
envKeys: {
url: ['TEST_INTEGRATION_URL', 'FALLBACK_INTEGRATION_URL'],
username: ['TEST_INTEGRATION_USERNAME'],
password: ['TEST_INTEGRATION_PASSWORD'],
},
}
describe('Integration Config Helper', () => {
setupTestDbHooks()
const originalEnv = process.env
beforeEach(() => {
process.env = { ...originalEnv }
delete process.env.TEST_INTEGRATION_URL
delete process.env.FALLBACK_INTEGRATION_URL
delete process.env.TEST_INTEGRATION_USERNAME
delete process.env.TEST_INTEGRATION_PASSWORD
})
afterEach(() => {
process.env = originalEnv
})
describe('resolveIntegrationConfig', () => {
it('prioritizes environment variables over database settings', async () => {
const team = await prisma.team.create({
data: {
name: 'Test Team',
settings: {
turntable: {
url: 'http://db-server:1234',
username: 'db-user',
password: 'db-password',
},
},
},
})
process.env.TEST_INTEGRATION_URL = 'http://env-server:5678/'
process.env.TEST_INTEGRATION_USERNAME = 'env-user'
process.env.TEST_INTEGRATION_PASSWORD = 'env-pass'
const config = await resolveIntegrationConfig(TEST_INTEGRATION_DEF, team.id)
expect(config).toEqual({
url: 'http://env-server:5678',
username: 'env-user',
password: 'env-pass',
isEnv: true,
})
})
it('falls back to second env key if first is missing', async () => {
process.env.FALLBACK_INTEGRATION_URL = 'http://fallback-server:8080'
const config = await resolveIntegrationConfig(TEST_INTEGRATION_DEF)
expect(config.url).toBe('http://fallback-server:8080')
expect(config.isEnv).toBe(true)
})
it('reads from team settings when environment variables are not set', async () => {
const team = await prisma.team.create({
data: {
name: 'Test Team',
settings: {
turntable: {
url: 'http://db-server:1234/',
username: 'db-user',
password: 'db-password',
},
},
},
})
const config = await resolveIntegrationConfig(TEST_INTEGRATION_DEF, team.id)
expect(config).toEqual({
url: 'http://db-server:1234',
username: 'db-user',
password: 'db-password',
isEnv: false,
})
})
it('returns isEnv: false and undefined fields when neither env nor DB are set', async () => {
const team = await prisma.team.create({
data: { name: 'Empty Team' },
})
const config = await resolveIntegrationConfig(TEST_INTEGRATION_DEF, team.id)
expect(config).toEqual({
isEnv: false,
})
})
})
describe('getIntegrationSettings', () => {
it('returns safe response masking password', async () => {
const team = await prisma.team.create({
data: {
name: 'Test Team',
settings: {
turntable: {
url: 'http://db-server:1234',
username: 'db-user',
password: 'db-password',
},
},
},
})
const res = await getIntegrationSettings(TEST_INTEGRATION_DEF, team.id)
expect(res).toEqual({
url: 'http://db-server:1234',
username: 'db-user',
hasPassword: true,
isEnvConfigured: false,
})
})
})
describe('updateIntegrationSettings', () => {
it('throws error when environment variable is configured', async () => {
const team = await prisma.team.create({
data: { name: 'Test Team' },
})
process.env.TEST_INTEGRATION_URL = 'http://env-server:5678'
await expect(
updateIntegrationSettings(TEST_INTEGRATION_DEF, team.id, {
url: 'http://new-url:1234',
}),
).rejects.toThrow(
'TestIntegration settings are configured via environment variables and cannot be edited',
)
})
it('updates settings and preserves existing password when not supplied', async () => {
const team = await prisma.team.create({
data: {
name: 'Test Team',
settings: {
turntable: {
url: 'http://original:1234',
username: 'original-user',
password: 'secret-password',
},
},
},
})
const res = await updateIntegrationSettings(TEST_INTEGRATION_DEF, team.id, {
url: 'http://updated:5678/',
username: 'new-user',
})
expect(res).toEqual({
url: 'http://updated:5678',
username: 'new-user',
hasPassword: true,
isEnvConfigured: false,
})
const config = await resolveIntegrationConfig(TEST_INTEGRATION_DEF, team.id)
expect(config.password).toBe('secret-password')
})
it('clears settings in database when URL is empty or whitespace', async () => {
const team = await prisma.team.create({
data: {
name: 'Test Team',
settings: {
turntable: {
url: 'http://to-be-cleared:1234',
username: 'user',
password: 'pwd',
},
},
},
})
const res = await updateIntegrationSettings(TEST_INTEGRATION_DEF, team.id, {
url: ' ',
})
expect(res).toEqual({
url: undefined,
username: undefined,
hasPassword: false,
isEnvConfigured: false,
})
const updatedTeam = await prisma.team.findUnique({
where: { id: team.id },
})
// Check that turntable is removed from team settings
expect((updatedTeam?.settings as Record<string, unknown>)?.turntable).toBeUndefined()
})
})
describe('resolveTestConnectionConfig', () => {
it('uses provided credentials directly when supplied', async () => {
const target = await resolveTestConnectionConfig(TEST_INTEGRATION_DEF, {
url: 'http://direct:1234',
username: 'direct-user',
password: 'direct-password',
})
expect(target).toEqual({
resolvedUrl: 'http://direct:1234',
resolvedUsername: 'direct-user',
resolvedPassword: 'direct-password',
})
})
it('falls back to saved database settings when testing with partial inputs', async () => {
const team = await prisma.team.create({
data: {
name: 'Test Team',
settings: {
turntable: {
url: 'http://db:1234',
username: 'db-user',
password: 'db-password',
},
},
},
})
const target = await resolveTestConnectionConfig(
TEST_INTEGRATION_DEF,
{ url: 'http://new-url:9999' },
team.id,
)
expect(target).toEqual({
resolvedUrl: 'http://new-url:9999',
resolvedUsername: 'db-user',
resolvedPassword: 'db-password',
})
})
})
describe('getIntegrationBasicAuthHeader', () => {
it('generates base64 basic auth header when username and password are provided', () => {
const header = getIntegrationBasicAuthHeader({ username: 'admin', password: 'secret' })
expect(header).toBe(`Basic ${Buffer.from('admin:secret').toString('base64')}`)
})
it('returns undefined when username or password is missing', () => {
expect(getIntegrationBasicAuthHeader({ username: 'admin' })).toBeUndefined()
expect(getIntegrationBasicAuthHeader({ password: 'secret' })).toBeUndefined()
expect(getIntegrationBasicAuthHeader({})).toBeUndefined()
})
})
})
+188
View File
@@ -0,0 +1,188 @@
import { prisma } from '@shumai/db'
import '@shumai/db/src/prisma-json-types'
import { logger } from '@shumai/core/src/logger'
export interface IntegrationEnvKeys {
url: string[]
username?: string[]
password?: string[]
}
export interface IntegrationDefinition<TKey extends keyof PrismaJson.Settings> {
name: string
settingsKey: TKey
envKeys: IntegrationEnvKeys
}
export interface ResolvedIntegrationConfig {
url?: string
username?: string
password?: string
isEnv: boolean
}
export interface IntegrationSettingsResponse {
url?: string
username?: string
hasPassword: boolean
isEnvConfigured: boolean
}
export interface UpdateIntegrationSettingsRequest {
url?: string
username?: string
password?: string
}
function getFirstEnvValue(keys?: string[]): string | undefined {
if (!keys) return undefined
for (const key of keys) {
const val = process.env[key]?.trim()
if (val) return val
}
return undefined
}
export async function resolveIntegrationConfig<TKey extends keyof PrismaJson.Settings>(
def: IntegrationDefinition<TKey>,
teamId?: string,
): Promise<ResolvedIntegrationConfig> {
const envUrl = getFirstEnvValue(def.envKeys.url)
const envUsername = getFirstEnvValue(def.envKeys.username)
const envPassword = getFirstEnvValue(def.envKeys.password)
if (envUrl) {
return {
url: envUrl.replace(/\/+$/, ''),
username: envUsername || undefined,
password: envPassword || undefined,
isEnv: true,
}
}
if (teamId) {
try {
const team = await prisma.team.findUnique({
where: { id: teamId },
select: { settings: true },
})
const teamSettings = team?.settings as PrismaJson.Settings | null
const serviceSettings = teamSettings?.[def.settingsKey] as
| { url?: string; username?: string; password?: string }
| undefined
if (serviceSettings?.url?.trim()) {
return {
url: serviceSettings.url.trim().replace(/\/+$/, ''),
username: serviceSettings.username?.trim() || undefined,
password: serviceSettings.password?.trim() || undefined,
isEnv: false,
}
}
} catch (err) {
logger.error({ err, teamId }, `Failed to load team ${def.name} settings`)
}
}
return {
isEnv: false,
}
}
export async function getIntegrationSettings<TKey extends keyof PrismaJson.Settings>(
def: IntegrationDefinition<TKey>,
teamId: string,
): Promise<IntegrationSettingsResponse> {
const config = await resolveIntegrationConfig(def, teamId)
return {
url: config.url,
username: config.username,
hasPassword: Boolean(config.password),
isEnvConfigured: config.isEnv,
}
}
export async function updateIntegrationSettings<TKey extends keyof PrismaJson.Settings>(
def: IntegrationDefinition<TKey>,
teamId: string,
req: UpdateIntegrationSettingsRequest,
): Promise<IntegrationSettingsResponse> {
const currentConfig = await resolveIntegrationConfig(def, teamId)
if (currentConfig.isEnv) {
throw new Error(
`${def.name} settings are configured via environment variables and cannot be edited`,
)
}
const team = await prisma.team.findUnique({
where: { id: teamId },
select: { settings: true },
})
const settings = { ...((team?.settings || {}) as PrismaJson.Settings) }
const prevSettings = (settings[def.settingsKey] || {}) as {
url?: string
username?: string
password?: string
}
const trimmedUrl = req.url?.trim()
if (!trimmedUrl) {
// If URL is empty/unset, clear the integration settings
delete settings[def.settingsKey]
} else {
const updatedServiceSettings: { url: string; username?: string; password?: string } = {
url: trimmedUrl.replace(/\/+$/, ''),
username:
req.username !== undefined ? req.username.trim() || undefined : prevSettings.username,
password: req.password !== undefined ? req.password : prevSettings.password,
}
settings[def.settingsKey] = updatedServiceSettings as unknown as PrismaJson.Settings[TKey]
}
await prisma.team.update({
where: { id: teamId },
data: { settings },
})
return getIntegrationSettings(def, teamId)
}
export async function resolveTestConnectionConfig<TKey extends keyof PrismaJson.Settings>(
def: IntegrationDefinition<TKey>,
config?: UpdateIntegrationSettingsRequest,
teamId?: string,
): Promise<{ resolvedUrl?: string; resolvedUsername?: string; resolvedPassword?: string }> {
let resolvedUrl = config?.url?.trim()
let resolvedUsername = config?.username?.trim()
let resolvedPassword = config?.password
if (teamId && (!resolvedUrl || resolvedPassword === undefined)) {
const savedConfig = await resolveIntegrationConfig(def, teamId)
if (!resolvedUrl) {
resolvedUrl = savedConfig.url
}
if (!resolvedUsername && savedConfig.username) {
resolvedUsername = savedConfig.username
}
if (resolvedPassword === undefined) {
resolvedPassword = savedConfig.password
}
}
return {
resolvedUrl,
resolvedUsername,
resolvedPassword,
}
}
export function getIntegrationBasicAuthHeader(config: {
username?: string
password?: string
}): string | undefined {
if (config.username && config.password) {
const credentials = Buffer.from(`${config.username}:${config.password}`).toString('base64')
return `Basic ${credentials}`
}
return undefined
}
@@ -0,0 +1,700 @@
import { describe, expect, it, vi, beforeEach } from 'vitest'
import nodemailer from 'nodemailer'
import { prisma } from '@shumai/db'
import { setupTestDbHooks } from '@shumai/db/test'
import { NotificationType } from '@shumai/db'
import { emailService, DEFAULT_EMAIL_SETTINGS } from './email'
const mockSendMail = vi.fn().mockResolvedValue({ messageId: '<mock-msg-id@shumai.test>' })
const mockVerify = vi.fn().mockResolvedValue(true)
const mockClose = vi.fn()
vi.mock('nodemailer', () => ({
default: {
createTransport: vi.fn(() => ({
sendMail: mockSendMail,
verify: mockVerify,
close: mockClose,
})),
},
}))
describe('EmailService', () => {
setupTestDbHooks()
beforeEach(() => {
vi.clearAllMocks()
mockSendMail.mockResolvedValue({ messageId: '<mock-msg-id@shumai.test>' })
mockVerify.mockResolvedValue(true)
})
describe('getEmailSettings', () => {
it('returns default settings when team settings are not set', async () => {
const team = await prisma.team.create({ data: { name: 'Team Alpha' } })
const settings = await emailService.getEmailSettings(team.id)
expect(settings).toEqual(DEFAULT_EMAIL_SETTINGS)
expect(settings.enabled).toBe(false)
expect(settings.port).toBe(587)
})
it('returns saved email settings when configured', async () => {
const team = await prisma.team.create({
data: {
name: 'Team Beta',
settings: {
emailNotification: {
enabled: true,
host: 'smtp.beta.test',
port: 465,
username: 'beta_user',
password: 'secret_password',
smtps: true,
ignoreCert: false,
from: 'beta@shumai.test',
replyTo: 'reply@shumai.test',
},
},
},
})
const settings = await emailService.getEmailSettings(team.id)
expect(settings.enabled).toBe(true)
expect(settings.host).toBe('smtp.beta.test')
expect(settings.port).toBe(465)
expect(settings.password).toBe('secret_password')
})
it('throws 404 if team does not exist', async () => {
await expect(emailService.getEmailSettings('non-existent-team-id')).rejects.toThrow()
})
})
describe('updateEmailSettings', () => {
it('updates email settings and saves to team.settings', async () => {
const team = await prisma.team.create({ data: { name: 'Team Gamma' } })
const updated = await emailService.updateEmailSettings(team.id, {
enabled: true,
host: 'smtp.gamma.test',
port: 587,
username: 'gamma',
password: 'gamma_password',
smtps: false,
ignoreCert: true,
from: 'gamma@shumai.test',
replyTo: '',
})
expect(updated.enabled).toBe(true)
expect(updated.host).toBe('smtp.gamma.test')
expect(updated.password).toBe('gamma_password')
expect(updated.ignoreCert).toBe(true)
const reloaded = await prisma.team.findUnique({ where: { id: team.id } })
expect(reloaded?.settings?.emailNotification?.host).toBe('smtp.gamma.test')
})
it('preserves existing password if empty string or undefined is provided in update', async () => {
const team = await prisma.team.create({
data: {
name: 'Team Delta',
settings: {
emailNotification: {
enabled: true,
host: 'smtp.delta.test',
port: 587,
username: 'delta',
password: 'original_super_secret',
smtps: false,
ignoreCert: false,
from: 'delta@shumai.test',
replyTo: '',
},
},
},
})
const updated = await emailService.updateEmailSettings(team.id, {
enabled: true,
host: 'smtp.delta-updated.test',
port: 587,
username: 'delta_new',
password: '', // Empty password should retain existing
smtps: true,
ignoreCert: false,
from: 'delta@shumai.test',
})
expect(updated.host).toBe('smtp.delta-updated.test')
expect(updated.password).toBe('original_super_secret')
const reloaded = await prisma.team.findUnique({ where: { id: team.id } })
expect(reloaded?.settings?.emailNotification?.password).toBe('original_super_secret')
})
})
describe('sendMail and verifySmtp', () => {
it('verifySmtp calls transport.verify and closes transport', async () => {
const config = { ...DEFAULT_EMAIL_SETTINGS, host: 'smtp.test' }
const ok = await emailService.verifySmtp(config)
expect(ok).toBe(true)
expect(mockVerify).toHaveBeenCalled()
expect(mockClose).toHaveBeenCalled()
})
it('sendMail sends email using nodemailer transport', async () => {
const config = { ...DEFAULT_EMAIL_SETTINGS, host: 'smtp.test', from: 'shumai@test.com' }
const res = await emailService.sendMail(config, {
to: 'user@test.com',
subject: 'Hello',
html: '<p>Hello World</p>',
text: 'Hello World',
})
expect(res.messageId).toBe('<mock-msg-id@shumai.test>')
expect(mockSendMail).toHaveBeenCalledWith(
expect.objectContaining({
from: 'shumai@test.com',
to: 'user@test.com',
subject: 'Hello',
}),
)
expect(mockClose).toHaveBeenCalled()
})
it('sendMail forwards attachments to nodemailer transport', async () => {
const config = { ...DEFAULT_EMAIL_SETTINGS, host: 'smtp.test', from: 'shumai@test.com' }
const attachments = [
{
filename: 'avatar.png',
content: Buffer.from('image-data'),
cid: 'avatar-123@shumai.internal',
contentType: 'image/png',
},
]
await emailService.sendMail(config, {
to: 'user@test.com',
subject: 'Hello',
html: '<img src="cid:avatar-123@shumai.internal" />',
text: 'Hello',
attachments,
})
expect(mockSendMail).toHaveBeenCalledWith(
expect.objectContaining({
attachments,
}),
)
})
})
describe('sendTestEmail', () => {
it('sends test email to the current admin user', async () => {
const team = await prisma.team.create({
data: {
name: 'Test Team',
settings: {
emailNotification: {
enabled: true,
host: 'smtp.test.com',
port: 587,
username: 'user',
password: 'pw',
smtps: false,
ignoreCert: false,
from: 'sender@test.com',
replyTo: '',
},
},
},
})
const currentUser = { id: 'admin1', email: 'admin@shumai.test', name: 'Admin' }
const result = await emailService.sendTestEmail(team.id, currentUser)
expect(result.success).toBe(true)
expect(result.messageId).toBe('<mock-msg-id@shumai.test>')
expect(mockVerify).toHaveBeenCalled()
expect(mockSendMail).toHaveBeenCalledWith(
expect.objectContaining({
to: 'admin@shumai.test',
subject: 'Test email from Shumai',
}),
)
})
it('allows overriding settings with input draft', async () => {
const team = await prisma.team.create({ data: { name: 'Test Team 2' } })
const currentUser = { id: 'admin2', email: 'admin2@shumai.test', name: 'Admin 2' }
const result = await emailService.sendTestEmail(team.id, currentUser, {
host: 'smtp.draft.com',
from: 'draft@test.com',
password: 'draft-password',
})
expect(result.success).toBe(true)
expect(nodemailer.createTransport).toHaveBeenCalledWith(
expect.objectContaining({
host: 'smtp.draft.com',
}),
)
expect(mockSendMail).toHaveBeenCalledWith(
expect.objectContaining({
from: 'draft@test.com',
to: 'admin2@shumai.test',
}),
)
})
it('throws error when host is missing', async () => {
const team = await prisma.team.create({ data: { name: 'Empty Host Team' } })
const currentUser = { id: 'admin3', email: 'admin3@shumai.test', name: 'Admin 3' }
await expect(
emailService.sendTestEmail(team.id, currentUser, { host: '', from: 'a@b.com' }),
).rejects.toThrow('SMTP host is required')
})
it('throws error when from address is missing', async () => {
const team = await prisma.team.create({ data: { name: 'Empty From Team' } })
const currentUser = { id: 'admin4', email: 'admin4@shumai.test', name: 'Admin 4' }
await expect(
emailService.sendTestEmail(team.id, currentUser, { host: 'smtp.test', from: '' }),
).rejects.toThrow('From address is required')
})
it('throws 400 when verify fails', async () => {
const team = await prisma.team.create({ data: { name: 'Failing Team' } })
const currentUser = { id: 'admin5', email: 'admin5@shumai.test', name: 'Admin 5' }
mockVerify.mockRejectedValueOnce(new Error('Connection refused'))
await expect(
emailService.sendTestEmail(team.id, currentUser, {
host: 'smtp.down.com',
from: 'admin@down.com',
}),
).rejects.toThrow('SMTP verification failed: Connection refused')
})
})
describe('renderNotificationEmail', () => {
it('renders comment notification', () => {
const rendered = emailService.renderNotificationEmail({
type: NotificationType.comment_created,
creatorName: 'Alice',
teamName: 'Designers',
teamId: 'team-1',
projectId: 'proj-1',
assetId: 'asset-1',
assetName: 'logo.png',
commentMessage: 'Please make it bigger',
})
expect(rendered.subject).toContain('Alice commented on "logo.png"')
expect(rendered.html).toContain('Please make it bigger')
expect(rendered.html).toContain('logo.png')
expect(rendered.text).toContain('Alice commented on "logo.png"')
})
it('renders mention notification', () => {
const rendered = emailService.renderNotificationEmail({
type: NotificationType.mention,
creatorName: 'Bob',
teamName: 'Developers',
teamId: 'team-2',
projectId: 'proj-2',
assetId: 'asset-2',
assetName: 'architecture.pdf',
commentMessage: 'Take a look at this',
})
expect(rendered.subject).toContain('Bob mentioned you')
expect(rendered.html).toContain('Bob mentioned you')
expect(rendered.html).toContain('Take a look at this')
})
it('replaces user mentions with user names instead of user ID in comment emails', () => {
const rendered = emailService.renderNotificationEmail({
