mirror of
https://github.com/danny-avila/LibreChat.git
synced 2026-05-13 07:46:47 +00:00
2024 commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
Danny Avila
|
68d80f3324
|
✨ v0.8.6-rc1 (#13094) | ||
|
Danny Avila
|
6b5596ec36
|
🍪 refactor: Refresh CloudFront Media Cookies (#13091)
* fix: refresh CloudFront media cookies * fix: satisfy changed-file lint * fix: centralize CloudFront image retry * fix: honor base path for CloudFront refresh * fix: bypass auth refresh for CloudFront cookie retry * fix: pass app auth header to CloudFront retry * test: cover CloudFront refresh with OpenID reuse * fix: avoid duplicate CloudFront refresh retries * fix: clear CloudFront scope cookie with matching flags |
||
github-actions[bot]
|
929082387f
|
🌍 i18n: Update translation.json with latest translations (#13080)
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> |
||
|
Danny Avila
|
3e7262cfe0
|
📦 chore: Bump @librechat/agents to v3.1.85 and mermaid to v11.15.0 (#13079)
* 📦 chore: Update @librechat/agents to version 3.1.85 in package-lock.json and package.json files * 📦 chore: Update mermaid to version 11.15.0 in package.json and package-lock.json |
||
|
Danny Avila
|
36e95353ed
|
🏷️ refactor: Rename Code Interpreter Labels To Run Code (#13071)
* fix: Rename Code Interpreter UI to Run Code * fix: Remove unused Run Code i18n keys * fix: Restore tool call labels * fix: Keep assistant Code Interpreter copy * fix: Update agent code environment copy * fix: Update code environment upload copy * fix: Use fresh run code locale keys * fix: Update code environment test copy * fix: Sort upload translation test keys |
||
|
Danny Avila
|
c385f2ba88
|
📦 feat: Configure Skill Import Size Limit (#13073)
* fix: configure skill import size limit * fix: validate skill import size in ui * fix: align skill import size boundary * fix: show exact skill import limit |
||
|
Danny Avila
|
b32b328b87
|
🛡️ fix: Harden Artifact Routing Lookups (#13069) | ||
|
github-actions[bot]
|
508168fa57
|
🌍 i18n: Update translation.json with latest translations (#13058)
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> |
||
|
Danny Avila
|
030dc98a1d
|
☁️ fix: Enable Azure Agent Provider Uploads (#13045) | ||
|
Danny Avila
|
1e9d0cbd0d
|
🖥️ style: Render Bash PTC Calls With Bash UI (#13046)
* fix: Render bash PTC calls with bash UI * fix: Group bash execution tools consistently |
||
|
Danny Avila
|
715a4a5fc1
|
🧰 refactor: Use Bash PTC for Agent Tools (#13042)
* fix: Use Bash PTC for programmatic agent tools * fix: Preserve legacy PTC event calls |
||
|
Danny Avila
|
c7a4e6d418
|
📦 chore: Bump @babel/preset-env to v7.29.5 (#13034)
|
||
Marco Beretta
|
26a6312917
|
🖼️ refactor: Tool Image Outputs outside of Tool Group Auto-Collapses (#12949)
* refactor(attachments): add variant prop to AttachmentGroup
* feat(tool-call): add hideImageAttachments prop to ToolCall
* fix(tool-call): keep MCP image outputs visible when tool group auto-collapses
* test(tool-call): verify MCP images hoist out of collapsed tool group
* fix(tool-call): hoist all grouped attachments and prevent ExecuteCode double-render
- rename hideImageAttachments -> hideAttachments and hide every attachment
in the inner tool when a group auto-collapses, then hoist them via
ToolCallGroup with default variant 'all' so non-image attachments survive
the collapse alongside images
- thread hideAttachments to ExecuteCode so it skips its inline AttachmentGroup
when grouped, preventing double-render when the group is expanded
- memoize sequentialParts and groupedParts in ContentParts (with
groupAttachments rolled into each tool-group entry) so we don't re-flatMap
on every render
* test(tool-call): cover hideAttachments contract and grouping integration
- ToolCall: assert AttachmentGroup is skipped when hideAttachments=true and
rendered when explicitly false, locking the prop's contract
- ToolCallGroup: update variant assertion to 'all' (now hoists images and
files together) and add a non-image-only hoist case
- ContentParts.integration: new test exercising the full
ContentParts -> Part -> ToolCall -> AttachmentGroup chain with realistic
MCP-shaped data (groups 2+ contiguous tool calls and hoists, single calls
render inline, mixed image+file hoists, empty attachments are a no-op)
* fix(tool-call): extend hideAttachments to bash/read_file/skill/subagent
When the post-rebase dev branch added BashCall, ReadFileCall, SkillCall,
and SubagentCall as dedicated tool renderers, each rendered its own
inline AttachmentGroup. Once the parent tool group hoists every
attachment, those inline groups would double-render, so they now honor
the same hideAttachments contract as ToolCall and ExecuteCode.
Also seed the new ToolCallGroup mocks (Users icon, getToolDisplayLabel)
so the existing hoist test suite keeps passing on dev.
* fix(image-gen): suppress inline image when attachments are hoisted
OpenAIImageGen renders the generated image directly via <Image>. When
its tool_call lands inside a grouped tool call, the parent now hoists
those attachments into ToolCallGroup's AttachmentGroup, and the inline
<Image> would render the same file a second time. Thread hideAttachments
through Part -> ImageGen (agent-style branch) so the agent-style image
slot stays out of the way once the parent has hoisted.
* refactor(tool-call): drop dead variant prop and flatten render-part hooks
- AttachmentGroup's variant prop ('images' / 'non-images') had no callers
after the final hoisting design landed, so remove the prop and the
filtering branches; everything passes the default 'all' behavior.
- Replace the makeRenderPart factory + dual useMemo with two plain
useCallbacks (renderPart, renderGroupedPart) sharing the same dep set.
- Tighten test mocks: drop 'any' in the new integration test, hoist the
MCP delimiter constant above its consumer, and remove the now-stale
data-variant attribute assertion.
* refactor(tool-call): extract getToolCallId helper and tidy imports
- Pull the (part?.[TOOL_CALL] as Agents.ToolCall)?.id chain into a single
getToolCallId helper in ContentParts so the three call sites stop
repeating the cast verbatim.
- Re-sort ToolCallGroup local imports longest-to-shortest per the project
convention.
- Add a Users mock to the integration test's lucide-react stub so future
subagent-group tests don't trip over an undefined glyph.
* refactor(tool-call): unnest ternaries in subagent and group labels
|
||
|
Danny Avila
|
a43bc45b73
|
🧭 fix: Preserve File Search Upload Target (#13019) | ||
|
Danny Avila
|
8f92ec012c
|
🧭 fix: Navigate Signed CDN Downloads (#12998)
* fix(files): navigate signed CDN downloads * fix(files): avoid popup target for signed downloads * test(files): restore download URL mock |
||
|
Danny Avila
|
65b63b889e
|
🪟 refactor: Improve Subagent Dialog Prompt Rendering (#12982)
* fix: Improve subagent dialog prompt rendering * fix: Preserve cancelled subagent traces * chore: Reuse generic prompt toggle labels * fix: Scope new-chat subagent cleanup exemption * fix: Use valid subagent prompt min-height * fix: Flatten subagent dialog conditionals * fix: Place subagent prompt in dialog scroll |
||
|
Danny Avila
|
ddf5879ccd
|
⏱️ fix: Align Auto-Refill Next Date (#12980)
* fix: Align auto-refill next date * style: Fix auto-refill lint formatting * refactor: Share auto-refill eligibility date * refactor: Consolidate refill interval units * fix: Guard malformed refill interval units * fix: Preserve refill unit fallback label |
||
Joseph Licata
|
5efbcb8b93
|
🌐 fix: Percent-encode X-File-Metadata header for Unicode filenames (#12983)
* 🌐 fix: Percent-encode X-File-Metadata header for Unicode filenames After #12977 preserved Unicode in filenames, the download route crashes with ERR_INVALID_CHAR because JSON.stringify(file) now contains non-ASCII characters that Node.js rejects in HTTP headers per RFC 7230. Wrap the header value in encodeURIComponent on the server and decodeURIComponent on the client before JSON.parse. * fix: Update file route tests after dev merge --------- Co-authored-by: Danny Avila <danny@librechat.ai> |
||
|
Danny Avila
|
9c81792d25
|
🔐 feat: Add Signed CloudFront File Downloads (#12970)
* feat: add signed CloudFront downloads * fix: preserve local IdP avatar paths * fix: address signed download review findings * fix: harden CloudFront cookie scope validation * fix: preserve URL save API compatibility * fix: store CDN SSO avatars under shared prefix * fix: Harden CloudFront tenant file access * fix: Preserve CloudFront download compatibility * fix: Address CloudFront review follow-ups * fix: Preserve file URL fallback user paths * fix: Address download review hardening * fix: Use file owner for S3 RAG cleanup * fix: Address final download review nits * fix: Clear stale avatar CloudFront cookies * fix: Align download filename helpers with dev * fix: Address final CloudFront review follow-ups * fix: Stream S3 URL uploads * fix: Set S3 stream upload length * fix: Preserve download metadata filepath * fix: Avoid remote content length for stream uploads * fix: Use bounded multipart URL uploads * fix: Harden S3 filename boundaries |
||
|
Marco Beretta
|
4bd5630651
|
🧭 feat: Add Message Navigation Strip & Redesign Scroll-to-Bottom (#12657)
* feat(ui): add message navigation strip and redesign scroll-to-bottom button
Add a floating vertical navigation strip on the right edge of the chat
area that lets users jump between messages quickly. Each message gets an
indicator line (wider for assistant, narrower for user) with HoverCard
previews showing truncated message text. IntersectionObserver tracks
which messages are currently visible and highlights their indicators.
