Per ajnart's review (and CodeRabbit withdrawing its own MCP comment on the
predecessor PR), the Open WebUI chat isn't a meaningful MCP surface:
- remove the .meta({ mcp }) metadata from every openWebUi procedure
Remaining CodeRabbit comments from the predecessor PR that weren't carried over:
- processWeb: constrain the URL to http(s) and reject loopback/metadata hosts
(private LAN ranges stay allowed for self-hosters); it is also creator-gated
- uploadFile/transcribe: bound the base64 payload and validate filename/type
- docs: align the example URL with the integration default port (8080)
- new chat: clear the composer draft so it doesn't leak into the fresh chat
- web ingestion: keep the typed URL when ingestion fails
- persistence: fall back to the default title for image-only opening turns, and
don't persist a partial reply from a failed turn as finished history
- multimodal round-trip: make the native chat/history contracts accept the same
text-or-parts content union as completions, and flatten (not drop) non-text
nodes when rebuilding transcripts so attachment-backed chats aren't lost.
Image data URLs are still kept out of persisted history by design.
API / router (packages/api):
- open-webui: expose serializable procedures as MCP tools via .meta({ mcp })
and gate knowledge/files/notes reads behind protectedProcedure + interact so
anonymous board viewers can't enumerate private source content
- integration create: wrap integration + creator-permission + secret writes in
a single cross-driver transaction so a partial failure can't leave an
unusable integration
- integration test: scope the permission lookup to the created integration
Integration (packages/integrations):
- open-webui: log only safe metadata (not raw completion chunks) on parse
failure, flush the final buffered SSE frame before treating the stream as
truncated, and apply a default request timeout (streaming opts out)
Widget (packages/widgets):
- collapse knowledge rows by default to avoid an N+1 getKnowledgeFiles burst
- strip markdown images from assistant output to block untrusted fetches
- reconcile the selected model when the model list changes
- clear draft attachments when switching to a saved chat
- dedupe async note/chat attaches after the fetch resolves
- discard stale persist responses via a session token so a cleared chat can't
be rebound to an old Open WebUI chat id
- release the mic stream on mid-start failures and mark the recording cancelled
on unmount so partial audio isn't transcribed
Resolve conflicts by keeping both openWebUi and dev-side registry entries
(integration/widget kinds, creators, docs sitemap). Take dev's next.config
turbopack settings. Reset SQLite migration base to dev's 0040_add_widget_secret;
the integration.creator_id migrations are regenerated for all dialects in a
follow-up commit.
- Wrap setSecret delete+insert in a cross-DB transaction
- Add MCP metadata to all widget secret procedures and register in mcp.ts
- Show configured tokens even when no repos use that provider
- Only clear token input after mutation succeeds, with loading state
Wrap each repository fetch in try/catch so one failure (Redis,
network, etc.) doesn't crash the entire query batch. Failed repos
now return a proper error response instead of causing all repos
to show "No response from provider".
The releases widget was crashing because throwIfActionForbiddenAsync
or the widgetSecrets query would throw before DB migrations are run.
Wrap in try/catch so the widget works without tokens in that case.
- Fix npm scoped package support (filter time metadata keys, make
description optional, exclude falsy releases from version selection)
- Fix timetable date formatting (use Intl.DateTimeFormat for stable
MM/DD/YYYY output) and named zod import
- Guard migration 0000 against unknown provider keys
- Fix migration 0003 to delete integration links by integrationId
- Reset immich album carousel index when photo list changes
- Add bounded limit to timetable router input
- Remove stale misspelled noProviderSeleceted translation key
- Fix provider name visibility in release widget settings
- Update docs: add ArchiveTeam Warrior note, fix typos, correct
showPhotoInfo default
The widget-only integrations refactor incorrectly removed the
ArchiveTeam Warrior integration entirely. It only needed to be
excluded from auto-seeding (by removing defaultUrl), not deleted.
This restores the integration class, reverts the widget/router/handler
to use integration middleware, and removes archiveTeamWarrior from the
widget-only-integrations migration.
Break the monolithic chat.tsx into focused hooks (use-attachments,
use-chat-stream, use-voice-recording, use-camera-capture) and
presentational components (chat-components), with shared types and
utils extracted to chat-types/chat-utils.
Review fixes applied on top:
- require auth for the streaming completion (protectedProcedure) so
anonymous board viewers cannot spend the owner's LLM credits
- rename image-only state to images and the picker query options object
to pickerQueryOptions for clarity
- extract AttachPopover, clearPending(), composerEmpty and
releaseAudioResources() to remove duplication
- Voice messages: record from the microphone and transcribe to the input
via Open WebUI's /api/v1/audio/transcriptions (mic button in composer).
- Camera capture: attach the stream via a ref callback so the preview is
no longer a black frame (the ref could be read before the video mounted).
- Notes: fetch the full note body on attach. The list endpoint truncates
content to 1000 chars, so attached notes were cut off; getNote returns
the complete markdown.
Replace the single attach popover with an Open WebUI-style multi-level
menu covering every attachment source:
- Upload Files: images stay inline (vision); other files are uploaded and
grounded via retrieval (`file-{id}` collection).
- Capture: take a photo from the webcam as an inline image.
- Attach Webpage: ingest a URL (existing).
- Attach Files: pick files already stored in Open WebUI.
- Attach Notes: inject the contents of a note as context.