type: NotificationType.comment_created,
creatorName: 'Bob',
teamName: 'Designers',
teamId: 'team-1',
projectId: 'proj-1',
assetId: 'asset-1',
assetName: 'logo.png',
commentMessage: '<@kUFrm1W4fKLCaLpw2XQP1OPe0gKACJpy> nice',
mentionedUserNames: {
// eslint-disable-next-line @typescript-eslint/naming-convention
kUFrm1W4fKLCaLpw2XQP1OPe0gKACJpy: 'Alice',
},
})
expect(rendered.html).toContain('@Alice nice')
expect(rendered.html).not.toContain('<@kUFrm1W4fKLCaLpw2XQP1OPe0gKACJpy>')
expect(rendered.text).toContain('@Alice nice')
expect(rendered.text).not.toContain('<@kUFrm1W4fKLCaLpw2XQP1OPe0gKACJpy>')
})
it('renders asset upload notification with project name, creator profile image, and upload time', () => {
const rendered = emailService.renderNotificationEmail({
type: NotificationType.successful_file_uploaded,
creatorName: 'Charlie',
creatorAvatarUrl: 'https://example.com/avatar.jpg',
teamName: 'Media',
teamId: 'team-3',
projectName: 'Marketing Campaign',
projectId: 'proj-3',
assetId: 'asset-3',
assetName: 'video.mp4',
fileCount: 1,
uploadTime: new Date('2026-09-27T15:20:00Z'),
})
expect(rendered.subject).toContain('Charlie uploaded "video.mp4" to Marketing Campaign')
expect(rendered.html).toContain('Marketing Campaign')
expect(rendered.html).toContain('https://example.com/avatar.jpg')
expect(rendered.html).toContain('Sep 27, 2026')
expect(rendered.text).toContain('Marketing Campaign')
expect(rendered.text).toContain('Sep 27, 2026')
})
it('renders single file uploaded notification', () => {
const rendered = emailService.renderNotificationEmail({
type: NotificationType.successful_file_uploaded,
creatorName: 'Charlie',
teamName: 'Media',
teamId: 'team-3',
projectId: 'proj-3',
assetId: 'asset-3',
assetName: 'video.mp4',
fileCount: 1,
})
expect(rendered.subject).toContain('Charlie uploaded "video.mp4"')
expect(rendered.html).toContain('video.mp4')
})
it('renders consolidated batch file upload notification', () => {
const rendered = emailService.renderNotificationEmail({
type: NotificationType.successful_file_uploaded,
creatorName: 'Charlie',
teamName: 'Media',
teamId: 'team-3',
projectName: 'Videos',
projectId: 'proj-3',
fileCount: 5,
fileNames: ['img1.png', 'img2.png', 'img3.png', 'img4.png', 'img5.png'],
})
expect(rendered.subject).toContain('Charlie uploaded 5 files to Videos')
expect(rendered.html).toContain('img1.png')
expect(rendered.html).toContain('and 2 more')
})
it('renders user joined team notification', () => {
const rendered = emailService.renderNotificationEmail({
type: NotificationType.new_user_join_team,
creatorName: 'Dave',
teamName: 'Acme Corp',
teamId: 'team-4',
})
expect(rendered.subject).toContain('Dave joined Acme Corp')
expect(rendered.html).toContain('Dave is now a member of Acme Corp')
})
it('renders kanban task notifications', () => {
const created = emailService.renderNotificationEmail({
type: NotificationType.kanban_task_created,
creatorName: 'Eva',
teamName: 'Engineering',
teamId: 'team-5',
kanbanTaskId: 'kt-1',
kanbanTaskTitle: 'Fix login bug',
})
expect(created.subject).toContain('Eva created task "Fix login bug"')
const assigned = emailService.renderNotificationEmail({
type: NotificationType.kanban_task_assigned,
creatorName: 'Frank',
teamName: 'Engineering',
teamId: 'team-5',
kanbanTaskId: 'kt-2',
kanbanTaskTitle: 'Design DB schema',
})
expect(assigned.subject).toContain('Frank assigned you to task "Design DB schema"')
const status = emailService.renderNotificationEmail({
type: NotificationType.kanban_task_status_updated,
creatorName: 'Grace',
teamName: 'Engineering',
teamId: 'team-5',
kanbanTaskId: 'kt-3',
kanbanTaskTitle: 'Refactor auth',
commentMessage: 'Moved to In Progress',
})
expect(status.subject).toContain('Grace updated status of task "Refactor auth"')
expect(status.html).toContain('Moved to In Progress')
})
it('generates correct action links without team id for project and asset URLs', () => {
const baseUrl = emailService.getBaseUrl()
// Project link
const projectEmail = emailService.renderNotificationEmail({
type: NotificationType.new_user_join_project,
creatorName: 'Alice',
teamName: 'Designers',
teamId: '01KY2A7DRK2N0MECBDHTGAVV84',
projectName: 'Project Alpha',
projectId: '01KY2ANCA1SXGSPR63NVVD07EJ',
})
expect(projectEmail.html).toContain(`href="${baseUrl}/projects/01KY2ANCA1SXGSPR63NVVD07EJ"`)
expect(projectEmail.html).not.toContain(
'/teams/01KY2A7DRK2N0MECBDHTGAVV84/projects/01KY2ANCA1SXGSPR63NVVD07EJ',
)
// Asset link
const assetEmail = emailService.renderNotificationEmail({
type: NotificationType.comment_created,
creatorName: 'Alice',
teamName: 'Designers',
teamId: '01KY2A7DRK2N0MECBDHTGAVV84',
projectName: 'Project Alpha',
projectId: '01KY2ANCA1SXGSPR63NVVD07EJ',
assetId: '01KY2ASSET1234567890ABCDEF',
assetName: 'logo.png',
commentMessage: 'Looks good',
})
expect(assetEmail.html).toContain(
`href="${baseUrl}/projects/01KY2ANCA1SXGSPR63NVVD07EJ/files/01KY2ASSET1234567890ABCDEF"`,
)
expect(assetEmail.html).not.toContain('/teams/')
})
})
describe('renderBatchedNotificationEmail', () => {
it('renders single notification with specific subject and detailed row', () => {
const rendered = emailService.renderBatchedNotificationEmail({
teamId: 'team-1',
teamName: 'Designers',
recipientName: 'Bob',
items: [
{
id: 'n-1',
type: NotificationType.successful_file_uploaded,
creatorName: 'Alice',
creatorAvatarUrl: 'https://example.com/alice.png',
teamId: 'team-1',
teamName: 'Designers',
projectId: 'p-1',
projectName: 'Website Redesign',
assetId: 'a-1',
assetName: 'hero.png',
assetThumbnailUrl: 'https://example.com/hero-thumb.png',
createdAt: new Date('2026-09-27T10:00:00Z'),
},
],
})
expect(rendered.subject).toBe('[Shumai] Alice uploaded "hero.png" to Website Redesign')
expect(rendered.html).toContain('Alice')
expect(rendered.html).toContain('Uploaded &quot;hero.png&quot; to Website Redesign')
expect(rendered.html).toContain('https://example.com/hero-thumb.png')
expect(rendered.html).toContain('https://example.com/alice.png')
expect(rendered.text).toContain('Alice: Uploaded "hero.png" to Website Redesign')
})
it('renders detailed rows with thumbnail when event count <= 3', () => {
const rendered = emailService.renderBatchedNotificationEmail({
teamId: 'team-1',
teamName: 'Designers',
recipientName: 'Bob',
items: [
{
id: 'n-1',
type: NotificationType.successful_file_uploaded,
creatorName: 'Alice',
teamId: 'team-1',
teamName: 'Designers',
projectId: 'p-1',
projectName: 'Website Redesign',
assetId: 'a-1',
assetName: 'img1.png',
assetThumbnailUrl: 'https://example.com/img1.png',
createdAt: new Date('2026-09-27T10:00:00Z'),
},
{
id: 'n-2',
type: NotificationType.successful_file_uploaded,
creatorName: 'Alice',
teamId: 'team-1',
teamName: 'Designers',
projectId: 'p-1',
projectName: 'Website Redesign',
assetId: 'a-2',
assetName: 'img2.png',
createdAt: new Date('2026-09-27T10:01:00Z'),
},
],
})
expect(rendered.subject).toBe('[Shumai] 2 new notifications')
expect(rendered.html).toContain('You have 2 new updates:')
expect(rendered.html).not.toContain('in <strong>Designers</strong>')
expect(rendered.html).toContain('Open Shumai')
expect(rendered.html).not.toContain('Open Designers')
expect(rendered.html).not.toContain('Team: Designers')
expect(rendered.text).toContain('You have 2 new update(s):')
expect(rendered.text).toContain('Open Shumai:')
expect(rendered.text).not.toContain('Team: Designers')
expect(rendered.html).toContain('img1.png')
expect(rendered.html).toContain('img2.png')
// img1 has thumbnail, img2 does not
expect(rendered.html).toContain('https://example.com/img1.png')
expect(rendered.text).toContain('img1.png')
expect(rendered.text).toContain('img2.png')
})
it('collapses events into one summary row and omits thumbnails when count > 3', () => {
const items = Array.from({ length: 5 }, (_, i) => ({
id: `n-${i}`,
type: NotificationType.successful_file_uploaded,
creatorName: 'Alice',
creatorAvatarUrl: 'https://example.com/alice.png',
teamId: 'team-1',
teamName: 'Designers',
projectId: 'p-1',
projectName: 'Website Redesign',
assetId: `a-${i}`,
assetName: `photo_${i}.png`,
assetThumbnailUrl: `https://example.com/thumb_${i}.png`,
createdAt: new Date(`2026-09-27T10:0${i}:00Z`),
}))
const rendered = emailService.renderBatchedNotificationEmail({
teamId: 'team-1',
teamName: 'Designers',
recipientName: 'Bob',
items,
})
expect(rendered.subject).toBe('[Shumai] 5 new notifications')
expect(rendered.html).toContain('Alice uploaded 5 assets to Website Redesign')
// Thumbnails must be omitted for collapsed groups (> 3)
expect(rendered.html).not.toContain('https://example.com/thumb_')
expect(rendered.text).toContain('Alice uploaded 5 assets to Website Redesign')
})
it('batches mixed events in 1 email, collapsing groups > 3 while showing <= 3 in detail', () => {
// 4 comments by UserB in Project Alpha (> 3)
const commentItems = Array.from({ length: 4 }, (_, i) => ({
id: `c-${i}`,
type: NotificationType.comment_created,
creatorName: 'UserB',
teamId: 'team-1',
teamName: 'Designers',
projectId: 'p-1',
projectName: 'Project Alpha',
assetId: `a-${i}`,
assetName: `asset_${i}`,
commentMessage: `Comment number ${i}`,
createdAt: new Date(`2026-09-27T11:0${i}:00Z`),
}))
// 1 upload by UserC in Project Alpha (<= 3)
const uploadItem = {
id: 'u-1',
type: NotificationType.successful_file_uploaded,
creatorName: 'UserC',
teamId: 'team-1',
teamName: 'Designers',
projectId: 'p-1',
projectName: 'Project Alpha',
assetId: 'a-upload',
assetName: 'diagram.png',
assetThumbnailUrl: 'https://example.com/diagram.png',
createdAt: new Date('2026-09-27T11:05:00Z'),
}
const rendered = emailService.renderBatchedNotificationEmail({
teamId: 'team-1',
teamName: 'Designers',
recipientName: 'Alice',
items: [...commentItems, uploadItem],
})
expect(rendered.subject).toBe('[Shumai] 5 new notifications')
// Collapsed comment group
expect(rendered.html).toContain('UserB left 4 comments in Project Alpha')
// Detailed upload item with thumbnail
expect(rendered.html).toContain('UserC')
expect(rendered.html).toContain('diagram.png')
expect(rendered.html).toContain('https://example.com/diagram.png')
})
it('generates correct project action links in batched notification emails', () => {
const baseUrl = emailService.getBaseUrl()
const rendered = emailService.renderBatchedNotificationEmail({
teamId: '01KY2A7DRK2N0MECBDHTGAVV84',
teamName: 'Designers',
recipientName: 'Bob',
items: [
{
id: 'n-1',
type: NotificationType.successful_file_uploaded,
creatorName: 'Alice',
teamId: '01KY2A7DRK2N0MECBDHTGAVV84',
teamName: 'Designers',
projectId: '01KY2ANCA1SXGSPR63NVVD07EJ',
projectName: 'Project Alpha',
assetId: '01KY2ASSET1234567890ABCDEF',
assetName: 'hero.png',
createdAt: new Date('2026-09-27T10:00:00Z'),
},
],
})
expect(rendered.html).toContain(
`href="${baseUrl}/projects/01KY2ANCA1SXGSPR63NVVD07EJ/files/01KY2ASSET1234567890ABCDEF"`,
)
expect(rendered.html).not.toContain(
'/teams/01KY2A7DRK2N0MECBDHTGAVV84/projects/01KY2ANCA1SXGSPR63NVVD07EJ',
)
// Test collapsed group (>3 items)
const collapsedItems = Array.from({ length: 4 }, (_, i) => ({
id: `n-${i}`,
type: NotificationType.successful_file_uploaded,
creatorName: 'Alice',
teamId: '01KY2A7DRK2N0MECBDHTGAVV84',
teamName: 'Designers',
projectId: '01KY2ANCA1SXGSPR63NVVD07EJ',
projectName: 'Project Alpha',
assetId: `asset-${i}`,
assetName: `file_${i}.png`,
createdAt: new Date('2026-09-27T10:00:00Z'),
}))
const renderedCollapsed = emailService.renderBatchedNotificationEmail({
teamId: '01KY2A7DRK2N0MECBDHTGAVV84',
teamName: 'Designers',
recipientName: 'Bob',
items: collapsedItems,
})
expect(renderedCollapsed.html).toContain(
`href="${baseUrl}/projects/01KY2ANCA1SXGSPR63NVVD07EJ"`,
)
expect(renderedCollapsed.html).not.toContain(
'/teams/01KY2A7DRK2N0MECBDHTGAVV84/projects/01KY2ANCA1SXGSPR63NVVD07EJ',
)
})
})
})
+912
View File
@@ -0,0 +1,912 @@
import nodemailer, { type Transporter } from 'nodemailer'
import { HTTPException } from 'hono/http-exception'
import { prisma, NotificationType } from '@shumai/db'
import type { EmailNotificationSettings, UpdateEmailNotificationSettings } from '@shumai/dtos'
import { logger } from '@shumai/core/src/logger'
export interface EmailAttachment {
filename: string
content?: Buffer | string
path?: string
cid?: string
contentType?: string
}
export interface SendMailOptions {
to: string
subject: string
html: string
text: string
attachments?: EmailAttachment[]
}
export interface NotificationEmailContext {
type: NotificationType
creatorName: string
creatorAvatarUrl?: string
teamName: string
teamId: string
projectName?: string
projectId?: string
assetName?: string
assetId?: string
kanbanTaskTitle?: string
kanbanTaskId?: string
commentMessage?: string
mentionedUserNames?: Record<string, string>
fileCount?: number
fileNames?: string[]
uploadTime?: string | Date
}
export interface BatchedNotificationItem {
id: string
type: NotificationType
creatorName: string
creatorAvatarUrl?: string
teamId: string
teamName: string
projectId?: string
projectName?: string
assetId?: string
assetName?: string
assetThumbnailUrl?: string
kanbanTaskId?: string
kanbanTaskTitle?: string
commentMessage?: string
mentionedUserNames?: Record<string, string>
uploadTime?: string | Date
createdAt: Date
}
export interface BatchedEmailContext {
teamId: string
teamName: string
recipientName: string
items: BatchedNotificationItem[]
}
export function getNotificationBatchGroupKey(item: BatchedNotificationItem): string {
return `${item.type}_${item.creatorName || ''}_${item.projectId || ''}_${item.teamId}`
}
export const DEFAULT_EMAIL_SETTINGS: EmailNotificationSettings = {
enabled: false,
host: '',
port: 587,
username: '',
password: '',
smtps: false,
ignoreCert: false,
from: '',
replyTo: '',
}
export class EmailService {
getBaseUrl(): string {
const raw =
process.env.APP_URL ||
process.env.BETTER_AUTH_URL ||
`http://localhost:${process.env.SHUMAI_SERVER_PORT || '3000'}`
return raw.replace(/\/+$/, '')
}
createTransport(config: EmailNotificationSettings): Transporter {
return nodemailer.createTransport({
host: config.host,
port: config.port,
secure: config.smtps,
tls: {
rejectUnauthorized: !config.ignoreCert,
},
auth:
config.username || config.password
? {
user: config.username,
pass: config.password,
}
: undefined,
connectionTimeout: 10000,
})
}
async verifySmtp(config: EmailNotificationSettings): Promise<boolean> {
const transport = this.createTransport(config)
try {
await transport.verify()
return true
} finally {
transport.close()
}
}
async sendMail(
config: EmailNotificationSettings,
options: SendMailOptions,
): Promise<{ messageId: string }> {
const transport = this.createTransport(config)
try {
logger.debug(
{ to: options.to, from: config.from, subject: options.subject },
'Sending email via nodemailer',
)
const info = await transport.sendMail({
from: config.from,
replyTo: config.replyTo || config.from,
to: options.to,
subject: options.subject,
html: options.html,
text: options.text,
attachments: options.attachments,
})
logger.info({ messageId: info.messageId, to: options.to }, 'Email sent successfully')
return { messageId: info.messageId }
} finally {
transport.close()
}
}
async getEmailSettings(teamId: string): Promise<EmailNotificationSettings> {
const team = await prisma.team.findUnique({
where: { id: teamId },
select: { settings: true },
})
if (!team) {
throw new HTTPException(404, { message: 'Team not found' })
}
const saved = team.settings?.emailNotification
return {
enabled: saved?.enabled ?? DEFAULT_EMAIL_SETTINGS.enabled,
host: saved?.host ?? DEFAULT_EMAIL_SETTINGS.host,
port: saved?.port ?? DEFAULT_EMAIL_SETTINGS.port,
username: saved?.username ?? DEFAULT_EMAIL_SETTINGS.username,
password: saved?.password ?? DEFAULT_EMAIL_SETTINGS.password,
smtps: saved?.smtps ?? DEFAULT_EMAIL_SETTINGS.smtps,
ignoreCert: saved?.ignoreCert ?? DEFAULT_EMAIL_SETTINGS.ignoreCert,
from: saved?.from ?? DEFAULT_EMAIL_SETTINGS.from,
replyTo: saved?.replyTo ?? DEFAULT_EMAIL_SETTINGS.replyTo,
}
}
async updateEmailSettings(
teamId: string,
input: UpdateEmailNotificationSettings,
): Promise<EmailNotificationSettings> {
const team = await prisma.team.findUnique({
where: { id: teamId },
select: { settings: true },
})
if (!team) {
throw new HTTPException(404, { message: 'Team not found' })
}
const currentSettings = (team.settings || {}) as PrismaJson.Settings
const existingEmailConfig = currentSettings.emailNotification
// If password is not provided or empty, retain existing password
const password =
input.password !== undefined && input.password !== ''
? input.password
: existingEmailConfig?.password || ''
const updatedEmailConfig: PrismaJson.EmailNotificationSettings = {
enabled: input.enabled,
host: input.host,
port: input.port,
username: input.username || '',
password,
smtps: input.smtps,
ignoreCert: input.ignoreCert,
from: input.from,
replyTo: input.replyTo || '',
}
currentSettings.emailNotification = updatedEmailConfig
const updatedTeam = await prisma.team.update({
where: { id: teamId },
data: { settings: currentSettings as PrismaJson.Settings },
select: { settings: true },
})
return updatedTeam.settings?.emailNotification as EmailNotificationSettings
}
async sendTestEmail(
teamId: string,
currentUser: { id: string; email: string; name: string },
overrides?: Partial<EmailNotificationSettings>,
): Promise<{ success: boolean; messageId: string }> {
const saved = await this.getEmailSettings(teamId)
const effectiveConfig: EmailNotificationSettings = {
...saved,
...overrides,
// If override password is empty, fall back to saved
password:
overrides?.password !== undefined && overrides.password !== ''
? overrides.password
: saved.password,
}
if (!effectiveConfig.host) {
throw new HTTPException(400, { message: 'SMTP host is required' })
}
if (!effectiveConfig.from) {
throw new HTTPException(400, { message: 'From address is required' })
}
try {
await this.verifySmtp(effectiveConfig)
} catch (err) {
const errMsg = err instanceof Error ? err.message : String(err)
logger.warn({ teamId, err: errMsg }, 'SMTP verification failed during test email')
throw new HTTPException(400, { message: `SMTP verification failed: ${errMsg}` })
}
const baseUrl = this.getBaseUrl()
const subject = 'Test email from Shumai'
const html = `
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<title>Test Email from Shumai</title>
<style>
body { font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif; background-color: #f9fafb; margin: 0; padding: 24px; color: #111827; }
.container { max-width: 600px; margin: 0 auto; background: #ffffff; border-radius: 8px; border: 1px solid #e5e7eb; padding: 32px; }
.header { font-size: 20px; font-weight: 700; margin-bottom: 16px; color: #111827; }
.content { font-size: 14px; line-height: 1.6; color: #4b5563; margin-bottom: 24px; }
.badge { display: inline-block; background-color: #ecfdf5; color: #065f46; font-size: 12px; font-weight: 600; padding: 4px 10px; border-radius: 9999px; margin-bottom: 16px; }
.footer { font-size: 12px; color: #9ca3af; border-top: 1px solid #f3f4f6; padding-top: 16px; margin-top: 24px; }
</style>
</head>
<body>
<div class="container">
<div class="header">Shumai Notification System</div>
<div class="badge">SMTP Verified</div>
<div class="content">
<p>Hello ${escapeHtml(currentUser.name || 'Admin')},</p>
<p>This is a test email confirming that your SMTP server settings are correctly configured for team notifications in Shumai.</p>
<p>Notifications for new comments, mentions, uploads, and Kanban tasks will now be delivered via this email service.</p>
</div>
<div class="footer">
Sent from Shumai at <a href="${baseUrl}" style="color: #6366f1; text-decoration: none;">${baseUrl}</a>
</div>
</div>
</body>
</html>
`.trim()
const text = `
Shumai Notification System - Test Email
Hello ${currentUser.name || 'Admin'},
This is a test email confirming that your SMTP server settings are correctly configured for team notifications in Shumai.