Redesign the scroll-to-bottom button: solid backgrounds instead of
semi-transparent, clean enter/exit animations without twist/rotate,
no hover float animation, positioned at the right edge of the chat
form instead of center.
* fix(ui): prevent message nav layout shift on scroll
Use a fixed-height container for each indicator so the nav strip
maintains consistent dimensions when indicators transition between
active and inactive states.
* fix(ui): debounce message nav refresh and persist visibility state
Debounce entry refresh (200ms) to avoid thrashing from rapid DOM
mutations during code block rendering. Persist the visible message
set across IntersectionObserver reconnections to prevent momentary
empty state that disabled navigation buttons.
* fix(ui): prevent nav buttons from disabling during fast scroll
- Fall back to last known active index when IntersectionObserver
reports no visible messages during rapid scrolling
- Lower intersection threshold from 10% to 1% for long messages
- Fix preview text to skip the message header (Prompt N: username)
* fix(ui): scroll to message start when using nav arrow buttons
Arrow buttons now use block: 'start' to always scroll to the top of
the target message. Indicator dots keep block: 'nearest' for minimal
repositioning on direct clicks.
* fix(ui): account for header offset when scrolling to messages
Use manual scrollTo with a 56px offset to prevent the fixed header
from covering the top of the target message when using arrow buttons.
* fix(ui): improve message nav scrolling and visual subtlety
- Up button scrolls to current message top first before jumping to
previous, preventing skipped messages on long content
- Down button consistently scrolls to the start of the next message
- Nav strip is faded (opacity 30%) by default, fully visible on hover
- Background, buttons, and indicators all appear on hover of the
nav area using group hover coordination
* fix(ui): use native scroll-margin-top for reliable message navigation
Replace manual scrollTo calculations with scrollIntoView + CSS
scroll-margin-top on .message-render elements. The browser handles
scroll offset natively, eliminating positioning errors during smooth
scroll animations.
* fix(ui): use firstActiveIndex for both nav directions
Use firstActiveIndex (topmost visible message) for both up and down
navigation. Down now advances one message at a time from what the user
is currently reading instead of jumping past all visible messages.
Remove unused lastActiveIndex.
* fix(ui): address PR review feedback
- Scope getMessageEntries query to scroll container instead of document
- Include preview text in entries equality check to catch content
updates during streaming/edits
- Move scroll button transition to base state so release animates
smoothly instead of snapping back
* fix(ui): make message nav scroll precise and chevrons reliable
- Bump .message-render scroll-margin-top from 1rem to 4rem so messages
land below the 52px absolute gradient header instead of behind it.
- Drive chevron jumps from live scrollTop + offsetTop comparison rather
than the IntersectionObserver-derived firstActiveIndex, which lagged
behind rapid clicks and treated any 1px-visible message as "current".
- Track canGoUp / canGoDown from the same scroll-position comparison so
the disabled state matches what the buttons will actually do.
- Auto-center the indicator column on the visible message range and
smooth-scroll it via rAF so 500+ indicators stay at 60fps.
- Pull entry data from useGetMessagesByConvoId (with a DOM fallback) so
previews are state-backed instead of scraped from rendered markup.
- Memoize MessageIndicator and filter MutationObserver to .message-render
add/remove only.
- Add 5 i18n keys (com_ui_message_nav*) for nav and indicator labels.
* perf(ui): skip off-screen message layout and fix resulting scroll drift
Large conversations used to freeze the main thread during sidebar
toggles because every animated frame had to relayout every message.
With ~3000 message elements on this branch: avg frame 650ms,
max 1701ms (~1.5fps) during the 300ms transition. Adding
`content-visibility: auto` with `contain-intrinsic-size: auto 200px`
on .message-render lets the browser skip layout/paint for messages
outside the viewport, dropping avg frame to 33ms and max to 74ms
(~30fps, feels responsive).
content-visibility comes with a trade-off though: off-screen messages
use the 200px intrinsic-size estimate until they're measured. That
broke indicator-click scrolling on long conversations, landing 1-2
messages off the target because scrollIntoView computed its target
scrollTop once with stale estimates, and intermediate messages
shrunk/grew as they rendered during the smooth scroll.
Replaced scrollIntoView with a manual rAF scroll that re-reads the
target's getBoundingClientRect every frame and eases toward the
*current* target. Verified drift=0 across fake-0, fake-50, fake-250,
fake-450 (messages near the bottom naturally land higher than
scroll-margin when the container is already at max scroll — expected).
Also two small MessageNav.tsx hot-path cleanups:
- Use col.children[i] instead of col.querySelector by data-msg-id for
the indicator-column centering lookup (entries map 1:1 to column
children since HoverCardTrigger asChild forwards to the button).
- Compare visibility set contents before setActiveIds, so an
IntersectionObserver flush with unchanged membership doesn't force
a re-render and 500x memo comparisons.
* revert(ui): drop content-visibility on .message-render
Didn't deliver the expected sidebar-toggle perf win in real-world
usage, and its intrinsic-size estimation introduced the exact kind of
scroll drift we then had to work around. The rAF scroll in MessageNav
is orthogonal to this and stays — it works fine with or without
content-visibility.
* fix(ui): address PR review — a11y, tests, and MessageNav correctness
- ScrollToBottom aria-label now runs through useLocalize instead of being
hardcoded English. Added com_ui_scroll_to_bottom translation key.
- MessageNav nav expands on keyboard focus-within, not just pointer hover.
- Indicator buttons expose aria-current="true" for the active message and
get a visible focus-visible ring. Chevron buttons get the same ring so
keyboard users can see focus.
- Cancel in-flight rAF scrolls when a new navigation starts, so clicking
a second indicator mid-animation doesn't race the first loop on
container.scrollTop.
- Invalidate the cached offsetsTop/offsetsBottom arrays via a
ResizeObserver on the scroll content. Previously heights that changed
after mount (code blocks rendering, images loading) left canGoUp /
canGoDown and the indicator-column centering reading stale positions.
- Observe IntersectionObserver entries incrementally. The observer is
now created once per scroll container and entries add/remove on
change instead of the whole observer being torn down and rebuilt for
every new message.
- memo() the default export so parent re-renders don't cascade through
MessageNav when entries/activeIds haven't changed.
- Add 18-test suite covering rendering threshold, user/assistant
indicator styling, preview sourcing (React Query vs DOM fallback vs
truncation), accessibility (aria-label, aria-current, chevron
disabled state), click-driven rAF scroll + cancellation, and observer
lifecycle (observe on mount, incremental sync, unobserve on removal,
disconnect on unmount).
* fix(ui): catch in-place message id mutations and react to layout shifts
Follow-ups from deep review:
- MutationObserver on .message-render now also watches the id attribute.
During the SSE lifecycle a single DOM node's id cycles through three
values (client UUID -> createdHandler id -> server id, see the comment
in MultiMessage.tsx), which meant the previous childList-only observer
never refreshed entries after a streaming response. Nav clicks on the
most recent message were silently failing because getElementById
returned null for the stale id.
- ResizeObserver now calls scheduleTick() instead of only flipping a
flag. The flag was only consumed inside the scroll handler's tick, so
heights that changed while the user wasn't scrolling (assistant message
streaming in, code blocks highlighting) left offsetsTop/offsetsBottom
stale and canGoUp / canGoDown wrong. Both handlers now route through
scheduleTick so a resize and a scroll share the same rAF slot.
- Unify scroll and resize callbacks on scheduleTick. Removes a duplicate
rAF path and makes the effect cleaner.
- Single-pass build of newIds during incremental IO sync (previously
entries.map().new Set() did two passes for no reason).
- CSSTransition timeouts drop from 550/700 to 300/250 to match the new
scroll-to-bottom animations. Old values left the button in the DOM
for up to 450ms after the exit animation finished.
- ScrollToBottom.tsx imports reordered to longest-first per project
convention.
- style.css: collapse split `border: 1px solid` + `border-color` into
one shorthand; dark variant still overrides border-color cleanly.
- Tests: add SSE-lifecycle test that mutates a .message-render id in
place and asserts the nav now shows an indicator for the new id and
none for the old one. HoverCard mock no longer spreads unknown props
to the DOM div (drops a React warning).
* fix(ui): address deep-review follow-ups on MessageNav
- Move activeScrollToken from module scope to a per-instance useRef
(scrollTokenRef). When LibreChat eventually mounts more than one
MessageNav side-by-side (multi-panel / added-convo view) a click in
one panel will no longer cancel an in-flight smooth scroll in another.
scrollToMessageStart is now an instance useCallback and the button
click path goes through an onSelect prop on MessageIndicator, keeping
the memoized indicator stable.
- messagesById goes through a ref (messagesByIdRef) so refreshEntries is
no longer recreated on every streaming token. Previously messagesById
landed in both the useMemo and the refreshEntries dep array, so each
streaming response rebuilt the MutationObserver effect dozens of times
per second. A separate small effect still calls refreshEntries when
messagesById changes, so previews stay fresh.
- Extract USER_TURN_SELECTOR constant and tighten the text-preview type
narrowing so we no longer need the `as { value?: string }` cast (TS
narrows string | TextData correctly through the `typeof object` +
property access guard).
- Cache the computed scroll margin (4rem = 64px) in scrollMarginRef so
the nav callbacks don't call getComputedStyle on every click.
- Tests: add a two-instance isolation test that verifies scroll tokens
don't cross between mounted MessageNavs. Drop the unused `import React
from 'react'` pattern in favor of local type aliases.