- Attach Knowledge: nested picker with checkboxes, expandable to a base's
files (replaces the flat checkbox list).
- Reference Chats: bring a previous conversation in as context.
Knowledge bases, web pages and files are grounded by retrieval; notes and
referenced chats are injected verbatim as context blocks. Adds the
supporting integration methods (list files/notes, knowledge files, upload
file), request handlers and tRPC procedures.
Extend the AI Chat widget with three ways to ground a conversation:
- Images: attach images to a message, sent inline as OpenAI multimodal
content parts for vision-capable models. A hint warns when the selected
model likely has no vision support.
- Knowledge bases: pick RAG collections exposed by /api/v1/knowledge/.
- Web pages: ingest a URL via /api/v1/retrieval/process/web.
Open WebUI does not run RAG from the chat-completions `files` parameter,
so attached collections (web pages and knowledge bases) are grounded by
retrieving the most relevant chunks at send time
(/api/v1/retrieval/query/collection) and injecting them as system context.
Also add a Retry action to re-run the last turn after a model failure, and
handle the paginated `{ items, total }` shape of the knowledge endpoint on
newer Open WebUI versions.
Adds a new "openWebUi" integration kind (under a new "ai" category) and an
AI Chat widget for chatting with models served by Open WebUI.
- Integration: Bearer API key auth, model list, native chat history CRUD, and
token-by-token streaming chat completions (SSE).
- Widget: in-input model picker, streaming responses, and an in-widget history
panel. Reconstructs conversations from Open WebUI's message tree with a flat
fallback.
- Streaming is delivered over the WebSocket subscription link.
- Adds integration and widget documentation.
* fix(beszel): harden cache layer and prevent blank widgets
- Add debug logging across Beszel integration, request handler, and tRPC
router (auth, fetch, SSE, per-procedure entry/exit) gated behind
LOG_LEVEL=debug
- cached-request-handler: retry with backoff, isValid predicate to skip
caching invalid data, in-flight dedup (single-flight per channel) to
avoid thundering herd after cache invalidation
- redis channel: set-then-publish so storage is always consistent even
when pubsub publish fails
- Beszel handlers: opt into fallbackToStaleOnError and retry so brief
upstream hiccups don't throw away good cached data
- beszel-system-stats: add refetchInterval matching each time period's
cache TTL so historical charts refresh in the background
- tRPC router: switch Promise.all -> Promise.allSettled for getSystems
and getAlerts; getSystemStats returns an error object instead of
throwing. Each result carries an optional error field so one failed
integration no longer nukes the whole widget
- Widgets: add explicit isPending loader, empty state for grid and
stats, and a shared per-integration error indicator (tooltip with
failed integration names). Alerts and table now only show their
empty state after loading completes
- Translation: add empty.noSystems key to beszelSystemGrid and
beszelSystemStats
Addresses #6030
* style: auto-fix lint and format
---------
Co-authored-by: ajnart <ajnart@users.noreply.github.com>
* feat(analytics): migrate from Umami to PostHog server-side analytics
Replace @umami/node with posthog-node for all server-side telemetry.
Add singleton PostHog client, feature tracking POC, DAU heartbeat,
and tRPC analytics router. Cron job is now locked (no manual trigger,
no custom interval). Update consent copy to explain what we collect
and why.
* style: auto-fix lint and format
* refactor(analytics): simplify + add client-side posthog-js init
- Inline constants.ts and track.ts into client.ts (single source)
- Remove dead exports (shutdownPostHogAsync, getPostHogClient)
- Flatten all DB queries into one Promise.all (no waterfall)
- Derive integration/widget totals from group-by (drop redundant counts)
- Add posthog-js with dynamic import, gated on enableGeneral
- Move Analytics component to body, pass consent from server settings
* style: auto-fix lint and format
* feat(cron): add hidden flag to cron jobs, hide analytics from task panel
The analytics job runs silently in the background — users don't need
to see or interact with it. Hidden jobs are filtered out of getAllAsync
so they never appear in the management UI.
* refactor(analytics): simplify to single toggle, always collect all data
Remove granular enableWidgetData/enableIntegrationData/enableUserData
toggles. Analytics is now a single on/off switch. When enabled, all
anonymous usage data is collected in one flat Promise.all call.
Side effect: existing users with enableGeneral=true who had granular
flags off will now send all data. This is by design — the granular
controls added complexity without meaningful privacy benefit since
the data is already anonymous counts.
* style: auto-fix lint and format
- Throw TRPCError instead of returning null when no integration
- Sort beszelRouter import alphabetically in mcp.ts
- Replace ternary with lookup table for background color
- Mention integration_all in MCP server instructions
- Add .default("1h") to timePeriod so MCP clients can omit it
- Include "1m" in the description's valid timePeriod values
- Clarify that getSystemStats uses only the first integrationId
- Use MantineSize type instead of plain string for itemRadius
Register beszelRouter in the MCP server with .meta({ mcp: enabled })
on getSystems, getAlerts, and getSystemStats procedures. AI assistants
can now query Beszel-monitored systems, view alerts, and fetch
historical metrics through the MCP protocol.
Ref #5968
Thomas Camlong
Josh PrestonandThomas
Anil K TiwariandThomas
Felix-AyushandThomas
Nerdy TechnicianandThomas
Lok3sh-Kumar
ajnart
Louis Bailleau
GlareandManuel