Sent from Shumai at ${baseUrl}
`.trim()
try {
const res = await this.sendMail(effectiveConfig, {
to: currentUser.email,
subject,
html,
text,
})
return { success: true, messageId: res.messageId }
} catch (err) {
const errMsg = err instanceof Error ? err.message : String(err)
logger.error({ teamId, err: errMsg }, 'Failed to send test email')
throw new HTTPException(400, { message: `Failed to send test email: ${errMsg}` })
}
}
renderNotificationEmail(ctx: NotificationEmailContext): {
subject: string
html: string
text: string
} {
const baseUrl = this.getBaseUrl()
let actionUrl = baseUrl
let actionText = 'View in Shumai'
if (ctx.projectId && ctx.assetId) {
actionUrl = `${baseUrl}/projects/${ctx.projectId}/files/${ctx.assetId}`
actionText = 'View Asset'
} else if (ctx.projectId) {
actionUrl = `${baseUrl}/projects/${ctx.projectId}`
actionText = 'View Project'
} else if (ctx.teamId && ctx.kanbanTaskId) {
actionUrl = `${baseUrl}/teams/${ctx.teamId}/kanban?taskId=${ctx.kanbanTaskId}`
actionText = 'View Task'
} else if (ctx.teamId) {
actionUrl = `${baseUrl}/teams/${ctx.teamId}`
actionText = 'Open Shumai'
}
let subject = '[Shumai] Notification'
let headline = 'You have a new notification'
let detail = ''
const actor = ctx.creatorName || 'Someone'
const actorInitial = (actor.trim()[0] || 'U').toUpperCase()
const formattedUploadTime = formatUploadTime(ctx.uploadTime)
let commentMessage = ctx.commentMessage
if (commentMessage) {
commentMessage = commentMessage.replace(/<@([^>]+)>/g, (match, uid) => {
const name = ctx.mentionedUserNames?.[uid]
return name ? `@${name}` : `@${uid}`
})
}
switch (ctx.type) {
case NotificationType.comment_created: {
const target = ctx.assetName ? `"${ctx.assetName}"` : 'an asset'
subject = `[Shumai] ${actor} commented on ${target}`
headline = `${actor} left a comment`
detail = commentMessage ? `"${commentMessage}"` : `New comment on ${target}`
break
}
case NotificationType.reply_created: {
const target = ctx.assetName ? `"${ctx.assetName}"` : 'an asset'
subject = `[Shumai] ${actor} replied to your comment on ${target}`
headline = `${actor} replied to your comment`
detail = commentMessage ? `"${commentMessage}"` : `Reply on ${target}`
break
}
case NotificationType.mention: {
subject = `[Shumai] ${actor} mentioned you`
headline = `${actor} mentioned you`
detail = commentMessage ? `"${commentMessage}"` : 'You were mentioned in a comment'
break
}
case NotificationType.successful_file_uploaded: {
const target = ctx.assetName ? `"${ctx.assetName}"` : 'a file'
const projectInfo = ctx.projectName ? ` to ${ctx.projectName}` : ''
if (ctx.fileCount && ctx.fileCount > 1) {
subject = `[Shumai] ${actor} uploaded ${ctx.fileCount} files${projectInfo}`
headline = `${actor} uploaded ${ctx.fileCount} new files${projectInfo}`
let filesSummary: string
if (ctx.fileNames && ctx.fileNames.length > 0) {
const previewList = ctx.fileNames.slice(0, 3).join(', ')
const remaining = ctx.fileNames.length - 3
filesSummary = remaining > 0 ? `${previewList}, and ${remaining} more` : previewList
} else {
filesSummary = `${ctx.fileCount} new files were uploaded`
}
const details: string[] = []
if (ctx.projectName) details.push(`Project: ${ctx.projectName}`)
if (formattedUploadTime) details.push(`Uploaded at: ${formattedUploadTime}`)
details.push(`Files: ${filesSummary}`)
detail = details.join('\n')
} else {
subject = `[Shumai] ${actor} uploaded ${target}${projectInfo}`
headline = `${actor} uploaded ${target}${projectInfo}`
const details: string[] = []
if (ctx.assetName) details.push(`Asset: ${ctx.assetName}`)
if (ctx.projectName) details.push(`Project: ${ctx.projectName}`)
if (formattedUploadTime) details.push(`Uploaded at: ${formattedUploadTime}`)
detail = details.join('\n')
}
break
}
case NotificationType.metadata_field_updated_status: {
const target = ctx.assetName ? `"${ctx.assetName}"` : 'an asset'
subject = `[Shumai] ${actor} updated status of ${target}`
headline = `${actor} updated status of ${target}`
detail = `Status update on ${target}`
break
}
case NotificationType.new_user_join_team: {
subject = `[Shumai] ${actor} joined ${ctx.teamName}`
headline = `${actor} joined the team`
detail = `${actor} is now a member of ${ctx.teamName}`
break
}
case NotificationType.new_user_join_project: {
const project = ctx.projectName || 'a project'
subject = `[Shumai] ${actor} joined project ${project}`
headline = `${actor} joined project ${project}`
detail = `${actor} is now a member of ${project}`
break
}
case NotificationType.kanban_task_created: {
const task = ctx.kanbanTaskTitle ? `"${ctx.kanbanTaskTitle}"` : 'a task'
subject = `[Shumai] ${actor} created task ${task}`
headline = `${actor} created task ${task}`
detail = commentMessage || `Task ${task} was created`
break
}
case NotificationType.kanban_task_assigned: {
const task = ctx.kanbanTaskTitle ? `"${ctx.kanbanTaskTitle}"` : 'a task'
subject = `[Shumai] ${actor} assigned you to task ${task}`
headline = `${actor} assigned you to task ${task}`
detail = `You have been assigned to ${task}`
break
}
case NotificationType.kanban_task_status_updated: {
const task = ctx.kanbanTaskTitle ? `"${ctx.kanbanTaskTitle}"` : 'a task'
subject = `[Shumai] ${actor} updated status of task ${task}`
headline = `${actor} updated status of ${task}`
detail = commentMessage || `Status changed for ${task}`
break
}
case NotificationType.kanban_task_updated: {
const task = ctx.kanbanTaskTitle ? `"${ctx.kanbanTaskTitle}"` : 'a task'
subject = `[Shumai] ${actor} updated task ${task}`
headline = `${actor} updated task ${task}`
detail = commentMessage || `Task ${task} was updated`
break
}
case NotificationType.kanban_task_deleted: {
const task = ctx.kanbanTaskTitle ? `"${ctx.kanbanTaskTitle}"` : 'a task'
subject = `[Shumai] ${actor} deleted task ${task}`
headline = `${actor} deleted task ${task}`
detail = `Task ${task} was deleted`
break
}
case NotificationType.kanban_task_comment_created: {
const task = ctx.kanbanTaskTitle ? `"${ctx.kanbanTaskTitle}"` : 'a task'
subject = `[Shumai] ${actor} commented on task ${task}`
headline = `${actor} commented on task ${task}`
detail = commentMessage ? `"${commentMessage}"` : `New comment on ${task}`
break
}
}
const avatarHtml = ctx.creatorAvatarUrl
? `<img src="${escapeHtml(ctx.creatorAvatarUrl)}" alt="${escapeHtml(actor)}" width="36" height="36" style="width: 36px; height: 36px; border-radius: 50%; object-fit: cover; display: block; border: 1px solid #e5e7eb;" />`
: `<table cellpadding="0" cellspacing="0" border="0" style="width: 36px; height: 36px; border-radius: 50%; background-color: #4f46e5; text-align: center;"><tr><td style="color: #ffffff; font-weight: 600; font-size: 14px; text-align: center; vertical-align: middle;">${escapeHtml(actorInitial)}</td></tr></table>`
const html = `
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<title>${escapeHtml(subject)}</title>
<style>
body { font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif; background-color: #f9fafb; margin: 0; padding: 24px; color: #111827; }
.container { max-width: 600px; margin: 0 auto; background: #ffffff; border-radius: 8px; border: 1px solid #e5e7eb; padding: 32px; }
.header { font-size: 18px; font-weight: 700; margin-bottom: 8px; color: #111827; }
.content { font-size: 14px; line-height: 1.6; color: #374151; margin-bottom: 24px; }
.detail-box { background-color: #f3f4f6; border-left: 4px solid #4f46e5; padding: 12px 16px; border-radius: 4px; margin: 16px 0; font-size: 14px; color: #1f2937; white-space: pre-wrap; }
.btn { display: inline-block; background-color: #4f46e5; color: #ffffff !important; padding: 10px 20px; border-radius: 6px; text-decoration: none; font-size: 14px; font-weight: 500; margin-top: 12px; }
.footer { font-size: 12px; color: #9ca3af; border-top: 1px solid #f3f4f6; padding-top: 16px; margin-top: 24px; }
</style>
</head>
<body>
<div class="container">
<table cellpadding="0" cellspacing="0" border="0" style="margin-bottom: 16px;">
<tr>
<td style="vertical-align: middle; padding-right: 12px; width: 36px;">
${avatarHtml}
</td>
<td style="vertical-align: middle;">
<div style="font-weight: 600; font-size: 15px; color: #111827; line-height: 1.2;">${escapeHtml(actor)}</div>
${formattedUploadTime ? `<div style="font-size: 12px; color: #6b7280; margin-top: 2px;">Uploaded at ${escapeHtml(formattedUploadTime)}</div>` : ''}
</td>
</tr>
</table>
<div class="header">${escapeHtml(headline)}</div>
<div class="content">
${detail ? `<div class="detail-box">${escapeHtml(detail)}</div>` : ''}
<a href="${actionUrl}" class="btn">${escapeHtml(actionText)}</a>
</div>
<div class="footer">
Sent from <a href="${baseUrl}" style="color: #6366f1; text-decoration: none;">Shumai</a>
</div>
</div>
</body>
</html>
`.trim()
const text = `
${subject}
${headline}
${formattedUploadTime ? `Uploaded at: ${formattedUploadTime}\n` : ''}${detail ? `\n${detail}\n` : ''}
${actionText}: ${actionUrl}
Sent from Shumai at ${baseUrl}
`.trim()
return { subject, html, text }
}
renderBatchedNotificationEmail(ctx: BatchedEmailContext): {
subject: string
html: string
text: string
} {
const baseUrl = this.getBaseUrl()
const teamUrl = `${baseUrl}/teams/${ctx.teamId}`
if (ctx.items.length === 0) {
return {
subject: `[Shumai] Notifications`,
html: `<p>No new notifications.</p>`,
text: `No new notifications.`,
}
}
// Determine subject
let subject = `[Shumai] ${ctx.items.length} new notifications`
if (ctx.items.length === 1) {
const single = ctx.items[0]
const actor = single.creatorName || 'Someone'
const target = single.assetName ? `"${single.assetName}"` : 'a file'
const projectInfo = single.projectName ? ` to ${single.projectName}` : ''
switch (single.type) {
case NotificationType.successful_file_uploaded:
subject = `[Shumai] ${actor} uploaded ${target}${projectInfo}`
break
case NotificationType.comment_created:
subject = `[Shumai] ${actor} commented on ${target}`
break
case NotificationType.reply_created:
subject = `[Shumai] ${actor} replied to your comment on ${target}`
break
case NotificationType.mention:
subject = `[Shumai] ${actor} mentioned you`
break
case NotificationType.kanban_task_assigned: {
const task = single.kanbanTaskTitle ? `"${single.kanbanTaskTitle}"` : 'a task'
subject = `[Shumai] ${actor} assigned you to task ${task}`
break
}
case NotificationType.kanban_task_created: {
const task = single.kanbanTaskTitle ? `"${single.kanbanTaskTitle}"` : 'a task'
subject = `[Shumai] ${actor} created task ${task}`
break
}
default:
subject = `[Shumai] 1 new notification`
break
}
}
// Group items: same type + actor + project
const groups = new Map<string, BatchedNotificationItem[]>()
for (const item of ctx.items) {
const key = getNotificationBatchGroupKey(item)
const list = groups.get(key)
if (list) {
list.push(item)
} else {
groups.set(key, [item])
}
}
const htmlRows: string[] = []
const textRows: string[] = []
for (const [, cluster] of groups) {
const first = cluster[0]
const actor = first.creatorName || 'Someone'
const avatarHtml = renderAvatarHtml(actor, first.creatorAvatarUrl)
if (cluster.length <= 3) {
// Render detailed individual rows
for (const item of cluster) {
const action = getItemAction(baseUrl, item)
const timeStr = formatUploadTime(item.uploadTime || item.createdAt)
const detail = getNotificationItemDetail(item)
const thumbHtml = item.assetThumbnailUrl
? `<div style="margin-top: 8px;"><img src="${escapeHtml(item.assetThumbnailUrl)}" alt="${escapeHtml(item.assetName || 'Thumbnail')}" width="140" style="max-width: 140px; max-height: 90px; border-radius: 6px; border: 1px solid #e5e7eb; object-fit: cover; display: block;" /></div>`
: ''
htmlRows.push(
`
<div style="padding: 16px 0; border-bottom: 1px solid #f3f4f6;">
<table cellpadding="0" cellspacing="0" border="0" style="width: 100%;">
<tr>
<td style="vertical-align: top; width: 44px; padding-right: 12px;">
${avatarHtml}
</td>
<td style="vertical-align: top;">
<div style="font-weight: 600; font-size: 14px; color: #111827;">${escapeHtml(actor)}</div>
<div style="font-size: 13px; color: #374151; margin-top: 2px;">${escapeHtml(detail.headline)}</div>
${timeStr ? `<div style="font-size: 12px; color: #6b7280; margin-top: 2px;">${escapeHtml(timeStr)}</div>` : ''}
${detail.message ? `<div style="background-color: #f9fafb; border-left: 3px solid #6366f1; padding: 8px 12px; border-radius: 4px; font-size: 13px; color: #4b5563; margin-top: 6px; white-space: pre-wrap;">${escapeHtml(detail.message)}</div>` : ''}
${thumbHtml}
<div style="margin-top: 8px;">
<a href="${action.url}" style="display: inline-block; font-size: 12px; font-weight: 500; color: #4f46e5; text-decoration: none;">${escapeHtml(action.text)} &rarr;</a>
</div>
</td>
</tr>
</table>
</div>
`.trim(),
)
textRows.push(
`- ${actor}: ${detail.headline}${timeStr ? ` (${timeStr})` : ''}${detail.message ? `\n "${detail.message}"` : ''}\n Link: ${action.url}`,
)
}
} else {
// Grouped summary row (count > 3, NO thumbnails)
const count = cluster.length
let summaryHeadline = `${actor} triggered ${count} updates`
const projectName = first.projectName
switch (first.type) {
case NotificationType.successful_file_uploaded:
summaryHeadline = `${actor} uploaded ${count} assets${projectName ? ` to ${projectName}` : ''}`
break
case NotificationType.comment_created:
summaryHeadline = `${actor} left ${count} comments${projectName ? ` in ${projectName}` : ''}`
break
case NotificationType.reply_created:
summaryHeadline = `${actor} replied ${count} times to your comments`
break
case NotificationType.mention:
summaryHeadline = `${actor} mentioned you ${count} times`
break
case NotificationType.metadata_field_updated_status:
summaryHeadline = `${actor} updated status of ${count} assets`
break
case NotificationType.kanban_task_created:
summaryHeadline = `${actor} created ${count} tasks${projectName ? ` in ${projectName}` : ''}`
break
case NotificationType.kanban_task_assigned:
summaryHeadline = `${actor} assigned you to ${count} tasks`
break
case NotificationType.kanban_task_status_updated:
summaryHeadline = `${actor} updated status of ${count} tasks`
break
case NotificationType.kanban_task_updated:
summaryHeadline = `${actor} updated ${count} tasks`
break
case NotificationType.kanban_task_comment_created:
summaryHeadline = `${actor} commented on ${count} tasks`
break
}
const latestTime = cluster.reduce((latest, it) => {
const t = new Date(it.uploadTime || it.createdAt).getTime()
return t > latest ? t : latest
}, 0)
const timeStr = latestTime > 0 ? formatUploadTime(new Date(latestTime)) : undefined
const targetUrl = first.projectId ? `${baseUrl}/projects/${first.projectId}` : teamUrl
htmlRows.push(
`
<div style="padding: 16px 0; border-bottom: 1px solid #f3f4f6;">
<table cellpadding="0" cellspacing="0" border="0" style="width: 100%;">
<tr>
<td style="vertical-align: top; width: 44px; padding-right: 12px;">
${avatarHtml}
</td>
<td style="vertical-align: top;">
<div style="font-weight: 600; font-size: 14px; color: #111827;">${escapeHtml(summaryHeadline)}</div>
${timeStr ? `<div style="font-size: 12px; color: #6b7280; margin-top: 2px;">Latest activity: ${escapeHtml(timeStr)}</div>` : ''}
<div style="margin-top: 8px;">
<a href="${targetUrl}" style="display: inline-block; font-size: 12px; font-weight: 500; color: #4f46e5; text-decoration: none;">View Project &rarr;</a>
</div>
</td>
</tr>
</table>
</div>
`.trim(),
)
textRows.push(
`- ${summaryHeadline}${timeStr ? ` (Latest: ${timeStr})` : ''}\n Link: ${targetUrl}`,
)
}
}
const html = `
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<title>${escapeHtml(subject)}</title>
<style>
body { font-family: -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, Helvetica, Arial, sans-serif; background-color: #f9fafb; margin: 0; padding: 24px; color: #111827; }
.container { max-width: 600px; margin: 0 auto; background: #ffffff; border-radius: 8px; border: 1px solid #e5e7eb; padding: 32px; }
.header { font-size: 18px; font-weight: 700; color: #111827; }
.subtitle { font-size: 14px; color: #6b7280; margin-top: 4px; margin-bottom: 20px; }
.btn { display: inline-block; background-color: #4f46e5; color: #ffffff !important; padding: 10px 20px; border-radius: 6px; text-decoration: none; font-size: 14px; font-weight: 500; margin-top: 20px; }
.footer { font-size: 12px; color: #9ca3af; border-top: 1px solid #f3f4f6; padding-top: 16px; margin-top: 24px; }
</style>
</head>
<body>
<div class="container">
<div class="header">Shumai Notifications</div>
<div class="subtitle">You have ${ctx.items.length} new update${ctx.items.length > 1 ? 's' : ''}:</div>
<div>
${htmlRows.join('\n')}
</div>
<div style="text-align: center; margin-top: 20px;">
<a href="${teamUrl}" class="btn">Open Shumai</a>
</div>
<div class="footer">
Sent from <a href="${baseUrl}" style="color: #6366f1; text-decoration: none;">Shumai</a>
</div>
</div>
</body>
</html>
`.trim()
const text = `
Shumai Notifications
You have ${ctx.items.length} new update(s):
${textRows.join('\n\n')}
Open Shumai: ${teamUrl}
Sent from Shumai at ${baseUrl}
`.trim()
return { subject, html, text }
}
}
function getNotificationItemDetail(item: BatchedNotificationItem): {
headline: string
message?: string
} {
const actor = item.creatorName || 'Someone'
const asset = item.assetName ? `"${item.assetName}"` : 'an asset'
const task = item.kanbanTaskTitle ? `"${item.kanbanTaskTitle}"` : 'a task'
const project = item.projectName ? ` in ${item.projectName}` : ''
let commentMessage = item.commentMessage
if (commentMessage && item.mentionedUserNames) {
commentMessage = commentMessage.replace(/<@([^>]+)>/g, (match, uid) => {
const name = item.mentionedUserNames?.[uid]
return name ? `@${name}` : `@${uid}`
})
}
switch (item.type) {
case NotificationType.successful_file_uploaded:
return {
headline: `Uploaded ${asset}${item.projectName ? ` to ${item.projectName}` : ''}`,
}
case NotificationType.comment_created:
return {
headline: `Commented on ${asset}${project}`,
message: commentMessage,
}
case NotificationType.reply_created:
return {
headline: `Replied to your comment on ${asset}${project}`,
message: commentMessage,
}
case NotificationType.mention:
return {
headline: `Mentioned you on ${item.kanbanTaskTitle ? `task ${task}` : asset}${project}`,
message: commentMessage,
}
case NotificationType.metadata_field_updated_status:
return {
headline: `Updated status of ${asset}${project}`,
}
case NotificationType.kanban_task_created:
return {
headline: `Created task ${task}${project}`,
message: commentMessage,
}
case NotificationType.kanban_task_assigned:
return {
headline: `Assigned you to task ${task}${project}`,
}
case NotificationType.kanban_task_status_updated:
return {
headline: `Updated status of task ${task}${project}`,
message: commentMessage,
}
case NotificationType.kanban_task_updated:
return {
headline: `Updated task ${task}${project}`,
message: commentMessage,
}
case NotificationType.kanban_task_deleted:
return {
headline: `Deleted task ${task}${project}`,
}
case NotificationType.kanban_task_comment_created:
return {
headline: `Commented on task ${task}${project}`,
message: commentMessage,
}
case NotificationType.new_user_join_team:
return {
headline: `Joined the team ${item.teamName}`,
}
case NotificationType.new_user_join_project:
return {
headline: `Joined project ${item.projectName || 'a project'}`,
}
default:
return {
headline: `New activity from ${actor}`,
}
}
}
function getItemAction(
baseUrl: string,
item: BatchedNotificationItem,
): { url: string; text: string } {
if (item.projectId && item.assetId) {
return {
url: `${baseUrl}/projects/${item.projectId}/files/${item.assetId}`,
text: 'View Asset',
}
} else if (item.projectId) {
return {
url: `${baseUrl}/projects/${item.projectId}`,
text: 'View Project',
}
} else if (item.teamId && item.kanbanTaskId) {
return {
url: `${baseUrl}/teams/${item.teamId}/kanban?taskId=${item.kanbanTaskId}`,
text: 'View Task',
}
} else if (item.teamId) {
return {
url: `${baseUrl}/teams/${item.teamId}`,
text: 'Open Shumai',
}
}
return { url: baseUrl, text: 'Open Shumai' }
}
function renderAvatarHtml(actor: string, avatarUrl?: string): string {
const actorInitial = (actor.trim()[0] || 'U').toUpperCase()
if (avatarUrl) {
return `<img src="${escapeHtml(avatarUrl)}" alt="${escapeHtml(actor)}" width="36" height="36" style="width: 36px; height: 36px; border-radius: 50%; object-fit: cover; display: block; border: 1px solid #e5e7eb;" />`
}
return `<table cellpadding="0" cellspacing="0" border="0" style="width: 36px; height: 36px; border-radius: 50%; background-color: #4f46e5; text-align: center;"><tr><td style="color: #ffffff; font-weight: 600; font-size: 14px; text-align: center; vertical-align: middle;">${escapeHtml(actorInitial)}</td></tr></table>`
}
function formatUploadTime(time: string | Date | undefined): string | undefined {
if (!time) return undefined
if (time instanceof Date) {
if (isNaN(time.getTime())) return undefined
return (
new Intl.DateTimeFormat('en-US', {
month: 'short',
day: 'numeric',
year: 'numeric',
hour: 'numeric',
minute: '2-digit',
hour12: true,
timeZone: 'UTC',
}).format(time) + ' UTC'
)
}
const parsed = new Date(time)
if (!isNaN(parsed.getTime())) {
return (
new Intl.DateTimeFormat('en-US', {
month: 'short',
day: 'numeric',
year: 'numeric',
hour: 'numeric',
minute: '2-digit',
hour12: true,
timeZone: 'UTC',
}).format(parsed) + ' UTC'
)
}
return time
}
function escapeHtml(str: string): string {
return str
.replace(/&/g, '&amp;')
.replace(/</g, '&lt;')
.replace(/>/g, '&gt;')
.replace(/"/g, '&quot;')
.replace(/'/g, '&#039;')
}
export const emailService = new EmailService()
+3
View File
@@ -0,0 +1,3 @@
export * from './notification'
export * from './email'
export * from './notification-job'
@@ -0,0 +1,576 @@
import { describe, expect, it, vi, beforeEach, afterEach } from 'vitest'
import { prisma, NotificationType, NotificationEmailStatus } from '@shumai/db'
import { setupTestDbHooks } from '@shumai/db/test'
import { emailService, type EmailAttachment } from './email'
import { notificationJobService } from './notification-job'
import { userMetadataService } from '@shumai/core/src/user-metadata/user-metadata'
import { s3Service } from '@shumai/core/src/s3/s3'
import '@shumai/db/src/prisma-json-types'
describe('NotificationJobService', () => {
setupTestDbHooks()
let sendMailSpy: ReturnType<typeof vi.spyOn>
beforeEach(() => {
sendMailSpy = vi
.spyOn(emailService, 'sendMail')
.mockResolvedValue({ messageId: '<mock-id@shumai.test>' })
})
afterEach(() => {
sendMailSpy.mockRestore()
notificationJobService.stop()
})
async function createTeamWithEmail(name = 'Email Team') {
return prisma.team.create({
data: {
name,
settings: {
emailNotification: {