- client/package.json: bump @babel/preset-typescript to ^7.28.5. The old
^7.22.15 constraint was resolving to 7.23.3 via hoisting, which can't
parse modern `import type` syntax on a clean install and was breaking
the test suite.
* fix(ui): address re-review — clean lockfile + ScrollToBottom ref target
- package-lock.json: the preset-typescript bump last commit pulled in
transitive Babel packages resolved through a local internal registry
(npm.internal.berry13.com). Rewrote those 31 entries back to the
public npmjs.org registry so CI and contributors can install cleanly.
Integrity hashes unchanged — content-addressed.
- ScrollToBottom now forwards its ref to the wrapping <div> instead of
the inner <button>. CSSTransition's nodeRef + unmountOnExit can now
add transition classes to the actual root element, so the layout
wrapper is what mounts/unmounts, not just the button. Updated
scrollToBottomRef type in MessagesView to HTMLDivElement.
- jumpToPrevious / jumpToNext skip the document.getElementById fallback
lookup when scrollMarginRef is already populated, which is the normal
case after the first scroll-tick effect run.
* fix(ui): preserve IntersectionObserver across in-place id mutations
The IO sync effect was observing new ids before unobserving old ones.
During the SSE lifecycle of a fresh chat, a single .message-render node
cycles through three ids (client UUID -> handler id -> server id). When
the id mutated on the same element, the effect would call observe(el)
then unobserve(el) on that element in the same pass — leaving it
permanently unobserved. The active-message highlight never updated for
the new id until a hard refresh rebuilt everything from scratch.
Switched to element-identity tracking. Build an element -> newId map
from entries, then for each currently observed [oldId, el]:
- if the element no longer appears in entries, unobserve and drop it
- if the element appears under a new id, migrate observed and
visibleSet keys in place — the IntersectionObserver keeps watching
the same DOM node uninterrupted
Genuinely new elements get observed afterward as before. Rename doesn't
fire an IO callback, so flush activeIds manually when at least one
migration happened. Existing convos already had this working because
their ids never mutate after load — only fresh chats hit the SSE id
cycle, which matches the reproduction.
* fix(ui): keep message nav current and pinned at bottom
|
||
|
github-actions[bot]
|
09c8c05c06
|
🌍 i18n: Update translation.json with latest translations (#12964)
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> |
||
|
Danny Avila
|
6c6c72def7
|
🚀 feat: Decouple File Attachment Persistence from Preview Rendering (#12957)
* 🗂️ feat: add `status` lifecycle to file records for two-phase previews
Schema and model foundation for decoupling the agent's final response
from CPU-heavy office-format HTML extraction.
- `MongoFile.status: 'pending' | 'ready' | 'failed'` (indexed) and
`previewError?: string` mirror the lifecycle: phase-1 emits the file
record at `pending` so the response is unblocked; phase-2 transitions
to `ready` (with text/textFormat) or `failed` (with previewError) in
the background. Absent for legacy records — clients treat that as
`ready` for back-compat.
- Mirror types added to `TFile` in data-provider so frontend cache
consumers see the new fields.
- New `sweepOrphanedPreviews(maxAgeMs)` method on the file model
recovers stale `pending` records left behind by a process restart
mid-extraction; transitions them to `failed` with
`previewError: 'orphaned'`. Cheap because `status` is indexed.
* ⚡ feat: two-phase code-execution preview flow (unblocks final response)
The agent's final response no longer waits on CPU-heavy office HTML
extraction. Phase-1 (download + storage save + DB record at
`status: 'pending'`) is awaited as before; phase-2 (extract +
`updateFile`) runs in the background with a hard 60s ceiling.
Three flows, all funneling through `processCodeOutput` and updated to
the new `{ file, finalize? }` return shape:
- `callbacks.js` (chat-completions + Open Responses streaming): emit
the phase-1 attachment immediately (carries `status: 'pending'` for
office buckets so the UI shows "preparing preview…"), then
fire-and-forget `finalize()`. If the SSE stream is still open when
phase-2 lands, push an `attachment` update event with the same
`file_id` so the client merges over the placeholder in place.
- `tools.js` direct endpoint: same split — return the phase-1
metadata immediately, run extraction in the background. Client
polls for the resolved record.
`finalize()` wraps the existing 12s per-render timeout in a 60s outer
`withTimeout`. The HTML-or-null contract from #12934 is preserved:
office types that fail extraction transition to `status: 'failed'`
with `previewError: 'parser-error' | 'timeout'` rather than falling
back to plain text (would be an XSS vector).
Promises continue running after the HTTP response closes (Node
doesn't kill them). The boot-time orphan sweep covers the only case
that loses progress — actual process restart mid-extraction.
`primeFiles` annotates the agent's `toolContext` line for prior-turn
files: `(preview not yet generated)` for pending, `(preview
unavailable: <reason>)` for failed. The model can volunteer "you can
still download it" instead of pretending the preview is fine.
`hasOfficeHtmlPath` exported from `@librechat/api` so `processCodeOutput`
can decide whether a file expects a preview at all.
* 🔍 feat: `GET /api/files/:file_id/preview` endpoint and boot orphan sweep
- New `GET /api/files/:file_id/preview` route returns
`{ status, text?, textFormat?, previewError? }`. The frontend's
`useFilePreview` React Query hook polls this while phase-2 is in
flight, then auto-stops on terminal status. ACL identical to the
download route (reuses `fileAccess` middleware). Defaults `status`
to `'ready'` for legacy records so back-compat is implicit.
`text` only included when `status === 'ready'` and non-null —
preserves the HTML-or-null security contract from #12934.
- `sweepOrphanedPreviews()` invoked on boot in both `server/index.js`
and `server/experimental.js`. Recovers any `pending` records left
behind by a process restart mid-extraction (the only case the
in-process two-phase flow can't handle on its own). Fire-and-forget
so a transient sweep failure doesn't block startup.
* 🖥️ feat: frontend two-phase preview consumer (polling + UI states)
Wires the React side to the new lifecycle so the user sees what's
happening with their file while phase-2 extraction runs in the
background and after the response stream closes.
- `useAttachmentHandler` upserts by `file_id` (was append-only) so
the phase-2 SSE update event merges over the pending placeholder
in place. Lightweight attachments without a `file_id`
(web_search / file_search citations) keep the legacy append path.
- `useFilePreview(file_id)` React Query hook with
`refetchInterval: (data) => data?.status === 'pending' ? 2500 : false`
so polling auto-stops on the first terminal response without the
caller having to flip `enabled`.
- `useAttachmentPreviewSync(attachment)` bridges polled data into
`messageAttachmentsMap`. Polling enabled iff
`status === 'pending' && isAnySubmitting` — per the design ask:
active polling while the LLM is still generating, then quiet.
Process-restart and post-stream cases are covered by polling on
the next interaction.
- `Attachment.tsx` renders a small `PreviewStatusIndicator` (spinner +
"Preparing preview…" for pending, alert icon + "Preview unavailable"
for failed) inside `FileAttachment`. Download button stays fully
functional in both states. Two new English locale keys.
- Data-provider scaffolding: `TFilePreview` type, `endpoints.filePreview`,
`dataService.getFilePreview`, `QueryKeys.filePreview`.
* 🧪 fix: stub `useAttachmentPreviewSync` in pre-existing Attachment test mocks
The new `useAttachmentPreviewSync` hook is called unconditionally inside
`FileAttachment` (added in the prior commit). Two pre-existing test
files mock `~/hooks` to provide `useLocalize` only — the un-mocked
preview hook reference resolved to undefined and crashed render with
`(0 , _hooks.useAttachmentPreviewSync) is not a function` on the
Ubuntu/Windows CI runners.
Fix is local to the test mocks: add a no-op stub that returns
`{ status: 'ready' }` so the component renders the legacy chip path.
The two-phase preview behavior itself has its own dedicated suites
(`useAttachmentHandler.spec.tsx`, `useAttachmentPreviewSync.spec.tsx`).
* 🐛 fix: route phase-2 attachment update to current-run messageId
Codex P1 review on PR #12957. `processCodeOutput` intentionally
preserves the original DB `messageId` across cross-turn filename reuse
so `getCodeGeneratedFiles` can still trace a file back to the
assistant message that originally produced it. The phase-1 SSE emit
already routes by the current run's messageId — `processCodeOutput`
runtime-overlays it via `Object.assign(file, { messageId, toolCallId })`
and the callback writes `result.file` directly.
Phase-2 was passing the raw `updateFile` return through
`attachmentFromFileMetadata`, which read `messageId` straight off the
DB record. On a turn-N run that re-emitted a filename from turn-1
(e.g. agent writes `output.csv` again), the phase-2 SSE update
routed to `turn-1-msg` instead of `turn-N-msg`. Frontend's
`useAttachmentHandler` upserts under the wrong messageAttachmentsMap
slot — turn-N's pending chip stays stuck at "preparing preview…"
while turn-1's already-resolved attachment gets re-merged.
Fix: thread `runtimeMessageId` through `attachmentFromFileMetadata`
and pass `metadata.run_id` from the phase-2 emit site. Mirrors how
phase-1 sources its messageId. Tests cover the cross-turn reuse case
plus the writableEnded / null-finalize / no-finalize paths to lock
in the broader phase-2 emit contract.
* 🛠️ refactor: address codex audit findings (wire-shape parity, DRY, defensive catch)
Comprehensive audit on PR #12957. Resolves all valid findings:
- **MAJOR #1 — Wire-shape parity**: phase-1 ships the full `fileMetadata`
record over SSE; phase-2 was using a tight `attachmentFromFileMetadata`
projection. Drop the projection and have phase-2 spread `{...updated,
messageId, toolCallId}` so both events match the long-standing
legacy phase-1 shape clients depend on.