enabled: true,
host: 'smtp.test.com',
port: 587,
username: 'user',
password: 'pw',
smtps: false,
ignoreCert: false,
from: 'noreply@test.com',
},
},
},
})
}
it('picks up pending notifications, sends batched email, and marks them processed', async () => {
const team = await createTeamWithEmail()
const sender = await prisma.user.create({
data: { name: 'Alice', email: 'alice@test.com', password: 'p' },
})
const recipient = await prisma.user.create({
data: { name: 'Bob', email: 'bob@test.com', password: 'p' },
})
await prisma.teamMember.create({
data: { teamId: team.id, userId: recipient.id, role: 'editor' },
})
const project = await prisma.project.create({
data: { name: 'Alpha Project', teamId: team.id },
})
const asset = await prisma.asset.create({
data: {
name: 'design.png',
type: 'file',
status: 'uploaded',
project: { connect: { id: project.id } },
},
})
// Create 2 pending notifications
const n1 = await prisma.notification.create({
data: {
type: NotificationType.successful_file_uploaded,
teamId: team.id,
projectId: project.id,
assetId: asset.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.pending,
},
})
const n2 = await prisma.notification.create({
data: {
type: NotificationType.comment_created,
teamId: team.id,
projectId: project.id,
assetId: asset.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.pending,
},
})
const processedCount = await notificationJobService.processPendingNotifications(500)
expect(processedCount).toBe(2)
// Verify Bob received exactly 1 batched email containing both items
expect(sendMailSpy).toHaveBeenCalledTimes(1)
expect(sendMailSpy).toHaveBeenCalledWith(
expect.objectContaining({ host: 'smtp.test.com' }),
expect.objectContaining({
to: 'bob@test.com',
subject: '[Shumai] 2 new notifications',
}),
)
// Verify notifications are marked processed in DB
const updated1 = await prisma.notification.findUnique({ where: { id: n1.id } })
const updated2 = await prisma.notification.findUnique({ where: { id: n2.id } })
expect(updated1?.emailStatus).toBe(NotificationEmailStatus.processed)
expect(updated2?.emailStatus).toBe(NotificationEmailStatus.processed)
})
it('skips already processed notifications', async () => {
const team = await createTeamWithEmail()
const sender = await prisma.user.create({
data: { name: 'Alice', email: 'alice2@test.com', password: 'p' },
})
await prisma.notification.create({
data: {
type: NotificationType.comment_created,
teamId: team.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.processed,
},
})
const processedCount = await notificationJobService.processPendingNotifications(500)
expect(processedCount).toBe(0)
expect(sendMailSpy).not.toHaveBeenCalled()
})
it('marks notifications processed without sending email if team SMTP is disabled', async () => {
// Team without emailNotification configured
const team = await prisma.team.create({
data: { name: 'No Email Team' },
})
const sender = await prisma.user.create({
data: { name: 'Alice', email: 'alice3@test.com', password: 'p' },
})
const notif = await prisma.notification.create({
data: {
type: NotificationType.new_user_join_team,
teamId: team.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.pending,
},
})
const processedCount = await notificationJobService.processPendingNotifications(500)
expect(processedCount).toBe(1)
expect(sendMailSpy).not.toHaveBeenCalled()
const updated = await prisma.notification.findUnique({ where: { id: notif.id } })
expect(updated?.emailStatus).toBe(NotificationEmailStatus.processed)
})
it('excludes the actor from receiving notification about their own action', async () => {
const team = await createTeamWithEmail()
const alice = await prisma.user.create({
data: { name: 'Alice', email: 'alice-self@test.com', password: 'p' },
})
await prisma.teamMember.create({
data: { teamId: team.id, userId: alice.id, role: 'editor' },
})
// Alice is the creator and only member
await prisma.notification.create({
data: {
type: NotificationType.comment_created,
teamId: team.id,
creatorId: alice.id,
emailStatus: NotificationEmailStatus.pending,
},
})
await notificationJobService.processPendingNotifications(500)
// Alice should not receive an email for her own comment
expect(sendMailSpy).not.toHaveBeenCalled()
})
it('excludes AI agent users from receiving email notifications', async () => {
const team = await createTeamWithEmail()
const sender = await prisma.user.create({
data: { name: 'Alice', email: 'alice-agent-test@test.com', password: 'p' },
})
const agentUser = await prisma.user.create({
data: {
name: 'AutoAgent',
email: 'agent-bot@shumai.ai',
password: 'p',
type: 'agent',
},
})
await prisma.teamMember.create({
data: { teamId: team.id, userId: agentUser.id, role: 'editor' },
})
await prisma.notification.create({
data: {
type: NotificationType.new_user_join_team,
teamId: team.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.pending,
},
})
await notificationJobService.processPendingNotifications(500)
expect(sendMailSpy).not.toHaveBeenCalled()
})
it('respects user notification preferences', async () => {
const team = await createTeamWithEmail()
const sender = await prisma.user.create({
data: { name: 'Alice', email: 'alice-pref@test.com', password: 'p' },
})
const bob = await prisma.user.create({
data: { name: 'Bob', email: 'bob-pref@test.com', password: 'p' },
})
await prisma.teamMember.create({
data: { teamId: team.id, userId: bob.id, role: 'editor' },
})
// Bob disabled comments in his notification settings
await userMetadataService.upsertMetadata(bob.id, team.id, 'notification_settings', {
comments: false,
replies: true,
mentions: true,
yourUploads: false,
otherUploads: true,
statusUpdates: true,
kanbanTasks: true,
kanbanComments: true,
})
await prisma.notification.create({
data: {
type: NotificationType.comment_created,
teamId: team.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.pending,
},
})
await notificationJobService.processPendingNotifications(500)
expect(sendMailSpy).not.toHaveBeenCalled()
})
it('filters notifications by project scope for project-scoped members', async () => {
const team = await createTeamWithEmail()
const sender = await prisma.user.create({
data: { name: 'Alice', email: 'alice-scope@test.com', password: 'p' },
})
const project1 = await prisma.project.create({
data: { name: 'Project 1', teamId: team.id },
})
const project2 = await prisma.project.create({
data: { name: 'Project 2', teamId: team.id },
})
const bob = await prisma.user.create({
data: { name: 'Bob', email: 'bob-scope@test.com', password: 'p' },
})
const bobMember = await prisma.teamMember.create({
data: { teamId: team.id, userId: bob.id, role: 'editor', scope: 'project' },
})
// Bob only belongs to Project 1
await prisma.projectMember.create({
data: { projectId: project1.id, teamMemberId: bobMember.id, role: 'editor' },
})
// Notification in Project 2
await prisma.notification.create({
data: {
type: NotificationType.successful_file_uploaded,
teamId: team.id,
projectId: project2.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.pending,
},
})
await notificationJobService.processPendingNotifications(500)
// Bob should not receive notification for Project 2
expect(sendMailSpy).not.toHaveBeenCalled()
})
it('batches 5 uploads and 1 comment into 1 email with collapsed upload summary', async () => {
const team = await createTeamWithEmail()
const sender = await prisma.user.create({
data: { name: 'UserB', email: 'userb@test.com', password: 'p' },
})
const recipient = await prisma.user.create({
data: { name: 'Recipient', email: 'recip@test.com', password: 'p' },
})
await prisma.teamMember.create({
data: { teamId: team.id, userId: recipient.id, role: 'editor' },
})
const project = await prisma.project.create({
data: { name: 'Project Alpha', teamId: team.id },
})
// 5 uploads by UserB in Project Alpha
for (let i = 0; i < 5; i++) {
const asset = await prisma.asset.create({
data: {
name: `img_${i}.png`,
type: 'file',
status: 'uploaded',
project: { connect: { id: project.id } },
},
})
await prisma.notification.create({
data: {
type: NotificationType.successful_file_uploaded,
teamId: team.id,
projectId: project.id,
assetId: asset.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.pending,
},
})
}
// 1 comment by UserB
await prisma.notification.create({
data: {
type: NotificationType.comment_created,
teamId: team.id,
projectId: project.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.pending,
},
})
const count = await notificationJobService.processPendingNotifications(500)
expect(count).toBe(6)
// Recipient receives exactly 1 email containing all 6 events
expect(sendMailSpy).toHaveBeenCalledTimes(1)
const call = sendMailSpy.mock.calls[0]
expect(call[1].to).toBe('recip@test.com')
expect(call[1].subject).toBe('[Shumai] 6 new notifications')
// Uploads collapsed into 1 row because count > 3
expect(call[1].html).toContain('UserB uploaded 5 assets to Project Alpha')
// Comment rendered as individual row
expect(call[1].html).toContain('Commented on')
})
it('embeds creator avatar and asset thumbnail as inline cid attachments', async () => {
const team = await createTeamWithEmail()
const sender = await prisma.user.create({
data: {
name: 'Alice Artist',
email: 'alice.artist@test.com',
password: 'p',
image: 'avatars/alice.png',
},
})
const recipient = await prisma.user.create({
data: { name: 'Bob Reviewer', email: 'bob.reviewer@test.com', password: 'p' },
})
await prisma.teamMember.create({
data: { teamId: team.id, userId: recipient.id, role: 'editor' },
})
const project = await prisma.project.create({
data: { name: 'Artwork Project', teamId: team.id },
})
const asset = await prisma.asset.create({
data: {
name: 'poster.png',
type: 'file',
status: 'uploaded',
project: { connect: { id: project.id } },
media: {
thumbnail: { key: 'thumbnails/poster.jpg' },
} as unknown as PrismaJson.MediaInfo,
},
})
await prisma.notification.create({
data: {
type: NotificationType.successful_file_uploaded,
teamId: team.id,
projectId: project.id,
assetId: asset.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.pending,
},
})
const getObjectSpy = vi
.spyOn(s3Service, 'getObject')
.mockImplementation(async (_bucket, key) => {
if (key === 'avatars/alice.png') {
return { buffer: Buffer.from('alice-avatar-bytes'), contentType: 'image/png' }
}
if (key === 'thumbnails/poster.jpg') {
return { buffer: Buffer.from('poster-thumb-bytes'), contentType: 'image/jpeg' }
}
throw new Error('Not found')
})
try {
const count = await notificationJobService.processPendingNotifications(500)
expect(count).toBe(1)
expect(sendMailSpy).toHaveBeenCalledTimes(1)
const call = sendMailSpy.mock.calls[0]
const options = call[1]
expect(options.to).toBe('bob.reviewer@test.com')
// Check attachments
expect(options.attachments).toHaveLength(2)
const avatarAtt = options.attachments.find((a: EmailAttachment) =>
a.cid?.startsWith('avatar-'),
)
const thumbAtt = options.attachments.find((a: EmailAttachment) => a.cid?.startsWith('thumb-'))
expect(avatarAtt).toBeDefined()
expect(avatarAtt.content).toEqual(Buffer.from('alice-avatar-bytes'))
expect(avatarAtt.cid).toBe(`avatar-${sender.id}@shumai.internal`)
expect(thumbAtt).toBeDefined()
expect(thumbAtt.content).toEqual(Buffer.from('poster-thumb-bytes'))
expect(thumbAtt.cid).toBe(`thumb-${asset.id}@shumai.internal`)
// Check HTML references the exact cid
expect(options.html).toContain(`src="cid:${avatarAtt.cid}"`)
expect(options.html).toContain(`src="cid:${thumbAtt.cid}"`)
} finally {
getObjectSpy.mockRestore()
}
})
it('omits thumbnail attachments when cluster has count > 3', async () => {
const team = await createTeamWithEmail()
const sender = await prisma.user.create({
data: {
name: 'Bulk Uploader',
email: 'bulk@test.com',
password: 'p',
image: 'avatars/bulk.png',
},
})
const recipient = await prisma.user.create({
data: { name: 'Recipient', email: 'bulk-recip@test.com', password: 'p' },
})
await prisma.teamMember.create({
data: { teamId: team.id, userId: recipient.id, role: 'editor' },
})
const project = await prisma.project.create({
data: { name: 'Bulk Project', teamId: team.id },
})
for (let i = 0; i < 4; i++) {
const asset = await prisma.asset.create({
data: {
name: `bulk_${i}.png`,
type: 'file',
status: 'uploaded',
project: { connect: { id: project.id } },
media: {
thumbnail: { key: `thumbnails/bulk_${i}.jpg` },
} as unknown as PrismaJson.MediaInfo,
},
})
await prisma.notification.create({
data: {
type: NotificationType.successful_file_uploaded,
teamId: team.id,
projectId: project.id,
assetId: asset.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.pending,
},
})
}
const getObjectSpy = vi
.spyOn(s3Service, 'getObject')
.mockImplementation(async (_bucket, key) => {
if (key === 'avatars/bulk.png') {
return { buffer: Buffer.from('bulk-avatar-bytes'), contentType: 'image/png' }
}
return { buffer: Buffer.from('thumb-bytes'), contentType: 'image/jpeg' }
})
try {
const count = await notificationJobService.processPendingNotifications(500)
expect(count).toBe(4)
expect(sendMailSpy).toHaveBeenCalledTimes(1)
const options = sendMailSpy.mock.calls[0][1]
// Count > 3, so only avatar attachment is included, NO thumbnail attachments!
expect(options.attachments).toHaveLength(1)
expect(options.attachments[0].cid).toBe(`avatar-${sender.id}@shumai.internal`)
expect(options.html).toContain(`src="cid:avatar-${sender.id}@shumai.internal"`)
expect(options.html).not.toContain('src="cid:thumb-')
} finally {
getObjectSpy.mockRestore()
}
})
it('gracefully handles missing images or S3 failure with fallback badge', async () => {
const team = await createTeamWithEmail()
const sender = await prisma.user.create({
data: {
name: 'Broken Image User',
email: 'broken@test.com',
password: 'p',
image: 'avatars/non-existent.png',
},
})
const recipient = await prisma.user.create({
data: { name: 'Recipient', email: 'broken-recip@test.com', password: 'p' },
})
await prisma.teamMember.create({
data: { teamId: team.id, userId: recipient.id, role: 'editor' },
})
const project = await prisma.project.create({
data: { name: 'Project', teamId: team.id },
})
const asset = await prisma.asset.create({
data: {
name: 'broken.png',
type: 'file',
status: 'uploaded',
project: { connect: { id: project.id } },
media: {
thumbnail: { key: 'thumbnails/missing.jpg' },
} as unknown as PrismaJson.MediaInfo,
},
})
await prisma.notification.create({
data: {
type: NotificationType.successful_file_uploaded,
teamId: team.id,
projectId: project.id,
assetId: asset.id,
creatorId: sender.id,
emailStatus: NotificationEmailStatus.pending,
},
})
const getObjectSpy = vi.spyOn(s3Service, 'getObject').mockRejectedValue(new Error('S3 error'))
try {
const count = await notificationJobService.processPendingNotifications(500)
expect(count).toBe(1)
expect(sendMailSpy).toHaveBeenCalledTimes(1)
const options = sendMailSpy.mock.calls[0][1]
// No attachments because both avatar and thumb failed to load
expect(options.attachments).toBeUndefined()
// HTML falls back to letter badge "B" for "Broken Image User"
expect(options.html).toContain('>B</td>')
expect(options.html).not.toContain('<img')
} finally {
getObjectSpy.mockRestore()
}
})
it('starts and stops gracefully', () => {
expect(() => {
notificationJobService.start(100000, 100000)
notificationJobService.stop()
}).not.toThrow()
})
})
@@ -0,0 +1,496 @@
import { prisma, NotificationType, NotificationEmailStatus } from '@shumai/db'
import { emailService, getNotificationBatchGroupKey } from '@shumai/core/src/notification/email'
import type { BatchedNotificationItem, EmailAttachment } from '@shumai/core/src/notification/email'
import { userMetadataService } from '@shumai/core/src/user-metadata/user-metadata'
import { s3Service } from '@shumai/core/src/s3/s3'
import { logger } from '@shumai/core/src/logger'
import type { NotificationSettings } from '@shumai/dtos'
import '@shumai/db/src/prisma-json-types'
function getExtensionFromMime(mime?: string): string {
if (!mime) return 'jpg'
if (mime.includes('png')) return 'png'
if (mime.includes('webp')) return 'webp'
if (mime.includes('gif')) return 'gif'
if (mime.includes('svg')) return 'svg'
return 'jpg'
}
export class NotificationJobService {
private isRunning = false
private timer: NodeJS.Timeout | null = null
private isProcessing = false
start(initialDelayMs = 5000, intervalMs = 3 * 60 * 1000): void {
if (this.isRunning) return
this.isRunning = true
logger.info('Starting notification email background job')
const run = async () => {
if (!this.isRunning) return
try {
await this.processPendingNotifications()
} catch (err) {
logger.error(
{ err: err instanceof Error ? err.message : String(err) },
'Error in notification email background job',
)
}
if (this.isRunning) {
this.timer = setTimeout(run, intervalMs)
}
}
this.timer = setTimeout(run, initialDelayMs)
}
stop(): void {
this.isRunning = false
if (this.timer) {
clearTimeout(this.timer)
this.timer = null
}
logger.info('Stopped notification email background job')
}
async processPendingNotifications(limit = 500): Promise<number> {
if (this.isProcessing) {
logger.debug('Notification job turn already in progress, skipping')
return 0
}
this.isProcessing = true
let batchIds: string[] = []
try {
// 1. Concurrency control: atomically lock up to limit pending notifications
const locked = await prisma.$queryRaw<{ id: string }[]>`
WITH candidates AS (
SELECT id
FROM notifications
WHERE email_status = 'pending'::"NotificationEmailStatus"
ORDER BY id ASC
LIMIT ${limit}
FOR UPDATE SKIP LOCKED
)
UPDATE notifications
SET email_status = 'processing'::"NotificationEmailStatus", updated_at = NOW()
FROM candidates
WHERE notifications.id = candidates.id
RETURNING notifications.id;
`
if (!locked || locked.length === 0) {
return 0
}
batchIds = locked.map((r) => r.id)
logger.info({ count: batchIds.length }, 'Picked up pending notifications for processing')
// 2. Fetch full records with relations
const notifications = await prisma.notification.findMany({
where: { id: { in: batchIds } },
include: {
creator: { select: { id: true, name: true, image: true, type: true } },
team: { select: { id: true, name: true } },
project: { select: { id: true, name: true } },
asset: {
select: {
id: true,
name: true,
media: true,
projectId: true,
createdAt: true,
project: { select: { id: true, name: true } },
},
},
user: { select: { id: true, name: true, email: true, type: true } },
kanbanTask: { select: { id: true, title: true, projectId: true } },
},
orderBy: { id: 'asc' },
})
// 3. Group notifications by teamId
const byTeam = new Map<string, typeof notifications>()
for (const n of notifications) {
const list = byTeam.get(n.teamId)
if (list) {
list.push(n)
} else {
byTeam.set(n.teamId, [n])
}
}
// 4. Process each team
for (const [teamId, teamNotifs] of byTeam) {
const emailSettings = await emailService.getEmailSettings(teamId)
if (!emailSettings.enabled || !emailSettings.host || !emailSettings.from) {
logger.debug(
{ teamId },
'Team email notifications disabled or unconfigured, skipping delivery',
)
continue
}
// Fetch team members excluding AI agents
const members = await prisma.teamMember.findMany({
where: {
teamId,
user: {
type: { not: 'agent' },
agent: null,
},
},
include: {
user: { select: { id: true, name: true, email: true, type: true } },
projectMembers: { select: { projectId: true } },
},
})
const avatarCache = new Map<
string,
{ cidUrl: string; attachment: EmailAttachment } | null
>()
const thumbCache = new Map<string, { cidUrl: string; attachment: EmailAttachment } | null>()
const cidAttachmentMap = new Map<string, EmailAttachment>()
const resolveAvatar = async (
creatorId?: string | null,
image?: string | null,
): Promise<string | undefined> => {
if (!image) return undefined
if (creatorId && avatarCache.has(creatorId)) {
const cached = avatarCache.get(creatorId)
return cached ? cached.cidUrl : undefined
}
const cid = `avatar-${creatorId || Math.random().toString(36).slice(2)}@shumai.internal`
const cidUrl = `cid:${cid}`
try {
if (image.startsWith('http://') || image.startsWith('https://')) {
const res = await fetch(image, { signal: AbortSignal.timeout(5000) })
if (res.ok) {
const arrayBuffer = await res.arrayBuffer()
const buffer = Buffer.from(arrayBuffer)
const contentType = res.headers.get('content-type') || 'image/jpeg'
const ext = getExtensionFromMime(contentType)
const resolved = {
cidUrl,
attachment: {
filename: `avatar-${creatorId || 'user'}.${ext}`,
content: buffer,
cid,
contentType,
},
}
if (creatorId) avatarCache.set(creatorId, resolved)
cidAttachmentMap.set(cidUrl, resolved.attachment)
return cidUrl
}
} else {
const bucket = process.env.S3_BUCKET || 'shumai'
const obj = await s3Service.getObject(bucket, image)
if (obj && obj.buffer && obj.buffer.length > 0) {
const contentType = obj.contentType || 'image/jpeg'
const ext = getExtensionFromMime(contentType)
const resolved = {
cidUrl,
attachment: {
filename: `avatar-${creatorId || 'user'}.${ext}`,
content: obj.buffer,
cid,
contentType,
},
}
if (creatorId) avatarCache.set(creatorId, resolved)
cidAttachmentMap.set(cidUrl, resolved.attachment)
return cidUrl
}
}
} catch (err) {
logger.debug(
{ err: err instanceof Error ? err.message : String(err), image, creatorId },
'Failed to resolve avatar image for inline email attachment',
)
}
if (creatorId) avatarCache.set(creatorId, null)
return undefined
}
const resolveThumbnail = async (
assetId?: string | null,
media?: unknown,
): Promise<string | undefined> => {
if (!assetId || !media) return undefined
if (thumbCache.has(assetId)) {
const cached = thumbCache.get(assetId)
return cached ? cached.cidUrl : undefined
}
const m = media as PrismaJson.MediaInfo | null
const key = m?.thumbnail?.key || m?.poster?.key
if (!key) return undefined
const cid = `thumb-${assetId}@shumai.internal`
const cidUrl = `cid:${cid}`
try {
const bucket = process.env.S3_BUCKET || 'shumai'
const obj = await s3Service.getObject(bucket, key)
if (obj && obj.buffer && obj.buffer.length > 0) {
const contentType = obj.contentType || 'image/jpeg'
const ext = getExtensionFromMime(contentType)
const resolved = {
cidUrl,
attachment: {
filename: `thumb-${assetId}.${ext}`,
content: obj.buffer,
cid,
contentType,
},
}
thumbCache.set(assetId, resolved)
cidAttachmentMap.set(cidUrl, resolved.attachment)
return cidUrl
}
} catch (err) {
logger.debug(
{ err: err instanceof Error ? err.message : String(err), assetId, key },
'Failed to resolve thumbnail image for inline email attachment',
)
}
thumbCache.set(assetId, null)
return undefined
}
const userItemsMap = new Map<
string,
{ user: (typeof members)[0]['user']; items: BatchedNotificationItem[] }
>()
for (const member of members) {
if (!member.user.email) continue
if (
member.user.type === 'agent' ||
member.user.email.endsWith('@shumai.ai') ||
member.user.email.startsWith('agent-')
) {
continue
}
const settingsMeta = await userMetadataService.getMetadata(
member.userId,
teamId,
'notification_settings',
)
const settings: NotificationSettings = settingsMeta
? (settingsMeta.value as NotificationSettings)
: {
comments: true,
replies: true,
mentions: true,
yourUploads: false,
otherUploads: true,
statusUpdates: true,
kanbanTasks: true,
kanbanComments: true,
}
for (const n of teamNotifs) {
// Actor check (do not notify self unless it's their upload and yourUploads is enabled)
if (n.creatorId === member.userId) {
if (n.type !== NotificationType.successful_file_uploaded || !settings.yourUploads) {
continue
}
}
// Targeted notification check
if (n.userId && n.userId !== member.userId) {
continue
}
// Project scope check
if (member.scope === 'project' && n.projectId) {
if (n.type !== NotificationType.mention && n.userId !== member.userId) {
const inProject = member.projectMembers.some((pm) => pm.projectId === n.projectId)
if (!inProject) continue
}
}
// Settings preference check
let allowed = false
switch (n.type) {
case NotificationType.comment_created:
allowed = settings.comments ?? true
break
case NotificationType.reply_created:
allowed = settings.replies ?? true
break
case NotificationType.mention:
allowed = settings.mentions ?? true
break
case NotificationType.successful_file_uploaded:
allowed =
member.userId === n.creatorId
? (settings.yourUploads ?? false)
: (settings.otherUploads ?? true)
break
case NotificationType.metadata_field_updated_status:
allowed = settings.statusUpdates ?? true
break
case NotificationType.kanban_task_created:
case NotificationType.kanban_task_assigned:
case NotificationType.kanban_task_status_updated:
case NotificationType.kanban_task_updated:
case NotificationType.kanban_task_deleted:
allowed = settings.kanbanTasks ?? true
break
case NotificationType.kanban_task_comment_created:
allowed = settings.kanbanComments ?? true
break
case NotificationType.new_user_join_team:
case NotificationType.new_user_join_project:
allowed = true
break
default:
allowed = true
break
}
if (!allowed) continue
const creatorAvatarUrl = await resolveAvatar(n.creatorId, n.creator?.image)
const assetThumbnailUrl = await resolveThumbnail(n.assetId, n.asset?.media)
const item: BatchedNotificationItem = {
id: n.id,
type: n.type || NotificationType.successful_file_uploaded,
creatorName: n.creator?.name || 'Someone',
creatorAvatarUrl,
teamId: n.teamId,
teamName: n.team?.name || 'Team',
projectId: n.projectId || n.asset?.projectId || undefined,
projectName: n.project?.name || n.asset?.project?.name || undefined,
assetId: n.assetId || undefined,
assetName: n.asset?.name || undefined,
assetThumbnailUrl,
kanbanTaskId: n.kanbanTaskId || undefined,
kanbanTaskTitle: n.kanbanTask?.title || undefined,
uploadTime: n.asset?.createdAt || n.createdAt,
createdAt: n.createdAt,
}
let entry = userItemsMap.get(member.userId)
if (!entry) {
entry = { user: member.user, items: [] }
userItemsMap.set(member.userId, entry)
}
entry.items.push(item)
}
}
// Deliver batched email to each user with pending items
for (const [userId, { user, items }] of userItemsMap) {
if (items.length === 0 || !user.email) continue
try {
const rendered = emailService.renderBatchedNotificationEmail({
teamId,
teamName: teamNotifs[0].team?.name || 'Team',
recipientName: user.name || 'Member',
items,
})
// Collect attachments for this email based on rendered items
const attachmentsMap = new Map<string, EmailAttachment>()
// Group items exactly as renderBatchedNotificationEmail does
const groups = new Map<string, BatchedNotificationItem[]>()
for (const item of items) {
const key = getNotificationBatchGroupKey(item)
const list = groups.get(key)
if (list) {
list.push(item)
} else {
groups.set(key, [item])
}
}
for (const [, cluster] of groups) {
const first = cluster[0]
if (first.creatorAvatarUrl) {
const att = cidAttachmentMap.get(first.creatorAvatarUrl)
if (att?.cid) attachmentsMap.set(att.cid, att)
}
if (cluster.length <= 3) {
// Detailed rows: individual avatars and thumbnails are rendered
for (const item of cluster) {
if (item.creatorAvatarUrl) {
const att = cidAttachmentMap.get(item.creatorAvatarUrl)
if (att?.cid) attachmentsMap.set(att.cid, att)
}
if (item.assetThumbnailUrl) {
const att = cidAttachmentMap.get(item.assetThumbnailUrl)
if (att?.cid) attachmentsMap.set(att.cid, att)
}
}
}
// If cluster.length > 3, thumbnails are not rendered, so we do not attach them
}
const attachments = Array.from(attachmentsMap.values())
await emailService.sendMail(emailSettings, {
to: user.email,
subject: rendered.subject,
html: rendered.html,
text: rendered.text,
attachments: attachments.length > 0 ? attachments : undefined,
})
} catch (err) {
logger.error(
{
userId,
email: user.email,
err: err instanceof Error ? err.message : String(err),
},
'Failed to deliver batched notification email to user',
)
}
}
}
// 5. Update batch status to processed
await prisma.notification.updateMany({
where: { id: { in: batchIds } },
data: { emailStatus: NotificationEmailStatus.processed },
})
return batchIds.length
} catch (err) {
logger.error(
{ err: err instanceof Error ? err.message : String(err), batchCount: batchIds.length },
'Unexpected failure in processPendingNotifications',
)
if (batchIds.length > 0) {
// Mark as processed so subsequent turns aren't permanently blocked
await prisma.notification
.updateMany({
where: { id: { in: batchIds } },
data: { emailStatus: NotificationEmailStatus.processed },
})
.catch((cleanErr) =>
logger.error({ cleanErr }, 'Failed to mark notifications processed on error'),
)
}
throw err
} finally {
this.isProcessing = false
}
}
}
export const notificationJobService = new NotificationJobService()
@@ -0,0 +1,256 @@
import { describe, expect, it } from 'vitest'
import { prisma, AssetType, WorkflowTaskStatus, WorkflowTaskType } from '@shumai/db'
import { setupTestDbHooks } from '@shumai/db/test'
import { agentService } from '@shumai/core/src/agent/agent'
import { embeddingService } from './embedding-service'
describe('EmbeddingService', () => {
setupTestDbHooks()
it('should skip teams without an enabled embedding agent', async () => {
const team = await prisma.team.create({
data: { name: 'Team Without Agent' },
})
const project = await prisma.project.create({
data: { name: 'Project 1', teamId: team.id },
})
await prisma.asset.create({
data: {
name: 'video.mp4',
type: AssetType.file,
status: 'processed',
mediaType: 'video/mp4',
projectId: project.id,
},
})
const result = await embeddingService.regenerateEmbeddings()
expect(result.tasksCreated).toBe(0)
expect(result.totalAssets).toBe(0)
expect(result.skippedTeams).toContain('Team Without Agent')
})
it('should skip teams where embedding agent exists but is disabled', async () => {
const team = await prisma.team.create({
data: { name: 'Team Disabled Agent' },
})
await agentService.createAgent({
teamId: team.id,
name: 'embedding-agent',
type: 'embedding',
avatar: 'avatar-1',
thinkingLevel: 'off',
enabled: false,
})
const project = await prisma.project.create({
data: { name: 'Project Disabled', teamId: team.id },
})
await prisma.asset.create({
data: {
name: 'photo.jpg',
type: AssetType.file,
status: 'processed',
mediaType: 'image/jpeg',
projectId: project.id,
},
})
const result = await embeddingService.regenerateEmbeddings()
expect(result.tasksCreated).toBe(0)
expect(result.totalAssets).toBe(0)
expect(result.skippedTeams).toContain('Team Disabled Agent')
})
it('should regenerate embeddings for eligible media assets and clean up old data', async () => {
const team = await prisma.team.create({
data: { name: 'Active Embedding Team' },
})
const agent = await agentService.createAgent({
teamId: team.id,
name: 'embedding-agent',
type: 'embedding',
avatar: 'avatar-1',
thinkingLevel: 'off',
enabled: true,
})
const project = await prisma.project.create({
data: { name: 'Active Project', teamId: team.id },
})
// 1. Eligible video asset
const videoAsset = await prisma.asset.create({
data: {
name: 'sample.mp4',
type: AssetType.file,
status: 'processed',
mediaType: 'video/mp4',
projectId: project.id,
},
})
// 2. Eligible image asset
const imageAsset = await prisma.asset.create({
data: {
name: 'sample.png',
type: AssetType.file,
status: 'processed',
mediaType: 'image/png',
projectId: project.id,
},
})
// 3. Eligible 3D asset with rendered views
const asset3d = await prisma.asset.create({
data: {
name: 'model.glb',
type: AssetType.file,
status: 'processed',
mediaType: 'model/gltf-binary',
media: {
proxyType: '3d',
imageTranscodes: [
{ key: 'files/3d/view.webp', format: 'webp', width: 100, height: 100, quality: 80 },
],
} as unknown as PrismaJson.MediaInfo,
projectId: project.id,
},
})
// 4. Ineligible asset (non-media document)
await prisma.asset.create({
data: {
name: 'document.pdf',
type: AssetType.file,
status: 'processed',
mediaType: 'application/pdf',
projectId: project.id,
},
})
// 5. Ineligible asset (trashed / deleted)
await prisma.asset.create({
data: {
name: 'deleted-video.mp4',
type: AssetType.file,
status: 'processed',
isDeleted: true,
mediaType: 'video/mp4',
projectId: project.id,
},
})
// 6. Ineligible asset (status not processed)
await prisma.asset.create({
data: {
name: 'uploading.mp4',
type: AssetType.file,
status: 'uploading',
mediaType: 'video/mp4',
projectId: project.id,
},
})
// Create preexisting embeddings and workflow task for videoAsset to verify cleanup
const dummyVec768 = `[${new Array(768).fill(0.1).join(',')}]`
await prisma.$executeRaw`
INSERT INTO asset_embeddings (id, asset_id, embedding_768, updated_at)
VALUES (gen_random_uuid()::text, ${videoAsset.id}, ${dummyVec768}::vector, NOW())
`
await prisma.workflowTask.create({
data: {
assetId: videoAsset.id,
type: WorkflowTaskType.ai_embedding,
status: WorkflowTaskStatus.processing,
teamId: team.id,
projectId: project.id,
},
})
const initialEmbeddings = await prisma.assetEmbedding.count({
where: { assetId: videoAsset.id },
})
expect(initialEmbeddings).toBe(1)
// Run regeneration
const result = await embeddingService.regenerateEmbeddings()
expect(result.totalAssets).toBe(3)
expect(result.tasksCreated).toBe(3)
expect(result.skippedTeams).not.toContain('Active Embedding Team')
// Verify old embeddings were wiped upfront
const remainingEmbeddings = await prisma.assetEmbedding.count({
where: { assetId: videoAsset.id },
})
expect(remainingEmbeddings).toBe(0)
// Verify tasks created
const createdTasks = await prisma.workflowTask.findMany({
where: {
type: WorkflowTaskType.ai_embedding,
assetId: { in: [videoAsset.id, imageAsset.id, asset3d.id] },
},
})
expect(createdTasks).toHaveLength(3)
for (const task of createdTasks) {
expect(task.status).toBe(WorkflowTaskStatus.pending)
expect(task.teamId).toBe(team.id)
expect(task.projectId).toBe(project.id)
const payload = task.payload as { agent?: { agentId?: string } }
expect(payload?.agent?.agentId).toBe(agent!.id)
}
})
it('should handle multiple teams where only some have active embedding agents', async () => {
// Team 1: has active agent
const team1 = await prisma.team.create({ data: { name: 'Team One' } })
await agentService.createAgent({
teamId: team1.id,
name: 'embedding-agent',
type: 'embedding',
avatar: 'avatar-1',
thinkingLevel: 'off',
enabled: true,
})
const project1 = await prisma.project.create({ data: { name: 'P1', teamId: team1.id } })
await prisma.asset.create({
data: {
name: 'v1.mp4',
type: AssetType.file,
status: 'processed',
mediaType: 'video/mp4',
projectId: project1.id,
},
})
// Team 2: no active agent
const team2 = await prisma.team.create({ data: { name: 'Team Two' } })
const project2 = await prisma.project.create({ data: { name: 'P2', teamId: team2.id } })
await prisma.asset.create({
data: {
name: 'v2.mp4',
type: AssetType.file,
status: 'processed',
mediaType: 'video/mp4',
projectId: project2.id,
},
})
const result = await embeddingService.regenerateEmbeddings()
expect(result.totalAssets).toBe(1)
expect(result.tasksCreated).toBe(1)
expect(result.skippedTeams).toContain('Team Two')
expect(result.skippedTeams).not.toContain('Team One')
const tasks = await prisma.workflowTask.findMany({
where: {
type: WorkflowTaskType.ai_embedding,
teamId: { in: [team1.id, team2.id] },
},
})
expect(tasks).toHaveLength(1)
expect(tasks[0].teamId).toBe(team1.id)
})
})
@@ -0,0 +1,135 @@
import { prisma, WorkflowTaskStatus, WorkflowTaskType } from '@shumai/db'
import { logger } from '@shumai/core/src/logger'
export interface RegenerateEmbeddingsResult {
tasksCreated: number
totalAssets: number
skippedTeams: string[]
}
export class EmbeddingService {
/**
* Regenerates vector embeddings across all teams.
*
* Skips teams that do not have an active (enabled) embedding agent and logs a warning.
* For teams with an active embedding agent, purges existing embeddings and pending/processing
* embedding tasks upfront, then creates fresh ai_embedding workflow tasks for all eligible assets.
*/
async regenerateEmbeddings(): Promise<RegenerateEmbeddingsResult> {
const teams = await prisma.team.findMany({
select: { id: true, name: true },
})
const activeAgents = await prisma.agent.findMany({
where: {
type: 'embedding',
enabled: true,
},
select: { id: true, teamId: true },
})
const activeAgentByTeamId = new Map(
activeAgents
.filter((a): a is typeof a & { teamId: string } => typeof a.teamId === 'string')
.map((a) => [a.teamId, a]),
)
const activeTeamIds: string[] = []
const skippedTeams: string[] = []
for (const team of teams) {
if (activeAgentByTeamId.has(team.id)) {
activeTeamIds.push(team.id)
} else {
skippedTeams.push(team.name || team.id)
logger.warn(
{ teamId: team.id, teamName: team.name },
'Skipping embedding regeneration for team: no active embedding agent found',
)
}
}
if (activeTeamIds.length === 0) {
logger.info('No teams found with an active embedding agent. 0 tasks created.')
return { tasksCreated: 0, totalAssets: 0, skippedTeams }
}
const candidateAssets = await prisma.asset.findMany({
where: {
isDeleted: false,
status: 'processed',
type: { in: ['file', 'version_stack'] },
project: { teamId: { in: activeTeamIds } },
},
include: { project: true },
})
const eligibleAssets = candidateAssets.filter((asset) => {
const isVideo = asset.mediaType?.startsWith('video/')
const mediaInfo = asset.media as PrismaJson.MediaInfo | null
const isImage =
asset.mediaType?.startsWith('image/') ||
(mediaInfo?.proxyType === '3d' && (mediaInfo.imageTranscodes?.length ?? 0) > 0)
return isVideo || isImage
})
const eligibleAssetIds = eligibleAssets.map((a) => a.id)
if (eligibleAssetIds.length > 0) {
// Clean up previous embedding tasks to avoid duplicate execution
await prisma.workflowTask.deleteMany({
where: {
assetId: { in: eligibleAssetIds },
type: WorkflowTaskType.ai_embedding,
},
})
// Purge existing embeddings so everything regenerates cleanly with the configured provider
await prisma.assetEmbedding.deleteMany({
where: {
assetId: { in: eligibleAssetIds },
},
})
}
let tasksCreated = 0
for (const asset of eligibleAssets) {
const teamId = asset.project?.teamId
if (!teamId) continue
const agent = activeAgentByTeamId.get(teamId)
if (!agent) continue
await prisma.workflowTask.create({
data: {
assetId: asset.id,
type: WorkflowTaskType.ai_embedding,
status: WorkflowTaskStatus.pending,
teamId,
projectId: asset.projectId,
payload: {
projectId: asset.projectId ?? '',
agent: { agentId: agent.id },
},
},
})
tasksCreated++
}
logger.info(
{
totalAssets: eligibleAssets.length,
tasksCreated,
skippedTeamsCount: skippedTeams.length,
},
'Successfully queued embedding regeneration tasks',
)
return {
tasksCreated,
totalAssets: eligibleAssets.length,
skippedTeams,
}
}
}
export const embeddingService = new EmbeddingService()
+248 -13
View File
@@ -4,7 +4,7 @@ import { setupTestDbHooks } from '@shumai/db/test'
import { SearchService } from './search'
import { SearchConditionOperator } from '@shumai/dtos'
import { metadataService } from '@shumai/core/src/metadata/metadata'
import { workflowService } from '@shumai/workflow-core'
import { textEmbeddingService } from './text-embedding'
import { AssetType } from '@shumai/db'
import { assetService } from '@shumai/core/src/asset/asset'
@@ -805,20 +805,14 @@ describe('SearchService', () => {
)
`
// Mock the embedding generation workflow task.
// Mock the embedding generation
// Our query vector points along the first dimension [1, 0, 0...]
const queryVector = Array(1536).fill(0)
queryVector[0] = 1.0
const mockExecuteWait = vi
.spyOn(workflowService, 'executeWait')
.mockImplementation(async (task) => {
return {
...task,
status: 'completed',
output: {
embedding: queryVector,
},
} as unknown as typeof task
const mockGenerateEmbedding = vi
.spyOn(textEmbeddingService, 'generateEmbedding')
.mockResolvedValue({
embedding: queryVector,
})
// Search using semantic mode
@@ -855,7 +849,248 @@ describe('SearchService', () => {
expect(result.data[2].startTime).toBe(10)
expect(result.data[2].endTime).toBe(20)
mockExecuteWait.mockRestore()
mockGenerateEmbedding.mockRestore()
})
it('matches and sorts 768-dimensional embeddings using embedding_768 column', async () => {
const user = await prisma.user.create({
data: { name: 'semantic-user-768', email: 'semantic-user-768@example.com', type: 'human' },
})
const team = await prisma.team.create({
data: { name: 'semantic-team-768' },
})
const project = await prisma.project.create({
data: { name: 'semantic-project-768', teamId: team.id },
})
const botUser = await prisma.user.create({
data: { name: 'Embedding Bot 768', email: 'emb-768@example.com', type: 'agent' },
})
const provider = await prisma.provider.create({
data: {
name: 'gemma',
teamId: team.id,
config: { api: 'openai', apiKey: 'key' } as unknown as PrismaJson.ProviderConfig,
},
})
const model = await prisma.model.create({
data: {
modelId: 'embeddinggemma-2',
name: 'EmbeddingGemma 2',
providerId: provider.id,
config: {
reasoning: false,
input: ['text'],
contextWindow: 1000,
maxTokens: 1000,
cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
} as unknown as PrismaJson.ModelConfig,
},
})
await prisma.agent.create({
data: {
id: botUser.id,
teamId: team.id,
type: 'embedding',
enabled: true,
providerId: provider.id,
modelId: model.id,
config: { provider: 'gemma-embedding-2-openai', model: 'embeddinggemma-2' },
},
})
const rootFolder = await prisma.asset.create({
data: {
name: 'root-768',
type: AssetType.folder,
projectId: project.id,
creatorId: user.id,
status: 'uploaded',
},
})
const videoAssetA = await prisma.asset.create({
data: {
name: 'VideoA_768.mp4',
type: AssetType.file,
mediaType: 'video/mp4',
projectId: project.id,
parentId: rootFolder.id,
creatorId: user.id,
status: 'uploaded',
},
})
const videoAssetB = await prisma.asset.create({
data: {
name: 'VideoB_768.mp4',
type: AssetType.file,
mediaType: 'video/mp4',
projectId: project.id,
parentId: rootFolder.id,
creatorId: user.id,
status: 'uploaded',
},
})
const arrA1 = Array(768).fill(0)
arrA1[0] = 1.0
const vectorA1 = JSON.stringify(arrA1)
const arrB1 = Array(768).fill(0)
arrB1[0] = 1.0
arrB1[1] = 1.0
const vectorB1 = JSON.stringify(arrB1)
await prisma.$executeRaw`
INSERT INTO asset_embeddings (id, asset_id, embedding_768, start_time, end_time, created_at, updated_at)
VALUES (
'chunk-768-a1',
${videoAssetA.id},
${vectorA1}::vector,
0.0,
10.0,
NOW(),
NOW()
)
`
await prisma.$executeRaw`
INSERT INTO asset_embeddings (id, asset_id, embedding_768, start_time, end_time, created_at, updated_at)
VALUES (
'chunk-768-b1',
${videoAssetB.id},
${vectorB1}::vector,
5.0,
15.0,
NOW(),
NOW()
)
`
// Query vector with 768 dims pointing along dimension 0
const queryVector = Array(768).fill(0)
queryVector[0] = 1.0
const mockGenerateEmbedding = vi
.spyOn(textEmbeddingService, 'generateEmbedding')
.mockResolvedValue({
embedding: queryVector,
})
const result = await searchService.search(rootFolder.id, {
recursively: true,
assetType: 'file',
operator: 'AND',
conditions: [],
query: 'find 768 dim match',
isSemantic: true,
})
expect(result.data.length).toBeGreaterThanOrEqual(2)
expect(result.data[0].id).toBe(videoAssetA.id)
expect(result.data[1].id).toBe(videoAssetB.id)
mockGenerateEmbedding.mockRestore()
})
it('throws 502 if embedding generation fails during semantic search', async () => {
const user = await prisma.user.create({
data: { name: 'semantic-user-fail', email: 'semantic-fail@example.com', type: 'human' },
})
const team = await prisma.team.create({
data: { name: 'semantic-team-fail' },
})
const project = await prisma.project.create({
data: { name: 'semantic-project-fail', teamId: team.id },
})
const botUser = await prisma.user.create({
data: { name: 'Embedding Bot Fail', email: 'emb-fail@example.com', type: 'agent' },
})
await prisma.agent.create({
data: {
id: botUser.id,
teamId: team.id,
type: 'embedding',
enabled: true,
config: { provider: 'gemini-embedding-2', model: 'gemini-embedding-2' },
},
})
const rootFolder = await prisma.asset.create({
data: {
name: 'root-fail',
type: AssetType.folder,
projectId: project.id,
creatorId: user.id,
status: 'uploaded',
},
})
const mockGenerateEmbedding = vi
.spyOn(textEmbeddingService, 'generateEmbedding')
.mockRejectedValue(new Error('Rate limit exceeded'))
await expect(
searchService.search(rootFolder.id, {
recursively: true,
assetType: 'file',
operator: 'AND',
conditions: [],
query: 'fail query',
isSemantic: true,
}),
).rejects.toThrow('Failed to generate query embedding: Rate limit exceeded')
mockGenerateEmbedding.mockRestore()
})
it('rejects semantic search when ENABLE_EMBEDDING_AGENT is false', async () => {
const user = await prisma.user.create({
data: { name: 'u-emb-false', email: 'u-emb-false@example.com', type: 'human' },
})
const team = await prisma.team.create({ data: { name: 't-emb-false' } })
const project = await prisma.project.create({
data: { name: 'p-emb-false', teamId: team.id },
})
const rootFolder = await prisma.asset.create({
data: {
name: 'root-false',
type: AssetType.folder,
projectId: project.id,
creatorId: user.id,
status: 'uploaded',
},
})
// Even if an embedding agent is enabled in db
const bot = await prisma.user.create({
data: { name: 'Bot', email: 'b-emb-false@example.com', password: 'pw', type: 'agent' },
})
await prisma.agent.create({
data: {
id: bot.id,
teamId: team.id,
type: 'embedding',
enabled: true,
config: { provider: '', model: '' },
},
})
process.env.ENABLE_EMBEDDING_AGENT = 'false'
try {
await expect(
searchService.search(rootFolder.id, {
recursively: true,
assetType: 'file',
operator: 'AND',
conditions: [],
query: 'test query',
isSemantic: true,
}),
).rejects.toThrow('Embedding agent not configured or disabled for this team.')