- **MAJOR #2 — DRY**: extract `runPhase2Finalize({ finalize, fileId,
onResolved })` into `process.js` (alongside `processCodeOutput` whose
contract it pairs with). Both `callbacks.js` paths and `tools.js`
now flow through it. Single catch path eliminates divergence
surface — the fix landed in 01704d4f0 (cross-turn messageId routing)
was a symptom of this duplication risk.
- **MINOR #3 — JSDoc accuracy**: `finalizePreview`'s buffer is bounded
by `fileSizeLimit`, not the 1MB extractor cap. Updated and added a
note about peak heap from queued buffers.
- **MINOR #4 — Defensive catch**: `runPhase2Finalize`'s catch attempts
a best-effort `updateFile({ status: 'failed', previewError:
'unexpected' })` for the file_id, so a programming bug in
`finalizePreview` doesn't leave the record stuck `'pending'` until
the next boot-time orphan sweep.
- **NIT #6 — Stale PR refs**: 12952 → 12957 in 3 places.
- **NIT #7 — Schema bound**: `previewError` capped at `maxlength: 200`
to prevent a future codepath from accidentally persisting a stack
trace.
Skipped per audit verdict (non-blocking):
- #5 (memory pressure): documented in JSDoc; impl change was reviewer's
"consider", not actionable.
- #8 (double DB query per poll): low cost, indexed by_id, polling is
gated narrow.
- #9 (TAttachment cast): the union type is intentional; the casts are
safe widening, refactoring TAttachment is invasive and out of scope.
Tests: 11 new (7 `runPhase2Finalize` unit tests covering happy path,
null-finalize, throws, double-fail, no-fileId, no-onResolved; +4
wire-shape parity assertions in the existing cross-turn test). 328
backend tests pass; 528 frontend tests pass; lint and typecheck clean.
* 🛡️ refactor: address codex P1+P2 + rename to drop phase-1/2 jargon
Codex round 2 review on PR #12957 caught two race conditions and one
recovery gap, all triggered by cross-turn filename reuse (`claimCodeFile`
intentionally returns the same `file_id` for the same
`(filename, conversationId)` across turns). Plus naming cleanup the
user requested — internal "phase 1 / phase 2" vocabulary leaks across
sprints, replace it everywhere with terms describing what's actually
happening.
P1 — stale render overwrites newer revision (process.js)
Two turns reusing `output.csv` share a `file_id`. If turn-1's
background render resolves AFTER turn-2's persist step, the
unconditional `updateFile` writes turn-1's stale text/status over
turn-2's pending placeholder. Fix: stamp a fresh `previewRevision`
UUID on every emit, thread it through `finalizePreview`, and make
the commit conditional via a new optional `extraFilter` argument
on `updateFile` (`{ previewRevision: <expected> }`). The defensive
`updateFile` in `runPreviewFinalize`'s catch uses the same guard
so a programming error from an older render also can't override a
newer turn.
P1 — stale React Query cache on pending remount (queries.ts)
Same root cause from the frontend side. Cache key
`[QueryKeys.filePreview, file_id]` may hold a prior turn's `'ready'`
payload; with `refetchOnMount: false` and the polling gate on
`pending`, polling never starts for the new placeholder. Fix:
`useAttachmentHandler` invalidates that query whenever an attachment
with a `file_id` arrives. Both initial-emit and update events
trigger invalidation — uniform gate.
P2 — quick-restart orphans skipped by boot sweep (files.js)
Boot `sweepOrphanedPreviews` uses a 5-min cutoff for multi-instance
safety. A crash + restart inside the cutoff leaves `pending` records
that never get touched again. Fix: lazy sweep inside the preview
endpoint — if a polled record is `pending` and `updatedAt` is older
than 5 min, mark it `failed:orphaned` on the spot before responding.
Conditional on the same `updatedAt` we observed so a concurrent
legitimate update wins. Cheap, bounded by user activity.
Naming cleanup
- `runPhase2Finalize` → `runPreviewFinalize`
- `PHASE_TWO_TIMEOUT_MS` → `PREVIEW_FINALIZE_TIMEOUT_MS`
- All `phase-1` / `phase-2` / `two-phase` prose replaced with
"the immediate emit", "the deferred render", "the persist step",
"the deferred preview", etc. Skill-feature `phase 1/2` references
(different feature) left alone.
Tests: 10 new (4 lazy-sweep × preview endpoint, 3 cache-invalidation ×
useAttachmentHandler, 3 extraFilter × updateFile data-schemas).
Backend 332/332, frontend 531/531, data-schemas 37/37, lint clean.
* 🛠️ refactor: address comprehensive review (round 3) — stale-cache MAJOR + 3 minors
Comprehensive review on PR #12957 caught a P1 follow-on bug from the
prior `invalidateQueries` fix, plus 3 maintainability findings.
MAJOR: stale React Query cache not actually fixed by `invalidateQueries`
The previous fix called `invalidateQueries` to flush stale cached
preview data on cross-turn filename reuse. But `useFilePreview` had
`refetchOnMount: false`, which made the new observer read the
stale-marked 'ready' data without refetching. The polling
`refetchInterval` then evaluated against stale 'ready' → returned
`false` → polling never started → user stuck on stale content.
Fix (belt-and-suspenders):
a) `useAttachmentHandler` switched to `removeQueries` — drops the
cache entry entirely so the next mount has nothing to read and
must fetch.
b) `useFilePreview` no longer sets `refetchOnMount: false`, so the
React Query default (`true`) kicks in — second line of defense
if any future codepath observes stale data before the handler
has a chance to evict.
MINOR: `finalizePreview` JSDoc missing `previewRevision` param
Added with explanation of the conditional update guard.
MINOR: asymmetric stream-writable guard between SSE protocols
Chat-completions delegated the gate to `writeAttachmentUpdate`;
Open Responses inlined `!res.writableEnded && res.headersSent`.
Extracted `isStreamWritable(res, streamId)` predicate; both paths
+ `writeAttachmentUpdate` now share the single source of truth.
NIT: `(data as Partial<TFile>).file_id` cast repeated 4 times
Extracted to a `fileId` local at the top of the handler.
Tests: existing 9 invalidate-tests rewritten as remove-tests; +1 new
lock-in test asserts removeQueries is called and invalidateQueries
is NOT (regression guard against round-3 finding). 332 backend pass,
532 frontend pass, lint clean.
Skipped findings (deferred / acceptable):
- MINOR: post-submission pending state has no auto-recovery — the
`isAnySubmitting` polling gate was the user's explicit design;
LLM context surfaces failed/pending so the model can volunteer.
Worth a follow-up if real users hit it.
- NIT: double DB query per preview poll — reviewer marked acceptable;
changing `fileAccess` middleware is out of scope.
* 🛡️ test: address comprehensive review NITs (initial-emit guard + isStreamWritable coverage)
NIT — chat-completions initial emit skips writableEnded check
The Open Responses initial emit was switched to use the new
`isStreamWritable` predicate in the round-3 commit, but the
chat-completions initial emit kept the older narrower check
(`streamId || res.headersSent`). On a client disconnect mid-stream
(`writableEnded === true`) it would still hit `res.write` and
raise `ERR_STREAM_WRITE_AFTER_END` — caught by the outer IIFE
catch but logged as noise. Switch this site to `isStreamWritable`
too so both initial-emit paths share the same gate as the
deferred update emits.
NIT — `isStreamWritable` not directly unit-tested
The predicate was only covered indirectly via the deferred-preview
SSE tests (writableEnded skip, headersSent check). Export from
`callbacks.js` and add 5 parametric tests pinning down each branch
(streamId truthy, res null, !headersSent, writableEnded, happy
path) so a future condition addition can't silently regress.
* 🐛 fix: stuck "Preparing preview…" + inline the chip subtitle
Two related fixes for a stuck-spinner bug a user reported in manual
testing of PR #12957.
**Stuck spinner (the bug)**
The deferred preview render can complete a few seconds AFTER the SSE
stream closes (typical case: PPTX render finishes ~3s after the LLM
emits FINAL). When that happens, the SSE update is silently dropped
(`isStreamWritable` returns false on a closed stream) and polling is
the only recovery path.
The earlier polling gate was `status === 'pending' && isAnySubmitting`,
which mirrored the original design intent ("only query while the LLM
is still generating"). But `isAnySubmitting` flips false the moment
the model emits FINAL — milliseconds before the deferred render
commits. Polling never runs, the chip stays "Preparing preview…"
forever even though the DB has `status: 'ready'` with valid HTML.
Drop the `isAnySubmitting` part of the gate. `useFilePreview`'s
`refetchInterval` is already a function-form that returns `false` on
the first terminal response, so polling auto-stops within one tick of
resolution. The server-side render ceiling (60s) plus the lazy sweep
in the preview endpoint cap the worst case to ~24 polls per pending
attachment. Polling itself never blocks UX — the gate's purpose was
"don't waste cycles", and capping by terminal status is the correct
expression of that.
**Inline the chip subtitle (the visual)**
The previous design rendered "Preparing preview…" as a loose-feeling
spinner+text BELOW the file chip. The chip itself looked done while a
floating annotation said it wasn't.
`FileContainer` gains an optional `subtitle?: ReactNode` prop that
overrides the default file-type label. `Attachment.tsx` passes a
`PreviewStatusSubtitle` (spinner + "Preparing preview…" / alert +
"Preview unavailable") into that slot when the file's preview is
pending or failed. The chip footprint stays identical to its `'ready'`
form — just the second row swaps from "PowerPoint Presentation" to
the status indicator. No floating element, no layout shift.