} finally {
delete process.env.ENABLE_EMBEDDING_AGENT
}
})
})
+36 -21
View File
@@ -1,18 +1,19 @@
import { prisma } from '@shumai/db'
import { Prisma, AssetType, WorkflowTaskType } from '@shumai/db'
import { Prisma, AssetType } from '@shumai/db'
import { AssetService, assetService } from '@shumai/core/src/asset/asset'
import { AssetInfo } from '@shumai/dtos'
import { SearchRequest } from '@shumai/dtos'
import { PaginatedData, decodeCursor, encodeCursor, PageInfo } from '@shumai/core/src/pagination'
import { generateSearchNgrams } from '@shumai/core/src/utils/ngram'
import { workflowService } from '@shumai/workflow-core'
import { HTTPException } from 'hono/http-exception'
import { SqlQueryBuilder } from './sql-query-builder'
import { TextEmbeddingService, textEmbeddingService } from './text-embedding'
export class SearchService {
constructor(
private readonly prismaClient: typeof prisma = prisma,
private readonly assetSvc: AssetService = assetService,
private readonly embeddingSvc: TextEmbeddingService = textEmbeddingService,
) {}
async search(folderId: string, req: SearchRequest): Promise<PaginatedData<AssetInfo[]>> {
@@ -34,6 +35,12 @@ export class SearchService {
const teamId = team?.project?.teamId
if (!teamId) throw new Error('Team ID not found for folder')
if (process.env.ENABLE_EMBEDDING_AGENT === 'false') {
throw new HTTPException(422, {
message: 'Embedding agent not configured or disabled for this team.',
})
}
const embeddingAgent = await this.prismaClient.agent.findFirst({
where: { teamId, type: 'embedding', enabled: true },
})
@@ -44,36 +51,43 @@ export class SearchService {
})
}
// 2. Generate query embedding via workflow
const task = await this.prismaClient.workflowTask.create({
data: {
type: WorkflowTaskType.query_embedding_for_search,
// 2. Generate query embedding directly using configured environment provider
let queryVector: number[]
try {
const result = await this.embeddingSvc.generateEmbedding(req.query, {
teamId,
assetId: folderId,
payload: {
projectId: '',
queryEmbeddingForSearch: { text: req.query },
} as PrismaJson.WorkflowTaskPayload,
status: 'pending',
},
})
})
queryVector = result.embedding
} catch (err) {
const msg = err instanceof Error ? err.message : String(err)
throw new HTTPException(502, {
message: `Failed to generate query embedding: ${msg}`,
})
}
const completedTask = await workflowService.executeWait(task)
const output = completedTask.output as Record<string, unknown> | null
const queryVector = output?.embedding as number[] | undefined
if (!queryVector) {
throw new Error('Failed to generate query embedding')
if (!queryVector || queryVector.length === 0) {
throw new HTTPException(502, {
message: 'Failed to generate query embedding: empty vector returned',
})
}
// 3. Construct raw SQL using SqlQueryBuilder
const vectorJson = JSON.stringify(queryVector)
const is768 = queryVector.length === 768
const distanceExpr = is768
? Prisma.sql`(ae.embedding_768 <=> ${vectorJson}::vector)`
: Prisma.sql`(ae.embedding <=> ${vectorJson}::vector)`
const notNullCond = is768
? Prisma.sql`ae.embedding_768 IS NOT NULL`
: Prisma.sql`ae.embedding IS NOT NULL`
const builder = new SqlQueryBuilder()
.select(
Prisma.sql`a.id as "assetId", ae.start_time as "startTime", ae.end_time as "endTime", (ae.embedding <=> ${vectorJson}::vector) as "distance"`,
Prisma.sql`a.id as "assetId", ae.start_time as "startTime", ae.end_time as "endTime", ${distanceExpr} as "distance"`,
)
.from(Prisma.sql`assets a JOIN asset_embeddings ae ON a.id = ae.asset_id`)
.addWhere(Prisma.sql`a.is_deleted = false`)
.addWhere(notNullCond)
if (targetFolderIds.length > 0) {
builder.addWhere(Prisma.sql`a.parent_id = ANY(${targetFolderIds})`)
@@ -158,6 +172,7 @@ export class SearchService {
.select(Prisma.sql`COUNT(*)`)
.from(Prisma.sql`assets a JOIN asset_embeddings ae ON a.id = ae.asset_id`)
.addWhere(Prisma.sql`a.is_deleted = false`)
.addWhere(notNullCond)
if (targetFolderIds.length > 0) {
countBuilder.addWhere(Prisma.sql`a.parent_id = ANY(${targetFolderIds})`)
@@ -0,0 +1,236 @@
import { describe, it, expect, beforeEach, afterEach, vi } from 'vitest'
import { TextEmbeddingService } from './text-embedding'
import { logger } from '@shumai/core/src/logger'
const mockEmbedContent = vi.fn()
vi.mock('@google/genai', () => {
return {
// eslint-disable-next-line @typescript-eslint/naming-convention
GoogleGenAI: class {
models = {
embedContent: mockEmbedContent,
}
},
}
})
describe('TextEmbeddingService', () => {
const originalEnv = { ...process.env }
let service: TextEmbeddingService
beforeEach(() => {
process.env = { ...originalEnv }
process.env.GEMINI_API_KEY = 'test-gemini-key'
delete process.env.EMBEDDING_PROVIDER
delete process.env.GEMMA_EMBEDDING_OPENAI_BASE_URL
mockEmbedContent.mockReset()
service = new TextEmbeddingService(500, 3600 * 1000)
})
afterEach(() => {
process.env = originalEnv
vi.restoreAllMocks()
})
it('throws an error if text is empty or whitespace', async () => {
await expect(service.generateEmbedding('')).rejects.toThrow('Text to embed must not be empty')
await expect(service.generateEmbedding(' ')).rejects.toThrow(
'Text to embed must not be empty',
)
})
it('generates Gemini embedding (1536d) by default', async () => {
const dummyEmbedding = Array(1536).fill(0.05)
mockEmbedContent.mockResolvedValueOnce({
embeddings: [{ values: dummyEmbedding }],
})
const result = await service.generateEmbedding('beautiful sunset')
expect(result.embedding).toEqual(dummyEmbedding)
expect(result.usage?.model).toBe('gemini-embedding-2')
expect(result.cached).toBeUndefined()
expect(mockEmbedContent).toHaveBeenCalledWith(
expect.objectContaining({
model: 'gemini-embedding-2',
contents: [
expect.objectContaining({
parts: [{ text: 'beautiful sunset' }],
}),
],
config: {
outputDimensionality: 1536,
},
}),
)
})
it('returns cached result on repeated call without re-invoking the provider', async () => {
const dummyEmbedding = Array(1536).fill(0.05)
mockEmbedContent.mockResolvedValueOnce({
embeddings: [{ values: dummyEmbedding }],
})
const first = await service.generateEmbedding('ocean breeze')
expect(first.embedding).toEqual(dummyEmbedding)
expect(mockEmbedContent).toHaveBeenCalledTimes(1)
// Second call with same text and extra whitespace
const second = await service.generateEmbedding(' ocean breeze ')
expect(second.embedding).toEqual(dummyEmbedding)
expect(second.cached).toBe(true)
expect(mockEmbedContent).toHaveBeenCalledTimes(1)
})
it('preserves case distinctions in cache keys', async () => {
const dummyApple = Array(1536).fill(0.1)
const dummyLowerApple = Array(1536).fill(0.2)
mockEmbedContent
.mockResolvedValueOnce({
embeddings: [{ values: dummyApple }],
})
.mockResolvedValueOnce({
embeddings: [{ values: dummyLowerApple }],
})
const upper = await service.generateEmbedding('Apple')
expect(upper.embedding).toEqual(dummyApple)
expect(mockEmbedContent).toHaveBeenCalledTimes(1)
const lower = await service.generateEmbedding('apple')
expect(lower.embedding).toEqual(dummyLowerApple)
expect(lower.cached).toBeUndefined()
expect(mockEmbedContent).toHaveBeenCalledTimes(2)
})
it('does not log raw query text on cache hits', async () => {
const dummyEmbedding = Array(1536).fill(0.05)
mockEmbedContent.mockResolvedValue({
embeddings: [{ values: dummyEmbedding }],
})
const debugSpy = vi.spyOn(logger, 'debug')
await service.generateEmbedding('confidential search query')
await service.generateEmbedding('confidential search query')
expect(debugSpy).toHaveBeenCalledWith(
expect.not.objectContaining({
key: expect.anything(),
text: expect.anything(),
}),
'Cache hit for text embedding',
)
})
it('clears cache when clearCache is called', async () => {
const dummyEmbedding = Array(1536).fill(0.05)
mockEmbedContent.mockResolvedValue({
embeddings: [{ values: dummyEmbedding }],
})
await service.generateEmbedding('query')
expect(mockEmbedContent).toHaveBeenCalledTimes(1)
service.clearCache()
await service.generateEmbedding('query')
expect(mockEmbedContent).toHaveBeenCalledTimes(2)
})
it('throws if GEMINI_API_KEY is not configured', async () => {
delete process.env.GEMINI_API_KEY
await expect(service.generateEmbedding('test')).rejects.toThrow(
'GEMINI_API_KEY environment variable is not configured',
)
})
it('throws if Gemini returns empty embedding values', async () => {
mockEmbedContent.mockResolvedValueOnce({
embeddings: [{ values: null }],
})
await expect(service.generateEmbedding('test')).rejects.toThrow(
'No embedding values returned from gemini-embedding-2',
)
})
it('generates Gemma embedding (768d) when provider is gemma-embedding-2-openai', async () => {
const dummyEmbedding = Array(768).fill(0.12)
const mockFetch = vi.fn().mockResolvedValue({
ok: true,
json: async () => ({
model: 'embeddinggemma-2',
data: [{ embedding: dummyEmbedding }],
// eslint-disable-next-line @typescript-eslint/naming-convention
usage: { prompt_tokens: 12 },
}),
})
vi.stubGlobal('fetch', mockFetch)
const result = await service.generateEmbedding('drone footage', {
provider: 'gemma-embedding-2-openai',
})
expect(result.embedding).toEqual(dummyEmbedding)
expect(result.usage?.model).toBe('embeddinggemma-2')
expect(result.usage?.inputTokens).toBe(12)
expect(mockFetch).toHaveBeenCalledWith(
'http://localhost:9379/v1/embeddings',
expect.objectContaining({
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
model: 'embeddinggemma-2',
input: 'task: search query | text: drone footage',
}),
}),
)
})
it('respects GEMMA_EMBEDDING_OPENAI_BASE_URL without /v1 suffix', async () => {
process.env.GEMMA_EMBEDDING_OPENAI_BASE_URL = 'http://custom-host:9379'
const dummyEmbedding = Array(768).fill(0.12)
const mockFetch = vi.fn().mockResolvedValue({
ok: true,
json: async () => ({
data: [{ embedding: dummyEmbedding }],
}),
})
vi.stubGlobal('fetch', mockFetch)
await service.generateEmbedding('test query', {
provider: 'gemma-embedding-2-openai',
})
expect(mockFetch).toHaveBeenCalledWith(
'http://custom-host:9379/v1/embeddings',
expect.anything(),
)
})
it('throws descriptive error when Gemma server returns non-ok status', async () => {
const mockFetch = vi.fn().mockResolvedValue({
ok: false,
status: 503,
text: async () => 'Service Unavailable',
})
vi.stubGlobal('fetch', mockFetch)
await expect(
service.generateEmbedding('test query', {
provider: 'gemma-embedding-2-openai',
}),
).rejects.toThrow('Embedding server returned status 503: Service Unavailable')
})
it('throws error when abort signal is triggered', async () => {
const controller = new AbortController()
controller.abort(new Error('Operation cancelled by user'))
await expect(
service.generateEmbedding('test query', {
signal: controller.signal,
}),
).rejects.toThrow('Operation cancelled by user')
})
})
+196
View File
@@ -0,0 +1,196 @@
import { GoogleGenAI } from '@google/genai'
import { LruTtlCache } from '@shumai/core/src/cache/lru-ttl-cache'
import { logger } from '@shumai/core/src/logger'
export interface TextEmbeddingUsage {
model: string
inputTokens: number
outputTokens: number
}
export interface TextEmbeddingResult {
embedding: number[]
usage?: TextEmbeddingUsage
cached?: boolean
}
export interface GenerateTextEmbeddingOptions {
provider?: string
model?: string
signal?: AbortSignal
teamId?: string
skipCache?: boolean
}
export class TextEmbeddingService {
private cache: LruTtlCache<string, number[]>
constructor(
cacheCapacity: number = 500,
private defaultTtlMs: number = 3600 * 1000,
) {
this.cache = new LruTtlCache<string, number[]>(cacheCapacity, 50)
}
/**
* Generates a text embedding vector directly using the configured provider.
* If a cached embedding exists for the text, provider, and model, it is returned immediately.
*/
async generateEmbedding(
text: string,
options?: GenerateTextEmbeddingOptions,
): Promise<TextEmbeddingResult> {
const trimmedText = text.trim()
if (!trimmedText) {
throw new Error('Text to embed must not be empty')
}
const provider = options?.provider || process.env.EMBEDDING_PROVIDER || 'gemini-embedding-2'
const model =
options?.model ||
(provider === 'gemma-embedding-2-openai' ? 'embeddinggemma-2' : 'gemini-embedding-2')
const cacheKey = `${provider}:${model}:${trimmedText}`
if (!options?.skipCache) {
const cachedVector = this.cache.get(cacheKey)
if (cachedVector) {
logger.debug({ provider, model }, 'Cache hit for text embedding')
return {
embedding: cachedVector,
usage: { model, inputTokens: 0, outputTokens: 0 },
cached: true,
}
}
}
const signal = options?.signal ?? AbortSignal.timeout(15000)
let result: TextEmbeddingResult
if (provider === 'gemma-embedding-2-openai') {
result = await this.callGemmaEmbeddingServer(trimmedText, model, signal)
} else {
result = await this.callGeminiEmbedding(trimmedText, model, signal)
}
if (!options?.skipCache) {
this.cache.set(cacheKey, result.embedding, this.defaultTtlMs)
}
return result
}
private async callGeminiEmbedding(
text: string,
model: string,
signal: AbortSignal,
): Promise<TextEmbeddingResult> {
const apiKey = process.env.GEMINI_API_KEY
if (!apiKey) {
throw new Error('GEMINI_API_KEY environment variable is not configured')
}
if (signal.aborted) {
throw signal.reason ?? new Error('Embedding generation aborted or timed out')
}
const resolvedApiKey = process.env[apiKey] || apiKey
const ai = new GoogleGenAI({ apiKey: resolvedApiKey })
const embedPromise = ai.models.embedContent({
model,
contents: [
{
parts: [{ text }],
role: 'user',
},
],
config: {
outputDimensionality: 1536,
},
})
const abortPromise = new Promise<never>((_, reject) => {
if (signal.aborted) {
reject(signal.reason ?? new Error('Embedding generation aborted or timed out'))
return
}
signal.addEventListener(
'abort',
() => {
reject(signal.reason ?? new Error('Embedding generation aborted or timed out'))
},
{ once: true },
)
})
const response = await Promise.race([embedPromise, abortPromise])
const values = response.embeddings?.[0]?.values
if (!values || !Array.isArray(values)) {
throw new Error(`No embedding values returned from ${model}`)
}
return {
embedding: values,
usage: {
model,
inputTokens: 0,
outputTokens: 0,
},
}
}
private async callGemmaEmbeddingServer(
text: string,
model: string,
signal: AbortSignal,
): Promise<TextEmbeddingResult> {
const rawBaseUrl = process.env.GEMMA_EMBEDDING_OPENAI_BASE_URL || 'http://localhost:9379/v1'
const normalized = rawBaseUrl.replace(/\/+$/, '')
const endpoint = normalized.endsWith('/embeddings')
? normalized
: normalized.endsWith('/v1')
? `${normalized}/embeddings`
: `${normalized}/v1/embeddings`
const response = await fetch(endpoint, {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({
model,
input: `task: search query | text: ${text}`,
}),
signal,
})
if (!response.ok) {
const errorText = await response.text()
throw new Error(`Embedding server returned status ${response.status}: ${errorText}`)
}
const json = (await response.json()) as {
model?: string
data?: Array<{ embedding?: number[] }>
// eslint-disable-next-line @typescript-eslint/naming-convention
usage?: { prompt_tokens?: number }
}
const embedding = json.data?.[0]?.embedding
if (!embedding || !Array.isArray(embedding)) {
throw new Error('Invalid embedding returned from embedding server')
}
return {
embedding,
usage: {
model: json.model || model,
inputTokens: json.usage?.prompt_tokens ?? 0,
outputTokens: 0,
},
}
}
clearCache(): void {
this.cache.clear()
}
}
export const textEmbeddingService = new TextEmbeddingService()
+135 -3
View File
@@ -2,6 +2,7 @@ import { prisma } from '@shumai/db'
import { setupTestDbHooks } from '@shumai/db/test'
import { teamService } from '@shumai/core/src/team/team'
import { describe, expect, it } from 'vitest'
import type { TeamSettingsResponse } from '@shumai/dtos'
describe('TeamService', () => {
setupTestDbHooks()
@@ -10,6 +11,7 @@ describe('TeamService', () => {
const team = await teamService.ensureDefaultTeam()
expect(team).toBeDefined()
expect(team.name).toBe('Default Team')
expect((team.settings as TeamSettingsResponse | null)?.transcode?.threads).toBe(0)
})
it('should create team', async () => {
@@ -363,7 +365,30 @@ describe('TeamService', () => {
})
const settings = await teamService.getSettings(team.id)
expect(settings).toEqual({ semanticSearchEnabled: false, appearance: { hideAgent: false } })
expect(settings).toEqual({
semanticSearchEnabled: false,
transcode: {
videoStrategy: 'best_match',
videoResolutions: ['480p', '720p', '1080p', '1440p', '2160p'],
hardwareAcceleration: 'off',
threads: 0,
hlsEnabled: false,
hlsResolutions: ['480p', '720p', '1080p'],
},
turntable: {
hasPassword: false,
isEnvConfigured: false,
url: undefined,
username: undefined,
},
gotenberg: {
hasPassword: false,
isEnvConfigured: false,
url: undefined,
username: undefined,
},
appearance: { hideAgent: false },
})
await teamService.updateSettings(team.id, 'theme', 'dark')
@@ -371,17 +396,64 @@ describe('TeamService', () => {
expect(newSettings).toEqual({
theme: 'dark',
semanticSearchEnabled: false,
transcode: {
videoStrategy: 'best_match',
videoResolutions: ['480p', '720p', '1080p', '1440p', '2160p'],
hardwareAcceleration: 'off',
threads: 0,
hlsEnabled: false,
hlsResolutions: ['480p', '720p', '1080p'],
},
turntable: {
hasPassword: false,
isEnvConfigured: false,
url: undefined,
username: undefined,
},
gotenberg: {
hasPassword: false,
isEnvConfigured: false,
url: undefined,
username: undefined,
},
appearance: { hideAgent: false },
})
await teamService.updateSettings(team.id, 'transcode.videoStrategy', 'all')
await teamService.updateSettings(team.id, 'transcode.videoStrategy', 'multi')
await teamService.updateSettings(team.id, 'transcode.videoResolutions', ['720p', '1080p'])
await teamService.updateSettings(team.id, 'transcode.hardwareAcceleration', 'auto')
await teamService.updateSettings(team.id, 'transcode.threads', 4)
await teamService.updateSettings(team.id, 'transcode.hlsEnabled', true)
await teamService.updateSettings(team.id, 'transcode.hlsResolutions', [
'720p',
'1080p',
'2160p',
])
await teamService.updateSettings(team.id, 'appearance.hideAgent', true)
const finalSettings = await teamService.getSettings(team.id)
expect(finalSettings).toEqual({
theme: 'dark',
transcode: { videoStrategy: 'all', hardwareAcceleration: 'auto' },
transcode: {
videoStrategy: 'multi',
videoResolutions: ['720p', '1080p'],
hardwareAcceleration: 'auto',
threads: 4,
hlsEnabled: true,
hlsResolutions: ['720p', '1080p', '2160p'],
},
turntable: {
hasPassword: false,
isEnvConfigured: false,
url: undefined,
username: undefined,
},
gotenberg: {
hasPassword: false,
isEnvConfigured: false,
url: undefined,
username: undefined,
},
appearance: { hideAgent: true },
semanticSearchEnabled: false,
})
@@ -501,4 +573,64 @@ describe('TeamService', () => {
expect(result).toBe(false)
})
})
it('should not leak emailNotification or credentials in getSettings', async () => {
const team = await prisma.team.create({
data: {
name: 'Team With Credentials',
settings: {
transcode: {
videoStrategy: 'best_match',
hardwareAcceleration: 'off',
threads: 4,
},
appearance: {
hideAgent: true,
},
emailNotification: {
enabled: true,
host: 'smtp.secret.com',
port: 587,
password: 'super-secret-smtp-password',
from: 'noreply@secret.com',
},
},
},
})
const settings = await teamService.getSettings(team.id)
expect(settings).toBeDefined()
expect(settings.transcode?.threads).toBe(4)
expect(settings.appearance?.hideAgent).toBe(true)
expect((settings as Record<string, unknown>).emailNotification).toBeUndefined()
})
it('should report semanticSearchEnabled false when ENABLE_EMBEDDING_AGENT is false even if agent is enabled in db', async () => {
const owner = await prisma.user.create({
data: { name: 'Owner', email: 'owner-emb-false@example.com', password: 'pw' },
})
const team = await teamService.createTeam(owner, { name: 'Emb False Team' })
const bot = await prisma.user.create({
data: { name: 'Bot', email: 'bot-emb-false@example.com', password: 'pw', type: 'agent' },
})
await prisma.agent.create({
data: {
id: bot.id,
teamId: team.id,
type: 'embedding',
enabled: true,
config: { provider: '', model: '' },
},
})
process.env.ENABLE_EMBEDDING_AGENT = 'false'
try {
const settings = await teamService.getSettings(team.id)
expect(settings.semanticSearchEnabled).toBe(false)
} finally {
delete process.env.ENABLE_EMBEDDING_AGENT
}
})
})
+92 -9
View File
@@ -6,7 +6,10 @@ import { notificationService } from '@shumai/core/src/notification/notification'
import { getAvatarUrl } from '@shumai/core/src/user/avatar'
import { HTTPException } from 'hono/http-exception'
import { getAllowedAgentRoles } from '@shumai/core/src/agent/permissions'
import { AgentService, agentService } from '@shumai/core/src/agent/agent'
import { sandboxService } from '@shumai/core/src/sandbox/sandbox-service'
import { turntableService } from '@shumai/core/src/turntable/turntable'
import { gotenbergService } from '@shumai/core/src/gotenberg/gotenberg'
import {
ServiceCreateTeamRequest,
ServiceGetUserTeamsRequest,
@@ -33,12 +36,18 @@ export class TeamService {
data: {
name: 'Default Team',
settings: {
transcode: { videoStrategy: 'best_match', hardwareAcceleration: 'off' },
transcode: {
videoStrategy: 'best_match',
videoResolutions: ['480p', '720p', '1080p', '1440p', '2160p'],
hardwareAcceleration: 'off',
threads: 0,
},
},
sandbox: { create: {} },
},
})
await providerService.initBuiltinProviders(team.id)
await agentService.ensureEmbeddingAgentForTeam(team.id)
}
return team
@@ -79,10 +88,12 @@ export class TeamService {
},
})
// We use 'any' here because ProviderService expects the full PrismaClient,
// We use 'any' here because ProviderService/AgentService expect the full PrismaClient,
// but the TransactionClient is compatible with the methods we use.