Tests: regression test pinning down "polling stays enabled after the
LLM finishes" so a future revert can't reintroduce the stuck-spinner
bug. Existing FileContainer tests pass unchanged (subtitle override
is opt-in). 522 frontend tests pass; lint clean.
* 🐛 fix: deferred-preview survives reload + matches artifact card chrome
Fixes the remaining stuck-pending case after the polling gate fix: on
a reloaded conversation, message.attachments come from the DB frozen at
the immediate-persist `status: 'pending'`, but `messageAttachmentsMap`
is empty because no SSE handler ever fired for that messageId. Polling
now INSERTS a new live entry when no record matches the file_id, and
`useAttachments` merges live entries onto DB entries by file_id so the
resolved text/textFormat reach `artifactTypeForAttachment` and the
chip routes through the proper PanelArtifact card.
Also replaces the small file chip used during the pending state with
a PreviewPlaceholderCard that mirrors ToolArtifactCard chrome, so the
transition to the resolved PanelArtifact no longer reshapes the UI.
* ✨ feat: auto-open panel when deferred preview resolves pending→ready
The legacy auto-open path is gated only on `isSubmitting`, so an
office-file preview that resolves *after* the SSE stream closes would
render in place but never auto-open the panel — even though that's
exactly the moment the result becomes meaningful to the user. Adds a
per-file_id one-shot signal that `useAttachmentPreviewSync` flips on
the pending→ready edge; `ToolArtifactCard` consumes it on mount and
auto-opens regardless of submission state. The signal is *only* set on
the actual transition (history loads of pre-resolved files don't
trigger it) and is consumed once (panel close + reopen on the same
card stays user-controlled).
* 🐛 fix: drop placeholder Terminal overlay + scope auto-open to fresh resolutions
Two fixes for issues spotted in manual testing of the deferred-preview
auto-open feature:
1. PreviewPlaceholderCard was passing `file={attachment}` to FilePreview,
which triggered SourceIcon's Terminal overlay (`metadata.fileIdentifier`
is set on every code-execution file). The artifact card itself doesn't
show that overlay; the placeholder shouldn't either, so the
pending→resolved transition is visually seamless.
2. The `previewJustResolved` flag flipped on every pending→ready
transition observed by the polling hook — including stale-pending
DB records that resolve via the first poll on a *history load*.
Conversations whose immediate-persist snapshot left attachments at
`status: 'pending'` would yank the panel open every revisit.
Adds `mountedDuringStreamRef` to the hook (mirroring ToolArtifactCard)
so the flag fires only when the hook itself was mounted during an
active turn — preserving the pre-PR contract that the panel only
auto-opens for results the user is actively waiting on, never for
history.
* 🐛 fix: don't downgrade preview to failed when only the SSE emit throws
Codex P2 finding on PR #12957: the original chain placed `.catch` after
`.then(onResolved)`, so a throw inside `onResolved` (transport-side
errors — SSE write race after stream close, an emitter listener
throwing) would propagate into the finalize catch and persist
`status: 'failed'` / `previewError: 'unexpected'`. That surfaced
"preview unavailable" in the UI for a perfectly valid file, and
degraded next-turn LLM context to reflect a non-existent failure.
Wraps `onResolved` in its own try/catch so emit errors are logged but
do not affect the file's persisted status. Extraction success and
emit success are now independent: if extraction succeeds and
`finalizePreview` writes the terminal status, the polling layer / next
page load surfaces the resolved preview even if this turn's SSE emit
didn't land.
* 🛡️ fix: run boot-time orphan sweep under system tenant context
Codex P2 finding on PR #12957: `File` is tenant-isolated, so under
`TENANT_ISOLATION_STRICT=true` the boot-time `sweepOrphanedPreviews`
threw `[TenantIsolation] Query attempted without tenant context in
strict mode` and the recovery path silently failed every restart.
Stale `status: 'pending'` records would be stuck until a user happened
to poll the preview endpoint and trigger the lazy sweep — which only
covers the file the user is currently looking at, not the bulk
candidate set the boot sweep is designed to recover.
Wraps the sweep in `runAsSystem(...)` in both boot paths
(`api/server/index.js` and `api/server/experimental.js`) and pins the
contract with regression tests in `file.spec.ts` — one test asserts
the bare call throws under strict mode, the other asserts the
`runAsSystem`-wrapped call succeeds.
* 🧹 chore: trim verbose comments from previous commit
* 🧹 chore: address review findings (dead branch, lazy-sweep cutoff, stale JSDoc)
- finalizePreview: drop unreachable !isOfficeBucket branch (caller
already gates on hasOfficeHtmlPath, so this path is always office)
- preview endpoint: drop lazy-sweep cutoff from 5min to 2min — anything
past the 60s render ceiling is definitively orphaned, and per-request
sweep can be tighter than the per-instance boot sweep
- strip stale `isSubmitting` references from JSDoc in 3 spots (the
client-side gate was removed in
|
||
|
Danny Avila
|
25a4556aee
|
⌨️ refactor: Clarify Bash Command Drafting State (#12963) | ||
|
Danny Avila
|
16a65b67fc
|
🗃️ refactor: Keep Code Artifacts Manual-Open (#12961) | ||
|
Danny Avila
|
f20419d0b7
|
📄 feat: Rich File Artifact Previews for DOCX, CSV, XLSX, PPTX (#12934)
* 📄 feat: Rich File Artifact Previews for DOCX, CSV, XLSX, PPTX Render office files emitted by tools as interactive previews in the artifact panel instead of raw extracted text. The backend produces a sanitized HTML document via mammoth (DOCX), SheetJS (CSV/XLSX/XLS/ODS), or yauzl-based slide extraction (PPTX) and ships it through the existing SSE attachment payload; the client routes it through the Sandpack `static` template's `index.html` slot — no new browser deps, no client-side blob fetch, no React renderer components. * 🔐 fix: Restrict data: URLs to <img> in office HTML sanitizer Codex review on #12934 caught that `data:` lived in the global `allowedSchemes`, which meant a smuggled `<a href="data:text/html, <script>...</script>">` would survive sanitization. The Sandpack iframe sandbox does not gate `target="_blank"` navigations, so a click would open attacker-controlled HTML in a new tab. Scope `data:` to `<img src>` only via `allowedSchemesByTag` (mammoth inlines DOCX images as base64 `data:image/...` URIs — that path still works). Add a regression suite (`sanitizeOfficeHtml security`) with 8 cases covering: <script> stripping, event-handler removal, javascript:/data: rejection on anchors, data:image preservation in <img>, http/https/mailto allowance, target=_blank rel=noopener enforcement, and <iframe> stripping. * 🔧 fix: Route extensionless office files by MIME alone Codex review on #12934 caught that the office-render gate in `extractCodeArtifactText` only fired when the extension was in `OFFICE_HTML_EXTENSIONS` or the category was `document`/`pptx`. A tool emitting `data` with `text/csv` (no extension) classifies as `utf8-text`, so the gate was skipped and raw CSV text shipped to the client — but the client routes by MIME to the SPREADSHEET bucket expecting a full HTML document, so the panel rendered broken text. Extract a shared `officeHtmlBucket(name, mime)` predicate from `html.ts` (returns the bucket name or null). Both `bufferToOfficeHtml` (the dispatcher) and the upstream gate in `extract.ts` now go through this single source of truth, so they can never drift apart again. The predicate already mirrors the dispatcher's extension/MIME logic (extension wins; MIME is the fallback for extensionless inputs). Adds: - 14 cases for the new `officeHtmlBucket` predicate covering the positive paths (each bucket via extension OR MIME) and the negative paths (txt, py, json, jpg, pdf, zip, odt, plain noext). - A direct regression test in `extract.spec.ts` for the Codex catch: `data` with `text/csv` + utf8-text category routes through the office HTML producer. - Parameterized cases for extensionless DOCX/XLSX/XLS/ODS/PPTX files identified by MIME alone. * 🛡️ fix: Enforce extension-wins precedence in officeHtmlBucket Codex review on #12934 caught that the predicate's if-chain interleaved extension and MIME checks for each bucket — e.g. CSV's branch was `ext === 'csv' || CSV_MIME_PATTERN.test(mimeType)`. A `deck.pptx` shipped with `text/csv` (sandboxed tools sometimes ship generic MIMEs) matched the CSV branch BEFORE the PPTX extension branch was reached, so a binary PPTX would have been handed to `csvToHtml` to parse as text — yielding garbage or a parse exception. Restructure to a strict two-pass dispatch: an exhaustive extension table first (one lookup, all known extensions), then MIME-only fallback for extensionless / unknown-ext inputs. The doc comment's "extension wins" claim is now actually enforced by the implementation. Add 7 regression cases covering the conflicting-MIME footgun for each bucket: deck.pptx + text/csv → pptx; workbook.xlsx + text/csv → spreadsheet; legacy.xls + pptx-MIME → spreadsheet; report.docx + text/csv → docx; data.csv + docx-MIME → csv; etc. * 🛡️ fix: Reject zip-bomb office files before in-process parsing (SEC) Addresses pre-existing availability vulnerability validated by SEC review (Codex finding 275344c5...) and made worse by this PR's HTML rendering path. A sub-1MiB compressed XLSX/DOCX/PPTX (highly compressed run-of-zeros) inflates to 200+ MiB of XML when handed to mammoth/xlsx — blocking the Node event loop for 10+ seconds and spiking RSS to ~1 GiB. The existing 8s `withTimeout` wrapper uses `Promise.race`, which can only return early; it cannot interrupt synchronous parser CPU/RAM consumption. PoC ran an authenticated execute_code call to OOM the API process. Add `assertSafeZipSize(buffer)` — a yauzl-based