// eslint-disable-next-line @typescript-eslint/no-explicit-any
await new ProviderService(tx as any).initBuiltinProviders(team.id)
// eslint-disable-next-line @typescript-eslint/no-explicit-any
await new AgentService(tx as any).ensureEmbeddingAgentForTeam(team.id)
return this.toTeamInfo(updatedTeam)
})
@@ -294,19 +305,56 @@ export class TeamService {
where: { id: teamId },
})
if (!team) throw new HTTPException(404, { message: 'team not found' })
const settings = (team.settings || {}) as TeamSettingsResponse
const rawSettings = { ...(team.settings || {}) } as PrismaJson.Settings & {
[key: string]: unknown
}
delete rawSettings.emailNotification
const settings = rawSettings as TeamSettingsResponse
const embeddingAgent = await prisma.agent.findFirst({
where: { teamId, type: 'embedding', enabled: true },
})
const embeddingAgent =
process.env.ENABLE_EMBEDDING_AGENT === 'false'
? null
: await prisma.agent.findFirst({
where: { teamId, type: 'embedding', enabled: true },
})
settings.semanticSearchEnabled = !!embeddingAgent
if (!settings.transcode) {
settings.transcode = {
videoStrategy: 'best_match',
videoResolutions: ['480p', '720p', '1080p', '1440p', '2160p'],
hardwareAcceleration: 'off',
threads: 0,
hlsEnabled: false,
hlsResolutions: ['480p', '720p', '1080p'],
}
} else {
const strategyStr = settings.transcode.videoStrategy as string | undefined
if (strategyStr === 'single' || strategyStr === 'disable') {
settings.transcode.videoStrategy = 'best_match'
} else if (strategyStr === 'full' || strategyStr === 'all') {
settings.transcode.videoStrategy = 'multi'
}
if (!settings.transcode.videoResolutions) {
settings.transcode.videoResolutions = ['480p', '720p', '1080p', '1440p', '2160p']
}
if (settings.transcode.hlsEnabled === undefined) {
settings.transcode.hlsEnabled = false
}
if (!settings.transcode.hlsResolutions) {
settings.transcode.hlsResolutions = ['480p', '720p', '1080p']
}
}
if (!settings.appearance) {
settings.appearance = { hideAgent: false }
} else if (settings.appearance.hideAgent === undefined) {
settings.appearance.hideAgent = false
}
settings.turntable = await turntableService.getSettings(teamId)
settings.gotenberg = await gotenbergService.getSettings(teamId)
return settings
}
@@ -320,10 +368,25 @@ export class TeamService {
if (!settings.transcode) {
settings.transcode = {}
}
settings.transcode.videoStrategy = value as NonNullable<
const rawStrategy = value as string
let strategy = rawStrategy
if (rawStrategy === 'single' || rawStrategy === 'disable') {
strategy = 'best_match'
} else if (rawStrategy === 'full' || rawStrategy === 'all') {
strategy = 'multi'
}
settings.transcode.videoStrategy = strategy as NonNullable<
TeamSettingsResponse['transcode']
>['videoStrategy']
delete settings['transcode.videoStrategy']
} else if (key === 'transcode.videoResolutions') {
if (!settings.transcode) {
settings.transcode = {}
}
settings.transcode.videoResolutions = value as NonNullable<
TeamSettingsResponse['transcode']
>['videoResolutions']
delete settings['transcode.videoResolutions']
} else if (key === 'transcode.hardwareAcceleration') {
if (!settings.transcode) {
settings.transcode = {}
@@ -332,6 +395,26 @@ export class TeamService {
TeamSettingsResponse['transcode']
>['hardwareAcceleration']
delete settings['transcode.hardwareAcceleration']
} else if (key === 'transcode.threads') {
if (!settings.transcode) {
settings.transcode = {}
}
settings.transcode.threads = Number(value)
delete settings['transcode.threads']
} else if (key === 'transcode.hlsEnabled') {
if (!settings.transcode) {
settings.transcode = {}
}
settings.transcode.hlsEnabled = Boolean(value)
delete settings['transcode.hlsEnabled']
} else if (key === 'transcode.hlsResolutions') {
if (!settings.transcode) {
settings.transcode = {}
}
settings.transcode.hlsResolutions = value as NonNullable<
TeamSettingsResponse['transcode']
>['hlsResolutions']
delete settings['transcode.hlsResolutions']
} else if (key === 'appearance.hideAgent') {
if (!settings.appearance) {
settings.appearance = {}
@@ -342,12 +425,12 @@ export class TeamService {
settings[key] = value
}
const updated = await prisma.team.update({
await prisma.team.update({
where: { id: teamId },
data: { settings: settings as unknown as PrismaJson.Settings },
})
return (updated.settings || {}) as TeamSettingsResponse
return this.getSettings(teamId)
}
async getSandboxSettings(teamId: string): Promise<SandboxSettings> {
@@ -0,0 +1,737 @@
import { describe, it, expect, vi, beforeEach } from 'vitest'
import * as fs from 'fs'
import {
extractEmbeddedJpeg,
validateExtractedJpeg,
validateExtractedImage,
decodeRawWithDcraw,
extractAndValidateRawPreview,
EXIF_ORIENTATION_TO_ROTATION,
withTimeout,
} from './raw-extract'
// Mock exiftool-vendored
vi.mock('exiftool-vendored', () => ({
exiftool: {
extractBinaryTag: vi.fn(),
read: vi.fn(),
},
}))
// Mock sharp
vi.mock('sharp', () => {
const mockMetadata = vi.fn()
const mockSharp = vi.fn(() => ({
metadata: mockMetadata,
}))
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(mockSharp as any)._mockMetadata = mockMetadata
return { default: mockSharp }
})
// Mock child_process
vi.mock('child_process', () => ({
execFile: vi.fn(),
}))
import { exiftool } from 'exiftool-vendored'
import sharp from 'sharp'
import { execFile } from 'child_process'
// eslint-disable-next-line @typescript-eslint/no-explicit-any
const mockMetadata = (sharp as any)._mockMetadata as ReturnType<typeof vi.fn>
const mockExtractBinaryTag = vi.mocked(exiftool.extractBinaryTag)
const mockRead = vi.mocked(exiftool.read)
const mockExecFile = vi.mocked(execFile)
describe('extractEmbeddedJpeg', () => {
beforeEach(() => {
vi.clearAllMocks()
})
it('extracts from JpgFromRaw2 first to a temporary file', async () => {
mockExtractBinaryTag.mockImplementationOnce(async (_tag, _src, dest) => {
fs.writeFileSync(dest, 'jpeg-data')
})
mockRead.mockResolvedValueOnce({ Orientation: 1 } as never)
const result = await extractEmbeddedJpeg('/path/to/photo.cr2')
expect(result).not.toBeNull()
expect(fs.existsSync(result!.previewPath)).toBe(true)
expect(result!.orientation).toBe(1)
expect(mockExtractBinaryTag).toHaveBeenCalledWith(
'JpgFromRaw2',
'/path/to/photo.cr2',
expect.stringContaining('preview-JpgFromRaw2'),
)
// Test cleanup
result!.cleanup()
expect(fs.existsSync(result!.previewPath)).toBe(false)
})
it('falls back to JpgFromRaw when JpgFromRaw2 fails', async () => {
mockExtractBinaryTag
.mockRejectedValueOnce(new Error('0 output files created'))
.mockImplementationOnce(async (_tag, _src, dest) => {
fs.writeFileSync(dest, 'jpeg-data')
})
mockRead.mockResolvedValueOnce({ Orientation: 6 } as never)
const result = await extractEmbeddedJpeg('/path/to/photo.nef')
expect(result).not.toBeNull()
expect(fs.existsSync(result!.previewPath)).toBe(true)
expect(result!.orientation).toBe(6)
expect(mockExtractBinaryTag).toHaveBeenCalledTimes(2)
expect(mockExtractBinaryTag).toHaveBeenNthCalledWith(
2,
'JpgFromRaw',
'/path/to/photo.nef',
expect.stringContaining('preview-JpgFromRaw'),
)
result!.cleanup()
expect(fs.existsSync(result!.previewPath)).toBe(false)
})
it('falls back to PreviewImage when JpgFromRaw2 and JpgFromRaw fail', async () => {
mockExtractBinaryTag
.mockRejectedValueOnce(new Error('0 output files created'))
.mockRejectedValueOnce(new Error('0 output files created'))
.mockImplementationOnce(async (_tag, _src, dest) => {
fs.writeFileSync(dest, 'preview-data')
})
mockRead.mockResolvedValueOnce({ Orientation: 3 } as never)
const result = await extractEmbeddedJpeg('/path/to/photo.arw')
expect(result).not.toBeNull()
expect(fs.existsSync(result!.previewPath)).toBe(true)
expect(result!.orientation).toBe(3)
expect(mockExtractBinaryTag).toHaveBeenCalledTimes(3)
expect(mockExtractBinaryTag).toHaveBeenNthCalledWith(
3,
'PreviewImage',
'/path/to/photo.arw',
expect.stringContaining('preview-PreviewImage'),
)
result!.cleanup()
expect(fs.existsSync(result!.previewPath)).toBe(false)
})
it('returns null and cleans up when all tags fail', async () => {
mockExtractBinaryTag
.mockRejectedValueOnce(new Error('0 output files created'))
.mockRejectedValueOnce(new Error('0 output files created'))
.mockRejectedValueOnce(new Error('0 output files created'))
const result = await extractEmbeddedJpeg('/path/to/photo.cr2')
expect(result).toBeNull()
expect(mockExtractBinaryTag).toHaveBeenCalledTimes(3)
})
it('returns undefined orientation when EXIF read fails', async () => {
mockExtractBinaryTag.mockImplementationOnce(async (_tag, _src, dest) => {
fs.writeFileSync(dest, 'jpeg-data')
})
mockRead.mockRejectedValueOnce(new Error('EXIF read failed'))
const result = await extractEmbeddedJpeg('/path/to/photo.cr2')
expect(result).not.toBeNull()
expect(result!.orientation).toBeUndefined()
result!.cleanup()
})
it('returns undefined orientation when orientation value is out of range', async () => {
mockExtractBinaryTag.mockImplementationOnce(async (_tag, _src, dest) => {
fs.writeFileSync(dest, 'jpeg-data')
})
mockRead.mockResolvedValueOnce({ Orientation: 9 } as never)
const result = await extractEmbeddedJpeg('/path/to/photo.cr2')
expect(result).not.toBeNull()
expect(result!.orientation).toBeUndefined()
result!.cleanup()
})
it('returns undefined orientation when orientation is not a number', async () => {
mockExtractBinaryTag.mockImplementationOnce(async (_tag, _src, dest) => {
fs.writeFileSync(dest, 'jpeg-data')
})
mockRead.mockResolvedValueOnce({ Orientation: 'Horizontal' } as never)
const result = await extractEmbeddedJpeg('/path/to/photo.cr2')
expect(result).not.toBeNull()
expect(result!.orientation).toBeUndefined()
result!.cleanup()
})
it('skips remaining tags and aborts early when extraction times out', async () => {
const origEnv = process.env.EXIFTOOL_TIMEOUT_MS
try {
process.env.EXIFTOOL_TIMEOUT_MS = '20'
// Hang on first tag
mockExtractBinaryTag.mockImplementationOnce(
() => new Promise((resolve) => setTimeout(resolve, 300)),
)
const result = await extractEmbeddedJpeg('/path/to/photo.cr2')
expect(result).toBeNull()
// Crucial: should break immediately after the first timeout instead of trying remaining 2 tags
expect(mockExtractBinaryTag).toHaveBeenCalledTimes(1)
} finally {
if (origEnv !== undefined) {
process.env.EXIFTOOL_TIMEOUT_MS = origEnv
} else {
delete process.env.EXIFTOOL_TIMEOUT_MS
}
}
})
})
describe('validateExtractedJpeg', () => {
beforeEach(() => {
vi.clearAllMocks()
})
it('returns dimensions for a valid JPEG file', async () => {
mockMetadata.mockResolvedValueOnce({ width: 6000, height: 4000 })
const result = await validateExtractedJpeg('/tmp/valid-jpeg.jpg')
expect(result).toEqual({ width: 6000, height: 4000 })
})
it('returns null when metadata has zero width', async () => {
mockMetadata.mockResolvedValueOnce({ width: 0, height: 4000 })
const result = await validateExtractedJpeg('/tmp/bad-jpeg.jpg')
expect(result).toBeNull()
})
it('returns null when metadata has no dimensions', async () => {
mockMetadata.mockResolvedValueOnce({})
const result = await validateExtractedJpeg('/tmp/bad-jpeg.jpg')
expect(result).toBeNull()
})
it('returns null when sharp throws', async () => {
mockMetadata.mockRejectedValueOnce(new Error('Invalid image'))
const result = await validateExtractedJpeg('/tmp/corrupt-jpeg.jpg')
expect(result).toBeNull()
})
})
describe('validateExtractedImage', () => {
it('is an alias for validateExtractedJpeg', () => {
expect(validateExtractedImage).toBe(validateExtractedJpeg)
})
})
describe('decodeRawWithDcraw', () => {
beforeEach(() => {
vi.clearAllMocks()
})
it('decodes RAW using camera white balance, sRGB, and TIFF output without -h when longest dimension <= 8192', async () => {
mockRead.mockResolvedValueOnce({
ImageWidth: 6000,
ImageHeight: 4000,
Orientation: 1,
} as never)
let executedArgs: string[] = []
mockExecFile.mockImplementationOnce((_bin, args, _opts, cb) => {
executedArgs = args as string[]
const tiffDest = (args as string[])[(args as string[]).indexOf('-Z') + 1]
fs.writeFileSync(tiffDest, 'tiff-data')
const callback = typeof _opts === 'function' ? _opts : cb
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(null, { stdout: '', stderr: '' })
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
const result = await decodeRawWithDcraw('/path/to/photo.cr2')
expect(result).not.toBeNull()
expect(fs.existsSync(result!.previewPath)).toBe(true)
expect(result!.orientation).toBeUndefined()
expect(result!.rawWidth).toBe(6000)
expect(result!.rawHeight).toBe(4000)
expect(executedArgs).toContain('-w')
expect(executedArgs).toContain('-o')
expect(executedArgs).toContain('1')
expect(executedArgs).toContain('-T')
expect(executedArgs).toContain('-Z')
expect(executedArgs).not.toContain('-h')
expect(executedArgs[executedArgs.length - 1]).toBe('/path/to/photo.cr2')
result!.cleanup()
expect(fs.existsSync(result!.previewPath)).toBe(false)
})
it('enables half-size (-h) when longest dimension > 8192px and computes upright raw dimensions', async () => {
mockRead.mockResolvedValueOnce({
ImageWidth: 9504,
ImageHeight: 6336,
Orientation: 6,
} as never)
let executedArgs: string[] = []
mockExecFile.mockImplementationOnce((_bin, args, _opts, cb) => {
executedArgs = args as string[]
const tiffDest = (args as string[])[(args as string[]).indexOf('-Z') + 1]
fs.writeFileSync(tiffDest, 'tiff-data')
const callback = typeof _opts === 'function' ? _opts : cb
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(null, { stdout: '', stderr: '' })
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
const result = await decodeRawWithDcraw('/path/to/photo.arw')
expect(result).not.toBeNull()
expect(executedArgs).toContain('-h')
expect(result!.orientation).toBeUndefined()
// For orientation 6 (90 CW), upright raw dimensions are swapped
expect(result!.rawWidth).toBe(6336)
expect(result!.rawHeight).toBe(9504)
result!.cleanup()
})
it('omits -h when longest dimension is exactly 8192px', async () => {
mockRead.mockResolvedValueOnce({
ImageWidth: 8192,
ImageHeight: 5464,
Orientation: 1,
} as never)
let executedArgs: string[] = []
mockExecFile.mockImplementationOnce((_bin, args, _opts, cb) => {
executedArgs = args as string[]
const tiffDest = (args as string[])[(args as string[]).indexOf('-Z') + 1]
fs.writeFileSync(tiffDest, 'tiff-data')
const callback = typeof _opts === 'function' ? _opts : cb
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(null, { stdout: '', stderr: '' })
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
const result = await decodeRawWithDcraw('/path/to/photo.nef')
expect(result).not.toBeNull()
expect(executedArgs).not.toContain('-h')
result!.cleanup()
})
it('defaults to full size without -h when EXIF metadata cannot be read', async () => {
mockRead.mockRejectedValueOnce(new Error('Exif read failure'))
let executedArgs: string[] = []
mockExecFile.mockImplementationOnce((_bin, args, _opts, cb) => {
executedArgs = args as string[]
const tiffDest = (args as string[])[(args as string[]).indexOf('-Z') + 1]
fs.writeFileSync(tiffDest, 'tiff-data')
const callback = typeof _opts === 'function' ? _opts : cb
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(null, { stdout: '', stderr: '' })
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
const result = await decodeRawWithDcraw('/path/to/photo.dng')
expect(result).not.toBeNull()
expect(executedArgs).not.toContain('-h')
result!.cleanup()
})
it('defaults to full size without -h when EXIF metadata read times out', async () => {
const origEnv = process.env.EXIFTOOL_TIMEOUT_MS
try {
process.env.EXIFTOOL_TIMEOUT_MS = '20'
mockRead.mockImplementationOnce(() => new Promise((resolve) => setTimeout(resolve, 300)))
let executedArgs: string[] = []
mockExecFile.mockImplementationOnce((_bin, args, _opts, cb) => {
executedArgs = args as string[]
const tiffDest = (args as string[])[(args as string[]).indexOf('-Z') + 1]
fs.writeFileSync(tiffDest, 'tiff-data')
const callback = typeof _opts === 'function' ? _opts : cb
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(null, { stdout: '', stderr: '' })
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
const result = await decodeRawWithDcraw('/path/to/photo.dng')
expect(result).not.toBeNull()
expect(executedArgs).not.toContain('-h')
result!.cleanup()
} finally {
if (origEnv !== undefined) {
process.env.EXIFTOOL_TIMEOUT_MS = origEnv
} else {
delete process.env.EXIFTOOL_TIMEOUT_MS
}
}
})
it('returns null and cleans up when dcraw_emu exits with non-zero code', async () => {
mockRead.mockResolvedValueOnce({ ImageWidth: 6000, ImageHeight: 4000 } as never)
mockExecFile.mockImplementationOnce((_bin, _args, _opts, cb) => {
const callback = typeof _opts === 'function' ? _opts : cb
const err = new Error('Command failed: dcraw_emu')
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(err as any).code = 1
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(err)
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
const result = await decodeRawWithDcraw('/path/to/photo.cr2')
expect(result).toBeNull()
})
it('returns null and cleans up when dcraw_emu times out', async () => {
mockRead.mockResolvedValueOnce({ ImageWidth: 6000, ImageHeight: 4000 } as never)
mockExecFile.mockImplementationOnce((_bin, _args, _opts, cb) => {
const callback = typeof _opts === 'function' ? _opts : cb
const err = new Error('Process timed out')
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(err as any).code = 'ETIMEDOUT'
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(err)
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
const result = await decodeRawWithDcraw('/path/to/photo.cr2')
expect(result).toBeNull()
})
it('returns null and cleans up when dcraw_emu is not found (ENOENT)', async () => {
mockRead.mockResolvedValueOnce({ ImageWidth: 6000, ImageHeight: 4000 } as never)
mockExecFile.mockImplementationOnce((_bin, _args, _opts, cb) => {
const callback = typeof _opts === 'function' ? _opts : cb
const err = new Error('spawn dcraw_emu ENOENT')
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(err as any).code = 'ENOENT'
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(err)
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
const result = await decodeRawWithDcraw('/path/to/photo.cr2')
expect(result).toBeNull()
})
it('returns null and cleans up when output TIFF is missing or empty', async () => {
mockRead.mockResolvedValueOnce({ ImageWidth: 6000, ImageHeight: 4000 } as never)
mockExecFile.mockImplementationOnce((_bin, _args, _opts, cb) => {
const callback = typeof _opts === 'function' ? _opts : cb
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(null, { stdout: '', stderr: '' })
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
const result = await decodeRawWithDcraw('/path/to/photo.cr2')
expect(result).toBeNull()
})
})
describe('extractAndValidateRawPreview', () => {
beforeEach(() => {
vi.clearAllMocks()
})
it('returns previewPath, dimensions, orientation, and cleanup for valid RAW using embedded JPEG first', async () => {
mockExtractBinaryTag.mockImplementationOnce(async (_tag, _src, dest) => {
fs.writeFileSync(dest, 'jpeg-data')
})
mockRead.mockResolvedValueOnce({ Orientation: 6, ImageWidth: 4000, ImageHeight: 3000 } as never)
mockMetadata.mockResolvedValueOnce({ width: 4000, height: 3000 })
const result = await extractAndValidateRawPreview('/path/to/photo.cr2')
expect(result).not.toBeNull()
expect(fs.existsSync(result!.previewPath)).toBe(true)
expect(result!.width).toBe(4000)
expect(result!.height).toBe(3000)
expect(result!.orientation).toBe(6)
expect(mockExecFile).not.toHaveBeenCalled()
result!.cleanup()
expect(fs.existsSync(result!.previewPath)).toBe(false)
})
it('falls back to dcraw_emu when embedded preview extraction fails', async () => {
mockExtractBinaryTag
.mockRejectedValueOnce(new Error('fail'))
.mockRejectedValueOnce(new Error('fail'))
.mockRejectedValueOnce(new Error('fail'))
mockRead.mockResolvedValueOnce({
ImageWidth: 6000,
ImageHeight: 4000,
Orientation: 1,
} as never)
mockExecFile.mockImplementationOnce((_bin, args, _opts, cb) => {
const tiffDest = (args as string[])[(args as string[]).indexOf('-Z') + 1]
fs.writeFileSync(tiffDest, 'tiff-data')
const callback = typeof _opts === 'function' ? _opts : cb
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(null, { stdout: '', stderr: '' })
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
mockMetadata.mockResolvedValueOnce({ width: 6000, height: 4000 })
const result = await extractAndValidateRawPreview('/path/to/photo.cr2')
expect(result).not.toBeNull()
expect(mockExecFile).toHaveBeenCalled()
expect(result!.width).toBe(6000)
expect(result!.height).toBe(4000)
expect(result!.rawWidth).toBe(6000)
expect(result!.rawHeight).toBe(4000)
expect(result!.orientation).toBeUndefined()
result!.cleanup()
expect(fs.existsSync(result!.previewPath)).toBe(false)
})
it('falls back to dcraw_emu when extracted embedded JPEG is corrupt', async () => {
let createdJpegPath = ''
mockExtractBinaryTag.mockImplementationOnce(async (_tag, _src, dest) => {
createdJpegPath = dest
fs.writeFileSync(dest, 'corrupt-jpeg')
})
mockRead.mockResolvedValueOnce({ Orientation: 1 } as never)
mockMetadata.mockRejectedValueOnce(new Error('Not a valid image'))
// Fallback dcraw_emu
mockRead.mockResolvedValueOnce({ ImageWidth: 6000, ImageHeight: 4000 } as never)
mockExecFile.mockImplementationOnce((_bin, args, _opts, cb) => {
const tiffDest = (args as string[])[(args as string[]).indexOf('-Z') + 1]
fs.writeFileSync(tiffDest, 'tiff-data')
const callback = typeof _opts === 'function' ? _opts : cb
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(null, { stdout: '', stderr: '' })
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
mockMetadata.mockResolvedValueOnce({ width: 6000, height: 4000 })
const result = await extractAndValidateRawPreview('/path/to/photo.cr2')
expect(result).not.toBeNull()
expect(fs.existsSync(createdJpegPath)).toBe(false) // Embedded jpeg was cleaned up
expect(fs.existsSync(result!.previewPath)).toBe(true)
result!.cleanup()
expect(fs.existsSync(result!.previewPath)).toBe(false)
})
it('returns null when both embedded extraction and dcraw_emu fail', async () => {
mockExtractBinaryTag
.mockRejectedValueOnce(new Error('fail'))
.mockRejectedValueOnce(new Error('fail'))
.mockRejectedValueOnce(new Error('fail'))
mockRead.mockRejectedValueOnce(new Error('fail'))
mockExecFile.mockImplementationOnce((_bin, _args, _opts, cb) => {
const callback = typeof _opts === 'function' ? _opts : cb
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(new Error('dcraw failed'))
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
const result = await extractAndValidateRawPreview('/path/to/photo.cr2')
expect(result).toBeNull()
})
it('returns null and cleans up when decoded TIFF is corrupt', async () => {
mockExtractBinaryTag
.mockRejectedValueOnce(new Error('fail'))
.mockRejectedValueOnce(new Error('fail'))
.mockRejectedValueOnce(new Error('fail'))
let createdTiff = ''
mockRead.mockResolvedValueOnce({ ImageWidth: 6000, ImageHeight: 4000 } as never)
mockExecFile.mockImplementationOnce((_bin, args, _opts, cb) => {
createdTiff = (args as string[])[(args as string[]).indexOf('-Z') + 1]
fs.writeFileSync(createdTiff, 'corrupt-tiff')
const callback = typeof _opts === 'function' ? _opts : cb
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(null, { stdout: '', stderr: '' })
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
mockMetadata.mockRejectedValueOnce(new Error('Invalid TIFF'))
const result = await extractAndValidateRawPreview('/path/to/photo.cr2')
expect(result).toBeNull()
expect(fs.existsSync(createdTiff)).toBe(false)
})
it('falls back to dcraw_emu when embedded preview extraction times out', async () => {
const origEnv = process.env.EXIFTOOL_TIMEOUT_MS
try {
process.env.EXIFTOOL_TIMEOUT_MS = '20'
// Hang on extractBinaryTag
mockExtractBinaryTag.mockImplementationOnce(
() => new Promise((resolve) => setTimeout(resolve, 300)),
)
mockRead.mockResolvedValueOnce({
ImageWidth: 6000,
ImageHeight: 4000,
Orientation: 1,
} as never)
mockExecFile.mockImplementationOnce((_bin, args, _opts, cb) => {
const tiffDest = (args as string[])[(args as string[]).indexOf('-Z') + 1]
fs.writeFileSync(tiffDest, 'tiff-data')
const callback = typeof _opts === 'function' ? _opts : cb
if (callback) {
// eslint-disable-next-line @typescript-eslint/no-explicit-any
;(callback as any)(null, { stdout: '', stderr: '' })
}
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return {} as any
})
mockMetadata.mockResolvedValueOnce({ width: 6000, height: 4000 })
const result = await extractAndValidateRawPreview('/path/to/photo.cr2')
expect(result).not.toBeNull()
expect(mockExecFile).toHaveBeenCalled()
expect(result!.width).toBe(6000)
expect(result!.height).toBe(4000)
result!.cleanup()
} finally {
if (origEnv !== undefined) {
process.env.EXIFTOOL_TIMEOUT_MS = origEnv
} else {
delete process.env.EXIFTOOL_TIMEOUT_MS
}
}
})
})
describe('withTimeout', () => {
it('resolves when promise settles before timeout', async () => {
const result = await withTimeout(Promise.resolve('ok'), 1000)
expect(result).toBe('ok')
})
it('rejects with timeout error when promise exceeds timeout', async () => {
const hanging = new Promise((resolve) => setTimeout(resolve, 500))
await expect(withTimeout(hanging, 20, 'TestTask')).rejects.toThrow(
'TestTask timed out after 20ms',
)
})
it('propagates underlying promise rejection before timeout', async () => {
const failing = Promise.reject(new Error('fail fast'))
await expect(withTimeout(failing, 1000, 'TestTask')).rejects.toThrow('fail fast')
})
it('prevents unhandled rejection if promise rejects after timeout', async () => {
const delayedReject = new Promise((_, reject) =>
setTimeout(() => reject(new Error('late error')), 40),
)
await expect(withTimeout(delayedReject, 10, 'TestTask')).rejects.toThrow(
'TestTask timed out after 10ms',
)
await new Promise((resolve) => setTimeout(resolve, 60))
})
})
describe('EXIF_ORIENTATION_TO_ROTATION', () => {
it('maps all 8 EXIF orientations', () => {
expect(Object.keys(EXIF_ORIENTATION_TO_ROTATION)).toHaveLength(8)
for (let i = 1; i <= 8; i++) {
expect(EXIF_ORIENTATION_TO_ROTATION[i]).toBeDefined()
}
})
it('has correct rotation for landscape (1)', () => {
expect(EXIF_ORIENTATION_TO_ROTATION[1]).toEqual({})
})
it('has correct rotation for 90° CW (6)', () => {
expect(EXIF_ORIENTATION_TO_ROTATION[6]).toEqual({ angle: 90 })
})
it('has correct rotation for 180° (3)', () => {
expect(EXIF_ORIENTATION_TO_ROTATION[3]).toEqual({ angle: 180 })
})
it('has correct rotation for 270° CW (8)', () => {
expect(EXIF_ORIENTATION_TO_ROTATION[8]).toEqual({ angle: 270 })
})
it('has correct rotation for mirror horizontal (2)', () => {
expect(EXIF_ORIENTATION_TO_ROTATION[2]).toEqual({ flop: true })
})
})
+395
View File
@@ -0,0 +1,395 @@
import { execFile } from 'child_process'
import { exiftool } from 'exiftool-vendored'
import * as fs from 'fs'
import * as os from 'os'
import * as path from 'path'
import sharp from 'sharp'
import { promisify } from 'util'
import { logger } from '@shumai/core/src/logger'
const execFileAsync = promisify(execFile)
export interface RawExtractResult {
previewPath: string
cleanup: () => void
/** EXIF orientation from the RAW container (1–8), or undefined if absent or already applied. */
orientation?: number
/** Original unoriented or oriented image dimensions from EXIF if available */
rawWidth?: number
rawHeight?: number
}
export interface RawPreviewResult {
previewPath: string
cleanup: () => void
width: number
height: number
orientation?: number
/**
* True when the file is already upright (dcraw_emu output), so downstream sharp must not
* auto-orient it even though `orientation` is undefined.