pre-flight that streams every entry with mid-inflate byte counting and bails on either a per-entry or total decompressed-size cap. Mid-inflate counting cannot be bypassed by falsifying the central directory's `uncompressedSize` field (the technique the PoC used). Defaults: 25 MiB per entry, 100 MiB total — generous headroom for legitimate image-heavy office files, well below the attack profile. Hook the check into every path that hands a buffer to mammoth/xlsx /yauzl: - New HTML producers (`wordDocToHtml`, `excelSheetToHtml`, `pptxToSlideListHtml`) — added by this PR - Legacy RAG text extractors (`wordDocToText`, `excelSheetToText` in `crud.ts`) — pre-existing path, also vulnerable Errors propagate as a tag-distinct `ZipBombError` so callers can distinguish a refused bomb from generic parse failures. The outer `extractCodeArtifactText` swallows the error and returns null, falling back to the regular download UI. `.xls` (BIFF/CFB binary, not ZIP) is detected by magic bytes and skipped — yauzl would reject it as malformed anyway. Adds 15 tests: - `zipSafety.spec.ts` (9): benign passes, per-entry cap, total cap, ZipBombError type-tagging, malformed-zip distinction, directory- entry handling, named-error surfacing, and the SEC-PoC pattern (sub-1 MiB compressed → 50 MiB inflated rejected on default caps). - `html.spec.ts` zip-bomb suite (5): each producer rejects a bomb; dispatcher propagates correctly; legitimate fixtures still render. - `extract.spec.ts` (1): outer extractor swallows ZipBombError and returns null so the download UI fallback fires. * 🧹 fix: Normalize MIME parameters; add legacy CSV MIME variant Two related Codex catches on PR #12934 — both about MIME-routing inconsistencies between backend and client that would cause extensionless CSV files to render as broken (raw text under an HTML slot) or skip the artifact panel entirely. P2 — backend MIME normalization: `officeHtmlBucket` matched MIME strings exactly, so a real-world `text/csv; charset=utf-8` Content-Type slipped through and the backend returned raw CSV text. The client's `baseMime` helper strips parameters before its own MIME lookup, so it routed the same file to the SPREADSHEET bucket expecting an HTML body that never arrived. Mirror the client's normalization on the backend (strip everything from `;` onward, lowercase) before bucket matching. P3 — client legacy CSV MIME: Backend's `CSV_MIME_PATTERN` accepts three variants (`text/csv`, `application/csv`, `text/comma-separated-values`); the client's `MIME_TO_TOOL_ARTIFACT_TYPE` only had the first two. An extensionless file with `text/comma-separated-values` would have backend HTML produced but the client would skip the artifact panel entirely. Add the missing variant. Tests: - 9 new parameterized-MIME cases on backend covering charset/ boundary/case variants for every bucket. - 1 new client routing case for `text/comma-separated-values`. * 🩹 fix: Try office HTML before short-circuiting on category=other Codex review on #12934 caught that the early `category === 'other'` return short-circuited before `hasOfficeHtmlPath` was checked. The classifier returns 'other' for inputs the new dispatcher can still route — extensionless `application/csv` (CSV MIMEs aren't in the classifier's text-MIME set and don't start with `text/`), and extensionless office MIMEs with parameters like `application/vnd... spreadsheetml.sheet; charset=binary` (the classifier's `isDocumentMime` exact-matches these MIMEs without parameter normalization). Both would route correctly through `officeHtmlBucket` but never reached it. Move the office-HTML attempt above the 'other' early return, and drop the `|| category === 'document' || category === 'pptx'` shortcut now that `hasOfficeHtmlPath` covers the same surface (with parameter normalization) and a wider one. ODT still routes through `extractDocument` unchanged — `hasOfficeHtmlPath` returns false for it and the `category === 'document'` branch below handles it. Adds 3 regression tests: - extensionless `application/csv` + category='other' → office HTML - extensionless parameterized office MIME + category='other' → office HTML - defense check: actual binary 'other' (image/jpeg) still returns null without invoking the office producer * 🛡️ fix: Office types are HTML-or-null (no text fallback → XSS) Codex P1 review on #12934 caught that when `renderOfficeHtml` failed (timeout, malformed file, zip-bomb rejection) for an office type, the extractor fell through to `extractDocument` and returned plain text. The client routes by extension/MIME to the office preview buckets and feeds `attachment.text` straight into the Sandpack iframe's `index.html`. A spreadsheet cell or document body containing the literal string `<script>alert(1)</script>` would have been injected as executable markup — direct XSS. The contract for office types is now HTML-or-null with no text fallback. Failed render returns null, the client's empty-text gate keeps the artifact off the panel, and the file falls back to the regular download UI (matching what PPTX already did). PDF and ODT still go through `extractDocument` because the client routes them to PLAIN_TEXT (which the markdown viewer escapes) or no artifact at all, so plain text is safe there. Test reshuffle: - `document` describe block now uses ODT/PDF for the legacy parseDocument-path tests (DOCX/XLSX/XLS/ODS bypass that path). - New "does NOT call parseDocument for office HTML types" test locks in the SEC contract for all four office HTML buckets. - "falls back to ..." tests rewritten as "returns null when ..." with explicit `parseDocumentCalls.length === 0` assertions to prove no text leaks back to the client. - New XSS regression test for the XLSX failure path. - Mock parseDocument failure-name match relaxed to `includes()` so ODT-named tests can use the same trigger. * 🧽 chore: Address follow-up review findings on PR #12934 Wraps up the 10-finding follow-up review. Two MAJOR + four MINOR + two NIT addressed; one NIT skipped after verifying it was a misread of the package.json structure. MAJOR - #1: Rewrite `renderOfficeHtml` JSDoc to document the HTML-or-null contract explicitly. The pre-fix doc described a text-fallback path that was the original XSS vector (commit b06f08a). A future maintainer trusting the stale doc could reintroduce the fallback. - #2: Replace byte-truncation of office HTML with a small "preview too large" banner document. Cutting at a UTF-8 boundary lands mid-tag (`<table><tr><td>con\n…[truncated]`) and ships malformed markup to the iframe — unpredictable rendering, occasional broken layouts on DOCX with embedded images / wide spreadsheets. MINOR - #4: Wrap `readSlidesFromZip`'s `zipfile.close()` in try/catch so a close-time exception (mid-flight stream) doesn't replace the original error. Mirrors the defensive pattern in zipSafety.ts. - #5: Refactor PPTX extraction to use `yauzl.fromBuffer` directly, eliminating the temp-file write/unlink the safety pre-flight already proved unnecessary. Removes 4 unused imports (os, path, fs/promises, randomUUID). - #6: Extract `isPreviewOnlyArtifact(type)` to `client/src/utils/ artifacts.ts` so the membership check is unit-testable without mounting the full Artifacts component (Recoil + Sandpack + media query). 15 new test cases covering positive types, negative types, null/undefined, and unknown strings. NIT - #3: Remove dead `stripColorStyles` / `COLOR_PROPERTY_PATTERN` — unused (sanitizer's `allowedStyles` config handles color implicitly). - #7: Remove dead `!_lc_csv_label` worksheet property write. - #9: Remove no-op `exclusiveFilter: () => false` sanitize-html config. - #10: Type-narrow `PREVIEW_ONLY_ARTIFACT_TYPES` to `ReadonlySet<ToolArtifactType>` so the membership table is compile-time checked against the enum. SKIPPED - #8: Reviewer flagged `sanitize-html` as duplicated in devDeps and dependencies. The package has no `dependencies` section — only `devDependencies` and `peerDependencies`. Existing convention (mammoth, xlsx, yauzl, pdfjs-dist) is to appear in BOTH. Removing the devDep entry would break local test runs. Tests: packages/api 4406/4406, client artifacts 128/128. * 🪞 chore: Fix isPreviewOnlyArtifact test description parameter order Follow-up review nit on PR #12934. Jest's `it.each` substitutes `%s` positionally, and the table rows were `[type, expected]` while the description template read `'returns %s for type %s'` — outputting "returns application/vnd.librechat.docx-preview for type true" instead of the intended "type ... returns true". Reorder the template to match the column order. Test runner output now reads naturally: "type application/vnd.librechat.docx-preview returns true". Pure cosmetic — runtime behavior unchanged. * ✨ feat: Improve DOCX rendering and surface filename in panel header Two UX improvements based on hands-on use of the office preview pipeline. DOCX rendering — mammoth strips the navy banners, cell shading, and column layouts that direct-formatted docs apply (python-docx-style output is a common case). The flat `<p><strong>X</strong></p>` and bare `<table><tr><td>` it emits looks washed out next to the source. Three targeted compensations: - Style map promotes `Title`, `Subtitle`, `Heading 1` thru `Heading 6`, and `Quote` paragraphs to their semantic HTML equivalents (mammoth's default only handles Heading 1-6, missing Title/Subtitle/Quote). - Extra CSS scoped to `.lc-docx` gives the first table row sticky- looking header styling regardless of `<thead>` (mammoth never emits `<thead>`), adds zebra striping, and treats the python-docx `<p><strong>X</strong></p>` section-heading idiom as a pseudo-h2 with a thin accent left border so document structure survives the round trip. Headings get a left accent or underline so they read as headings instead of just bold paragraphs. - Sanitizer's `allowedAttributes` opens `class` on the heading and block tags the styleMap and CSS heuristics rely on. `<script>`, event handlers, javascript: URLs, etc. are still stripped — the existing security regression suite catches any drift. Panel header — `Artifacts.tsx` showed a generic "Preview" pill for preview-only artifacts. Single-tab Radio is a no-op; surfacing the document filename there gives the user something useful in the chrome without taking real estate. `displayFilename` handles the sandbox dotfile suffix the upload pipeline applies. Tests: html.spec.ts +1 (new CSS-emission lock), 71/71. Backend files suite 428/428. Client 308/308. * ✨ feat: High-fidelity DOCX preview via docx-preview in iframe Switch the default DOCX render path from server-side mammoth → flat HTML to client-side `docx-preview` loaded inside the Sandpack iframe. Mammoth becomes the fallback for files above the cap. Why --- The Sandpack iframe is a real browser DOM. Server-side rendering ceiling for DOCX→HTML is well below the source's visual fidelity — mammoth strips cell shading, run colors, banners, and column layouts because Word's layout model doesn't fit HTML's flow model. Pushing the render into the iframe lifts that ceiling without paying the server-side cost of jsdom or LibreOffice. What ---- - New `wordDocToHtmlViaCdn(buffer)` builds a self-contained HTML doc that embeds the binary as base64 and lets `docx-preview@0.3.7` render it on load. CSS preserves dark/light mode handoff via `prefers-color-scheme`. Bootstrap script falls back to a "preview unavailable, please download" message if the CDN is unreachable or the parse throws. - `docx-preview` and its `jszip` peer dep are pinned to specific versions on jsdelivr with SRI sha384 integrity hashes and `crossorigin="anonymous"`. Refresh: re-fetch the file, run `openssl dgst -sha384 -binary FILE | openssl base64 -A`. - CSP locked down on the iframe: `default-src 'none'`, scripts only from jsdelivr (no eval), `connect-src 'none'` so a parser bug in docx-preview can't be turned into exfiltration of the embedded document, `base-uri 'none'`, `form-action 'none'`. Defense in depth on top of the Sandpack cross-origin sandbox. - `wordDocToHtml` dispatches by size: ≤ 350 KB binary → CDN path (high fidelity), larger → mammoth fallback (preserves the size cap on `attachment.text`). 350 KB chosen so worst-case base64-inflated output (~478 KB) plus wrapper overhead (~5 KB) fits under MAX_TEXT_CACHE_BYTES (512 KB) with 40 KB headroom. - Internal renderers exported as `_internal` for tests. Public API unchanged — callers still go through `wordDocToHtml`. PPTX intentionally NOT switched ------------------------------- Surveyed the available client-side PPTX libraries: - `pptx-preview@1.0.7` ships an ESM-only main entry plus a 1.36 MB UMD that references `require("stream"/"events"/"buffer"/"util")` — bundled for Node, not browser-clean. Could work but the runtime references to undefined Node globals are a fragility risk worth more validation than this PR can absorb. - `pptxjs` is jQuery-era, requires four separate UMD scripts in a specific order, less actively maintained. - The honest answer for PPTX is the LibreOffice sidecar (DOCX/XLSX/ PPTX → PDF → PDF.js), which is the architecture every major product (Google Drive, Claude.ai, ChatGPT) effectively uses and the only path to ~5/5 fidelity for arbitrary user decks. PPTX stays on the existing slide-list extraction for now. Open a follow-up issue for the LibreOffice/Gotenberg sidecar. Tests ----- - 6 new in CDN-rendered describe block: wrapper structure, base64 round-trip, SRI integrity + crossorigin, CSP locks (connect-src/eval/base-uri/form-action), fallback message wiring, size-threshold lock. - Adjusted 2 existing tests that asserted on mammoth-path artifacts (literal document text in `<article class="lc-docx">`) — those assertions move to the mammoth-fallback test that calls `_internal.wordDocToHtmlViaMammoth` directly. Dispatcher tests now assert CDN-path signatures instead. packages/api files: 434/434 ✅, full unit suite 4473/4473 ✅. * 🧷 fix: Address Codex P1 (MIME aliases) + P2 (CDN dependency) Two follow-up review findings on PR #12934, both real. P1 — Spreadsheet MIME aliases on client ---------------------------------------- Backend's `officeHtmlBucket` uses the broad `excelMimeTypes` regex from `librechat-data-provider` (covers `application/x-ms-excel`, `application/x-msexcel`, `application/msexcel`, `application/x-excel`, `application/x-dos_ms_excel`, `application/xls`, `application/x-xls`, plus the canonical sheet MIMEs). The client's exact-match `MIME_TO_TOOL_ARTIFACT_TYPE` only had three of those, so an extensionless XLS upload with a legacy MIME would have backend HTML produced but the client would fail to route the artifact at all — preview chip never registers. Fix: import the same regex on the client and add it as a fallback in `detectArtifactTypeFromFile` after the exact-match map miss. Stays in lock-step with the backend automatically. 7 new test cases — one per legacy alias. P2 — Hard CDN dependency on jsdelivr ------------------------------------- Air-gapped / corporate-filtered networks where jsdelivr is unreachable would see DOCX previews permanently degrade to "Preview unavailable" because the iframe could never load the renderer scripts. Mammoth was sitting right there on the server but the dispatcher always preferred the CDN path for files under 350 KB. Fix: `OFFICE_PREVIEW_DISABLE_CDN` env var. When truthy (`1`, `true`, `yes`, case-insensitive, whitespace-trimmed), `wordDocToHtml` short-circuits to the mammoth path regardless of file size. Operators on filtered networks set the env var; default behavior is unchanged. Read at function-call time (not module load) so jest can flip it in `beforeEach` without `jest.resetModules()`. The cost is one property access per render. 12 new test cases: env-unset uses CDN (default), all five truthy forms force mammoth, five non-truthy forms (`false`/`0`/`no`/empty/ arbitrary string) leave CDN active. Tests ----- packages/api/src/files: 446/446 ✅ (was 434, +12 from env-var matrix). client artifact suites: 235/235 ✅ (was 228, +7 from MIME aliases). * ✨ feat: High-fidelity PPTX preview via pptx-preview in iframe Mirrors the DOCX CDN architecture for PPTX: small files (≤350 KB binary) embed as base64 and render via `pptx-preview` loaded from jsdelivr inside the Sandpack iframe. Larger files and air-gapped deployments fall back to the existing slide-list extraction. Why --- PPTX is the format where the gap between LibreChat's preview and Claude.ai-style previews was most visible (slide-list of bullet points vs. rendered slide layouts). LibreOffice → PDF → PDF.js is still the eventual gold-standard answer for PPTX fidelity, but client-side rendering inside the Sandpack iframe gets us a meaningful intermediate step (~1.5/5 → ~3.5/5) without a sidecar. What ---- - `pptx-preview@1.0.7` (ISC license, ~1.36 MB UMD bundle that includes its echarts/lodash/uuid/jszip/tslib deps inline). Pinned to a specific version on jsdelivr with SHA-384 SRI and `crossorigin="anonymous"`. - `buildPptxCdnDocument` mirrors the DOCX wrapper: same CSP locks (`default-src 'none'`, `connect-src 'none'`, no eval, no base/form tampering), same `id="lc-doc-data"` base64 slot, same fallback message wiring (`typeof pptxPreview === 'undefined'` → "Preview unavailable"). - New public `pptxToHtml(buffer)` dispatcher; `bufferToOfficeHtml` switches its `'pptx'` case to call it. `pptxToSlideListHtml` stays exported as the slide-list-only path (still hit by tests directly and by the dispatcher fallback). - `OFFICE_PREVIEW_DISABLE_CDN=true` env-var hatch applies to PPTX too — air-gapped operators get the slide-list path. Same env-var read at call time, same matrix of truthy values (`1` / `true` / `yes` / case-insensitive / whitespace-trimmed). - `_internal` re-exports moved to after the PPTX section since the PPTX internals live further down in the file. Adds `pptxToHtmlViaCdn`, `MAX_PPTX_CDN_BINARY_BYTES`, `PPTX_PREVIEW_CDN`. Honest caveats -------------- - The 1.36 MB UMD bundle has `require("stream"/"events"/"buffer"/ "util")` references in its outer wrapper. Those are bundled-dep artifacts (likely from `tslib` / Node-shim transforms) and don't appear to execute on the browser code paths, but I haven't done manual e2e on a wide range of decks. If a class of files turns up that breaks rendering, the iframe-side fallback message catches it and operators have `OFFICE_PREVIEW_DISABLE_CDN=true` as the bail. - First-render CDN fetch is ~1.36 MB (browser-cached after). - PPTX with embedded media easily exceeds the 350 KB binary cap; those files take the slide-list path. Lifting the cap is a follow-up (tied to the broader self-hosting work). Tests ----- 11 new in two new describe blocks: - `pptxToHtml dispatcher`: routing predicate (small → CDN, env-set → slide-list). - `CDN-rendered path`: base64 round-trip, SRI integrity + crossorigin, CSP locks (connect/eval/base/form), fallback message, size-threshold lock at 350 KB. - `OFFICE_PREVIEW_DISABLE_CDN escape hatch`: env-var matrix for truthy values. packages/api/src/files: 457/457 ✅ (was 446, +11). * 🪟 fix: DOCX preview fills the artifact panel width docx-preview defaults to rendering at the document's native page width (8.5in for letter, 21cm for A4). In a wide artifact panel that left whitespace on either side; in a narrow one it forced horizontal scroll. Two changes: - Pass `ignoreWidth: true` to `docx.renderAsync` so the library skips the document's pageSize width and uses its container's width. - Defensive CSS overrides on `.docx-wrapper` and `.docx-wrapper > section.docx` in case a future library version regresses on the option, plus `padding: 0` on the wrapper to drop the page-edge whitespace docx-preview otherwise reserves. `renderHeaders`/`renderFooters`/etc. stay enabled — those still appear in the rendered output, just inside a container that fills the panel instead of a fixed-width "page." Tests