*/
orientationApplied?: boolean
rawWidth?: number
rawHeight?: number
}
export interface DecodeRawOptions {
timeoutMs?: number
signal?: AbortSignal
}
/**
* Maps EXIF orientation values (1–8) to Sharp rotation operations.
* Follows the same mapping as Immich's ORIENTATION_TO_SHARP_ROTATION.
*/
export const EXIF_ORIENTATION_TO_ROTATION: Record<
number,
{ angle?: number; flip?: boolean; flop?: boolean }
> = {
1: {}, // Horizontal (normal)
2: { flop: true }, // Mirror horizontal
3: { angle: 180 }, // Rotate 180°
4: { angle: 180, flop: true }, // Mirror vertical
5: { angle: 270, flip: true }, // Mirror horizontal + rotate 270° CW
6: { angle: 90 }, // Rotate 90° CW
7: { angle: 90, flip: true }, // Mirror horizontal + rotate 90° CW
8: { angle: 270 }, // Rotate 270° CW
}
const DEFAULT_EXIFTOOL_TIMEOUT_MS = 15_000
function getExifToolTimeoutMs(): number {
const envVal = process.env.EXIFTOOL_TIMEOUT_MS
if (envVal) {
const parsed = parseInt(envVal, 10)
if (!isNaN(parsed) && parsed > 0) {
return parsed
}
}
return DEFAULT_EXIFTOOL_TIMEOUT_MS
}
/**
* Wraps a promise with a timeout. If the timeout triggers first,
* the returned promise rejects with a timeout error and attaches a no-op handler
* to the underlying promise to prevent unhandled rejections.
*/
export function withTimeout<T>(
promise: Promise<T>,
timeoutMs: number,
operationName = 'Operation',
): Promise<T> {
let timer: ReturnType<typeof setTimeout> | undefined
const timeoutPromise = new Promise<never>((_, reject) => {
timer = setTimeout(() => {
reject(new Error(`${operationName} timed out after ${timeoutMs}ms`))
}, timeoutMs)
})
promise.catch(() => {})
return Promise.race([promise, timeoutPromise]).finally(() => {
if (timer) clearTimeout(timer)
})
}
/**
* Attempt to extract an embedded JPEG preview from a camera RAW file to a temporary file.
*
* Tries tags in priority order (matching Immich, skipping PreviewJXL):
* 1. JpgFromRaw2 — secondary full-resolution JPEG (newer Sony/Canon/Nikon)
* 2. JpgFromRaw — primary full-size embedded JPEG
* 3. PreviewImage — smaller embedded preview fallback
*
* Also reads the EXIF orientation from the RAW container so callers can
* apply it to the extracted buffer/file (which often lacks its own orientation tag).
*
* Returns null if no usable embedded JPEG exists.
*/
export async function extractEmbeddedJpeg(
rawFilePath: string,
targetDir?: string,
): Promise<RawExtractResult | null> {
const tags = ['JpgFromRaw2', 'JpgFromRaw', 'PreviewImage'] as const
const ownTempDir = !targetDir
const workDir = targetDir ?? fs.mkdtempSync(path.join(os.tmpdir(), 'raw-extract-'))
const cleanup = () => {
try {
if (ownTempDir && fs.existsSync(workDir)) {
fs.rmSync(workDir, { recursive: true, force: true })
}
} catch (err) {
logger.warn({ workDir, err }, 'Failed to clean up raw extraction temporary directory')
}
}
const exiftoolTimeoutMs = getExifToolTimeoutMs()
for (const tag of tags) {
const candidatePath = path.join(workDir, `preview-${tag}-${Date.now()}.jpg`)
try {
await withTimeout(
exiftool.extractBinaryTag(tag, rawFilePath, candidatePath),
exiftoolTimeoutMs,
`exiftool.extractBinaryTag(${tag})`,
)
if (fs.existsSync(candidatePath) && fs.statSync(candidatePath).size > 0) {
logger.debug({ tag, rawFilePath, candidatePath }, 'Extracted embedded JPEG from RAW')
// Read orientation and container dimensions from the RAW container's EXIF metadata.
let orientation: number | undefined
let rawWidth: number | undefined
let rawHeight: number | undefined
try {
const exif = await withTimeout(
exiftool.read(rawFilePath),
exiftoolTimeoutMs,
'exiftool.read',
)
const rawOrientation = exif.Orientation
if (typeof rawOrientation === 'number' && rawOrientation >= 1 && rawOrientation <= 8) {
orientation = rawOrientation
}
const w =
typeof exif.ImageWidth === 'number'
? exif.ImageWidth
: typeof exif.ExifImageWidth === 'number'
? exif.ExifImageWidth
: undefined
const h =
typeof exif.ImageHeight === 'number'
? exif.ImageHeight
: typeof exif.ExifImageHeight === 'number'
? exif.ExifImageHeight
: undefined
if (w && h && w > 0 && h > 0) {
rawWidth = w
rawHeight = h
}
} catch {
logger.debug({ rawFilePath }, 'Could not read EXIF orientation from RAW container')
}
return { previewPath: candidatePath, cleanup, orientation, rawWidth, rawHeight }
}
} catch (err) {
const isTimeout = err instanceof Error && err.message.includes('timed out')
logger.debug({ tag, rawFilePath, err, isTimeout }, `Could not extract ${tag} from RAW file`)
try {
if (fs.existsSync(candidatePath)) {
fs.rmSync(candidatePath, { force: true })
}
} catch {
// ignore
}
if (isTimeout) {
logger.warn(
{ rawFilePath, tag, timeoutMs: exiftoolTimeoutMs },
'ExifTool extraction timed out; skipping remaining tags to avoid further delays',
)
break
}
}
}
cleanup()
logger.info({ rawFilePath }, 'No usable embedded JPEG found in RAW file')
return null
}
/**
* Fallback RAW decoder using dcraw_emu (from LibRaw).
*
* Used when no usable embedded JPEG preview exists in the RAW file.
*
* Decoding options:
* - Camera white balance and sRGB output (-w -o 1)
* - Output temporary TIFF (-T -Z <tempPath>)
* - Half-size processing (-h) enabled only when longest dimension > 8192px
*
* LibRaw auto-rotates the output image to upright orientation based on camera EXIF,
* so the returned orientation is undefined to avoid double-rotation in downstream Sharp.
*/
export async function decodeRawWithDcraw(
rawFilePath: string,
targetDir?: string,
options?: DecodeRawOptions,
): Promise<RawExtractResult | null> {
const ownTempDir = !targetDir
const workDir = targetDir ?? fs.mkdtempSync(path.join(os.tmpdir(), 'raw-decode-'))
const tempTiffPath = path.join(
workDir,
`decoded-${Date.now()}-${Math.random().toString(36).slice(2, 8)}.tiff`,
)
const cleanup = () => {
try {
if (ownTempDir && fs.existsSync(workDir)) {
fs.rmSync(workDir, { recursive: true, force: true })
} else if (fs.existsSync(tempTiffPath)) {
fs.rmSync(tempTiffPath, { force: true })
}
} catch (err) {
logger.warn({ workDir, tempTiffPath, err }, 'Failed to clean up dcraw temporary files')
}
}
// 1. Probe RAW dimensions and orientation via ExifTool to check for 8192px threshold
let useHalfSize = false
let rawWidth: number | undefined
let rawHeight: number | undefined
try {
const exiftoolTimeoutMs = getExifToolTimeoutMs()
const exif = await withTimeout(exiftool.read(rawFilePath), exiftoolTimeoutMs, 'exiftool.read')
const w =
typeof exif.ImageWidth === 'number'
? exif.ImageWidth
: typeof exif.ExifImageWidth === 'number'
? exif.ExifImageWidth
: undefined
const h =
typeof exif.ImageHeight === 'number'
? exif.ImageHeight
: typeof exif.ExifImageHeight === 'number'
? exif.ExifImageHeight
: undefined
if (w && h && w > 0 && h > 0) {
const longest = Math.max(w, h)
if (longest > 8192) {
useHalfSize = true
}
// Compute upright raw dimensions matching dcraw_emu's auto-rotation
const rawOrientation = exif.Orientation
const isSwapped =
typeof rawOrientation === 'number' && rawOrientation >= 5 && rawOrientation <= 8
rawWidth = isSwapped ? h : w
rawHeight = isSwapped ? w : h
}
} catch (err) {
logger.debug({ rawFilePath, err }, 'Could not read EXIF metadata for RAW dimension check')
}
// 2. Build dcraw_emu arguments
const args = ['-w', '-o', '1']
if (useHalfSize) {
args.push('-h')
}
args.push('-T', '-Z', tempTiffPath, rawFilePath)
const dcrawBin = process.env.DCRAW_EMU_PATH || 'dcraw_emu'
const envTimeout = process.env.DCRAW_EMU_TIMEOUT_MS
? parseInt(process.env.DCRAW_EMU_TIMEOUT_MS, 10)
: undefined
const timeoutMs = options?.timeoutMs ?? (envTimeout && !isNaN(envTimeout) ? envTimeout : 60_000)
try {
logger.debug({ dcrawBin, args, rawFilePath }, 'Executing dcraw_emu fallback')
await execFileAsync(dcrawBin, args, {
timeout: timeoutMs,
signal: options?.signal,
})
if (!fs.existsSync(tempTiffPath) || fs.statSync(tempTiffPath).size === 0) {
logger.warn({ rawFilePath, tempTiffPath }, 'dcraw_emu did not produce a valid output TIFF')
cleanup()
return null
}
return {
previewPath: tempTiffPath,
cleanup,
orientation: undefined, // dcraw_emu already auto-rotates upright based on camera EXIF
rawWidth,
rawHeight,
}
} catch (err) {
logger.warn({ rawFilePath, err }, 'dcraw_emu failed to decode RAW file')
cleanup()
return null
}
}
/**
* Validate that an extracted or decoded image file (JPEG, TIFF, etc.) is decodable.
* Returns dimensions if valid, null if corrupt/unusable.
*/
export async function validateExtractedImage(
filePath: string,
): Promise<{ width: number; height: number } | null> {
try {
const metadata = await sharp(filePath).metadata()
if (metadata.width && metadata.height && metadata.width > 0 && metadata.height > 0) {
return { width: metadata.width, height: metadata.height }
}
return null
} catch {
return null
}
}
/** Backward compatibility alias */
export const validateExtractedJpeg = validateExtractedImage
/**
* Extract and validate a preview from a RAW file to a temporary file.
*
* 1. Attempts ExifTool embedded JPEG extraction first.
* 2. If no embedded preview is available or extraction fails, falls back to dcraw_emu decoding.
*
* Returns a validated file path + dimensions + orientation + cleanup function, or null.
*/
export async function extractAndValidateRawPreview(
rawFilePath: string,
targetDir?: string,
): Promise<RawPreviewResult | null> {
// 1. Try ExifTool embedded JPEG extraction first
const extracted = await extractEmbeddedJpeg(rawFilePath, targetDir)
if (extracted) {
const dims = await validateExtractedImage(extracted.previewPath)
if (dims) {
return {
previewPath: extracted.previewPath,
cleanup: extracted.cleanup,
orientation: extracted.orientation,
rawWidth: extracted.rawWidth,
rawHeight: extracted.rawHeight,
...dims,
}
}
logger.warn(
{ rawFilePath },
'Extracted JPEG from RAW is not decodable, falling back to dcraw_emu',
)
extracted.cleanup()
} else {
logger.info(
{ rawFilePath },
'No usable embedded JPEG found in RAW file, falling back to dcraw_emu',
)
}
// 2. Fallback to dcraw_emu
const decoded = await decodeRawWithDcraw(rawFilePath, targetDir)
if (!decoded) {
return null
}
const dims = await validateExtractedImage(decoded.previewPath)
if (!dims) {
logger.warn({ rawFilePath }, 'Decoded TIFF from dcraw_emu is not decodable, skipping preview')
decoded.cleanup()
return null
}
return {
previewPath: decoded.previewPath,
cleanup: decoded.cleanup,
orientation: decoded.orientation,
orientationApplied: true,
rawWidth: decoded.rawWidth,
rawHeight: decoded.rawHeight,
...dims,
}
}
@@ -0,0 +1,252 @@
import { describe, it, expect, beforeEach, afterEach, vi } from 'vitest'
import * as fs from 'fs'
import * as path from 'path'
import sharp from 'sharp'
import { transcodeService } from './transcode'
// Only the storage layer is mocked: these tests run the real sharp pipeline end to end.
vi.mock('@shumai/core/src/s3/s3', () => ({
s3Service: {
downloadToFile: vi.fn(),
putObject: vi.fn(),
resolveInput: vi.fn(),
presign: vi.fn(),
},
}))
// Only the RAW preview extraction is replaced, so the RAW branch of transcodeImage can be driven
// with a prepared file (the dcraw_emu fallback result) without ExifTool or dcraw_emu installed.
const extractRawMock = vi.hoisted(() => vi.fn())
vi.mock('./raw-extract', async (importOriginal) => ({
...(await importOriginal<typeof import('./raw-extract')>()),
extractAndValidateRawPreview: extractRawMock,
}))
// Windows keeps cached files open, which would make the temp dir cleanup fail with EBUSY.
sharp.cache(false)
type Corner = 'TL' | 'TR' | 'BL' | 'BR'
// Stored (sensor) pixels are landscape. Each quadrant has its own color so the direction of the
// rotation or flip can be read back from the corners of the output.
const STORED_WIDTH = 64
const STORED_HEIGHT = 48
const QUADRANT_COLORS: Record<Corner, [number, number, number]> = {
TL: [220, 30, 30], // red
TR: [30, 200, 30], // green
BL: [30, 30, 220], // blue
BR: [230, 230, 30], // yellow
}
/**
* For each EXIF orientation: the displayed size and which stored corner ends up at the displayed
* top-left and bottom-right. Orientations 5 to 8 swap width and height.
*/
const ORIENTATIONS: Array<{
orientation: number
name: string
swapped: boolean
topLeftIs: Corner
bottomRightIs: Corner
}> = [
{ orientation: 1, name: 'normal', swapped: false, topLeftIs: 'TL', bottomRightIs: 'BR' },
{
orientation: 2,
name: 'mirror horizontal',
swapped: false,
topLeftIs: 'TR',
bottomRightIs: 'BL',
},
{ orientation: 3, name: 'rotate 180', swapped: false, topLeftIs: 'BR', bottomRightIs: 'TL' },
{ orientation: 4, name: 'mirror vertical', swapped: false, topLeftIs: 'BL', bottomRightIs: 'TR' },
{ orientation: 5, name: 'transpose', swapped: true, topLeftIs: 'TL', bottomRightIs: 'BR' },
{ orientation: 6, name: 'rotate 90 CW', swapped: true, topLeftIs: 'BL', bottomRightIs: 'TR' },
{ orientation: 7, name: 'transverse', swapped: true, topLeftIs: 'BR', bottomRightIs: 'TL' },
{ orientation: 8, name: 'rotate 270 CW', swapped: true, topLeftIs: 'TR', bottomRightIs: 'BL' },
]
async function makeJpeg(orientation: number): Promise<Buffer> {
const halfW = STORED_WIDTH / 2
const halfH = STORED_HEIGHT / 2
const block = (corner: Corner) => ({
create: {
width: halfW,
height: halfH,
channels: 3 as const,
background: {
r: QUADRANT_COLORS[corner][0],
g: QUADRANT_COLORS[corner][1],
b: QUADRANT_COLORS[corner][2],
},
},
})
const tile = async (corner: Corner) => sharp(block(corner)).png().toBuffer()
return sharp({
create: {
width: STORED_WIDTH,
height: STORED_HEIGHT,
channels: 3,
background: { r: 0, g: 0, b: 0 },
},
})
.composite([
{ input: await tile('TL'), left: 0, top: 0 },
{ input: await tile('TR'), left: halfW, top: 0 },
{ input: await tile('BL'), left: 0, top: halfH },
{ input: await tile('BR'), left: halfW, top: halfH },
])
.jpeg({ quality: 100, chromaSubsampling: '4:4:4' })
.withMetadata({ orientation })
.toBuffer()
}
/** Which quadrant color is nearest to the pixel a few pixels inside the given output corner. */
async function cornerColor(file: string, where: 'TL' | 'BR'): Promise<Corner> {
const { data, info } = await sharp(fs.readFileSync(file))
.removeAlpha()
.raw()
.toBuffer({ resolveWithObject: true })
const inset = 3
const x = where === 'TL' ? inset : info.width - 1 - inset
const y = where === 'TL' ? inset : info.height - 1 - inset
const offset = (y * info.width + x) * info.channels
const pixel = [data[offset], data[offset + 1], data[offset + 2]]
let best: Corner = 'TL'
let bestDistance = Infinity
for (const corner of Object.keys(QUADRANT_COLORS) as Corner[]) {
const distance = QUADRANT_COLORS[corner].reduce(
(sum, channel, i) => sum + (channel - pixel[i]) ** 2,
0,
)
if (distance < bestDistance) {
best = corner
bestDistance = distance
}
}
return best
}
describe('EXIF orientation (real sharp)', () => {
let tempDir: string
beforeEach(() => {
tempDir = transcodeService.createTempDir('orientation-test-')
})
afterEach(() => {
transcodeService.removeDir(tempDir)
})
const expectedSize = (swapped: boolean) =>
swapped
? { width: STORED_HEIGHT, height: STORED_WIDTH }
: { width: STORED_WIDTH, height: STORED_HEIGHT }
it.each(ORIENTATIONS)(
'getImageInfo reports the displayed size for orientation $orientation ($name)',
async ({ orientation, swapped }) => {
const inputPath = path.join(tempDir, `in-${orientation}.jpg`)
fs.writeFileSync(inputPath, await makeJpeg(orientation))
const info = await transcodeService.getImageInfo(inputPath)
const expected = expectedSize(swapped)
expect(info.originalWidth).toBe(expected.width)
expect(info.originalHeight).toBe(expected.height)
},
)
it.each(ORIENTATIONS)(
'transcodeImage outputs an upright image for orientation $orientation ($name)',
async ({ orientation, swapped, topLeftIs, bottomRightIs }) => {
const inputPath = path.join(tempDir, `in-${orientation}.jpg`)
const outputPath = path.join(tempDir, `out-${orientation}.webp`)
fs.writeFileSync(inputPath, await makeJpeg(orientation))
await transcodeService.transcodeImage(inputPath, outputPath, 1000, 100)
const meta = await sharp(fs.readFileSync(outputPath)).metadata()
const expected = expectedSize(swapped)
expect(meta.width).toBe(expected.width)
expect(meta.height).toBe(expected.height)
// Direction, not just size: the stored corner that lands top-left and bottom-right.
expect(await cornerColor(outputPath, 'TL')).toBe(topLeftIs)
expect(await cornerColor(outputPath, 'BR')).toBe(bottomRightIs)
},
)
it('transcodeImage sizes a rotated preview from the displayed dimensions', async () => {
// Stored 64x48 with orientation 6 is displayed 48x64 (portrait): the short edge is the width.
const inputPath = path.join(tempDir, 'preview-in.jpg')
const outputPath = path.join(tempDir, 'preview-out.webp')
fs.writeFileSync(inputPath, await makeJpeg(6))
await transcodeService.transcodeImage(inputPath, outputPath, 24, 80, { isPreview: true })
const meta = await sharp(fs.readFileSync(outputPath)).metadata()
expect(meta.width).toBeLessThan(meta.height ?? 0)
})
describe('RAW inputs', () => {
const rawPreview = (previewPath: string, extra: Record<string, unknown>) => ({
previewPath,
cleanup: () => {},
width: STORED_WIDTH,
height: STORED_HEIGHT,
...extra,
})
it('never auto-orients dcraw_emu output that still carries an Orientation tag', async () => {
// The dcraw fallback reports orientation undefined (already upright) but flags it as applied.
const tiffLike = path.join(tempDir, 'dcraw.jpg')
const outputPath = path.join(tempDir, 'dcraw-out.webp')
fs.writeFileSync(tiffLike, await makeJpeg(6))
extractRawMock.mockResolvedValueOnce(
rawPreview(tiffLike, { orientation: undefined, orientationApplied: true }),
)
await transcodeService.transcodeImage(path.join(tempDir, 'photo.cr2'), outputPath, 1000, 100)
const meta = await sharp(fs.readFileSync(outputPath)).metadata()
// Stored landscape stays landscape and the stored corners stay put: no second rotation.
expect(meta.width).toBe(STORED_WIDTH)
expect(meta.height).toBe(STORED_HEIGHT)
expect(await cornerColor(outputPath, 'TL')).toBe('TL')
expect(await cornerColor(outputPath, 'BR')).toBe('BR')
})
it('still auto-orients an embedded preview that has no container orientation', async () => {
const embedded = path.join(tempDir, 'embedded.jpg')
const outputPath = path.join(tempDir, 'embedded-out.webp')
fs.writeFileSync(embedded, await makeJpeg(6))
extractRawMock.mockResolvedValueOnce(rawPreview(embedded, { orientation: undefined }))
await transcodeService.transcodeImage(path.join(tempDir, 'photo.cr2'), outputPath, 1000, 100)
const meta = await sharp(fs.readFileSync(outputPath)).metadata()
expect(meta.width).toBe(STORED_HEIGHT)
expect(meta.height).toBe(STORED_WIDTH)
})
})
it('overlayAnnotationsOnBuffer keeps the displayed orientation', async () => {
const annotations = [
{
type: 'box',
color: '#ffffff',
points: [
[0.1, 0.1],
[0.4, 0.4],
],
},
] as unknown as PrismaJson.AnnotationList
const annotated = await transcodeService.overlayAnnotationsOnBuffer(
await makeJpeg(6),
annotations,
)
const meta = await sharp(annotated).metadata()
expect(meta.width).toBe(STORED_HEIGHT)
expect(meta.height).toBe(STORED_WIDTH)
})
})
File diff suppressed because it is too large Load Diff
File diff suppressed because it is too large Load Diff

Some files were not shown because too many files have changed in this diff Show More