unchanged (100/100); manual e2e ahead of merge. * 🩹 fix: PPTX black screen — allow blob: workers + harden bootstrap Manual e2e of the PPTX CDN renderer surfaced a black screen with "Could not establish connection. Receiving end does not exist." unhandled-rejection — characteristic of a Web Worker that couldn't start. Root cause: pptx-preview's bundled echarts dep spins up Web Workers via blob: URLs for chart rendering. Our CSP had `default-src 'none'` and no `worker-src`, so workers fell back to default → blocked. The async failure deep inside echarts didn't surface through the outer `previewer.preview()` promise, so my bootstrap's `.catch` never fired, the loading state was removed, and the iframe sat with the body background showing through (dark navy in dark mode = "black screen"). Three changes: - Add `worker-src blob:` to the PPTX CSP. Allows blob:-only worker creation without permitting arbitrary worker URLs. - Bootstrap: window-level `unhandledrejection` and `error` listeners so rejections from inside bundled-dep async pipelines surface as the user-facing "Preview unavailable" fallback instead of going silent. - Bootstrap: 8-second timeout that checks `container.children.length` — if the renderer hasn't appended anything visible by then, assume silent failure and show the fallback. Also wipe `container.innerHTML` when showing the fallback so a partial render doesn't compete with the message. DOCX wrapper unchanged: docx-preview doesn't use workers, so the worker-src directive doesn't apply, and the existing fallback path already covers its failure modes. Tests ----- - Existing PPTX CSP test now also asserts `worker-src blob:` is present. - Existing fallback-message test extended to cover the new unhandledrejection/error/timeout listeners. packages/api/src/files: 467/467 ✅. * 🔒 fix: gate office HTML routing on backend trust flag (textFormat) Codex P1 review on PR #12934: routing .docx/.csv/.xlsx/.xls/.ods/.pptx into the office preview buckets assumed `attachment.text` was already sanitized full-document HTML, but that guarantee only existed for the new code-output extractor path. Existing stored attachments and other non-code paths can still carry plain extracted text — `useArtifactProps` would then inject that as `index.html` inside the Sandpack iframe. Adds a `textFormat: 'html' | 'text' | null` trust flag persisted on the file record by the code-output extractor, surfaced over the SSE attachment payload and the TFile API type. The client's routing in `detectArtifactTypeFromFile` requires `textFormat === 'html'` before landing on an office HTML bucket; everything else (legacy attachments, RAG-extracted plain text from `parseDocument`, explicitly-marked 'text' entries) falls back to the PLAIN_TEXT bucket where the markdown viewer escapes content rather than executing it. Tests: new `getExtractedTextFormat` helper has 14 cases covering all office paths, legacy XLS MIME aliases, parseDocument fallthroughs, and null-input. Client `artifacts.test.ts` adds three security-gate tests proving downgrade behavior for missing/null/'text' textFormat, plus a `fileToArtifact` test that legacy office attachments without the flag end up in PLAIN_TEXT with their content escaped. * 🌐 fix: air-gapped DOCX preview — embed mammoth fallback in CDN doc Codex P2 review on PR #12934: the CDN-rendered DOCX path always pulled docx-preview + jszip from cdn.jsdelivr.net. Air-gapped or corporate- filtered networks where jsdelivr is blocked would degrade to a static "Preview unavailable" message even though the server already had a local mammoth renderer that could produce readable output. Now the dispatcher renders mammoth first and embeds the sanitized output inside the CDN document as a hidden `#lc-fallback` block. The iframe's existing `typeof docx === 'undefined'` check (which fires when the CDN scripts can't load) un-hides the fallback so the user sees a real preview. CDN-success path is unchanged: high-fidelity docx-preview output owns the viewport, mammoth fallback stays hidden. Two new safeguards in the dispatcher: - Size budget: if base64(binary) + mammoth body + wrapper > 512 KB (the `attachment.text` cache cap), drop to mammoth-only so a giant document still renders. The `OFFICE_HTML_OUTPUT_CAP` constant mirrors `MAX_TEXT_CACHE_BYTES` from extract.ts (separate constant to avoid a circular import; pinned by a unit test). - `lc-render` is hidden when fallback shows so the empty padded slot doesn't sit above the mammoth content. Tests: existing CDN-path tests updated for the new `wordDocToHtmlViaCdn(buffer, mammothBody)` signature; new test for the embedded fallback structure (`#lc-fallback`, mammoth body content, "High-fidelity renderer unavailable" notice, render-slot hide); new constant pin and per-fixture cap-respect assertion. * 🧪 feat: LibreOffice → PDF preview path (POC, opt-in via env) Per the plan-mode discussion: prove out a LibreOffice subprocess pipeline as an alternative to the docx-preview / pptx-preview CDN renderers. LibreOffice handles every office format Microsoft and LibreOffice itself can open (DOCX, PPTX, XLSX, ODT, ODP, ODS, RTF, many more), produces a PDF, and the host browser's built-in PDF viewer renders it inside the Sandpack iframe via a `data:` URI. No client-side JS dependency, no CDN dependency, true high fidelity for any feature LibreOffice supports. Off by default. Operators opt in by setting both: - `OFFICE_PREVIEW_LIBREOFFICE=true` - LibreOffice (`soffice` or `libreoffice`) on the server's `$PATH` When either is missing, the dispatcher falls through to the existing CDN/mammoth/slide-list pipeline so a misconfiguration doesn't break previews. Hardening (`packages/api/src/files/documents/libreoffice.ts`): - Fresh subprocess per call with isolated temp dir, stripped env (PATH/HOME/TMPDIR only), and `-env:UserInstallation` so concurrent conversions can't collide on shared `~/.config/libreoffice` locks - 30-second wall-time cap; SIGKILL on timeout - 50 MB PDF output cap to bound disk pressure - 512 KB output cap on the wrapped HTML so the SSE/cache contract stays intact (base64 inflates ~33%, effective PDF cap ~380 KB) - Macros disabled by default flags (`--norestore --invisible --nodefault --nofirststartwizard --nolockcheck`) - Tag-distinct `LibreOfficeUnavailableError` / `LibreOfficeConversionError` so callers can swallow appropriately Iframe wrapper (`buildPdfEmbedDocument`): - Native browser PDF viewer via `<iframe src="data:application/pdf; base64,...">` — works in Chrome, Edge, Safari, Firefox - CSP locks the iframe to `default-src 'none'; frame-src data:; connect-src 'none'; script-src 'unsafe-inline'` — no outbound network, no eval, no external scripts - `#view=FitH` for first-paint sizing - 4-second heuristic timer that swaps to a "Preview unavailable" fallback when the browser's PDF viewer is disabled (kiosk mode, Brave Shields, etc.) Wired into `wordDocToHtml` and `pptxToHtml` as the first branch — returns null when disabled / unavailable / oversized so the existing pipeline takes over. XLSX intentionally NOT routed through this path: SheetJS's HTML output is already excellent for spreadsheets (sortable, sticky headers) and PDF rendering of sheets is awkward. Tests (`libreoffice.spec.ts`, 30 cases — 25 always run, 5 conditional on the binary): env-gating parser semantics matching `OFFICE_PREVIEW_DISABLE_CDN`, fallthrough contract (never throws, returns null on any failure), CSP lock-down, fallback structure, binary probe caching + missing-binary path, error tagging, and integration tests that engage when `soffice`/`libreoffice` is on PATH (DOCX→PDF, PPTX→PDF, output-cap fallthrough). Integration tests skip cleanly on bare CI. * 🩹 fix: CI — preserve legacy download path for empty-text office attachments Two regressions surfaced after the textFormat security gate landed. 1. **Client** (`LogContent.test.tsx` "falls back to the legacy download branch for an office file with no extracted text"): When the security gate downgraded an office type without `textFormat: 'html'` to PLAIN_TEXT, the lenient empty-text gate on PLAIN_TEXT then accepted a missing `text` field and rendered a half-empty panel card. The historical contract is "office type + no text → legacy download UI"; the downgrade should only fire when there's actual plain text that needs safe-escaping. Fix in `detectArtifactTypeFromFile`: short-circuit to null when the office type lands in the security-gate branch with no text. The PLAIN_TEXT downgrade still fires for legacy attachments that DO carry plain text. 2. **API** (`process.spec.js` + `process-traversal.spec.js`): the `@librechat/api` mocks didn't expose `getExtractedTextFormat`, so `processCodeOutput` called `undefined(...)` → TypeError → tests got undefined results. Added the helper to both mocks with a faithful default (returns 'text' for non-null extractor output, null otherwise). Tests: new regression in `artifacts.test.ts` pinning the empty-text + no-textFormat → null contract for all four office types (.docx/.csv/.xlsx/.pptx), so a future refactor can't silently re-introduce the half-empty card. * 🩹 fix: PPTX slides scale to fit panel width (no horizontal scroll) Manual e2e on PR #12934: pptx-preview rendered slides at their native init dimensions (960×540 default). The artifact panel is much narrower than that, so the iframe got a horizontal scrollbar and only a corner of each slide showed at any time — the user had to drag-scroll across each slide to read it. Fix: keep pptx-preview's init at 960×540 so its internal layout math stays correct, then post-process each rendered slide: - Cache the slide's native width/height on its dataset BEFORE applying any transform (so subsequent re-fits don't measure the already-transformed box). - Wrap the slide in `.lc-slide-wrap` with explicit width/height set inline to the scaled dimensions; the wrap shrinks the layout space the slide occupies. - Apply `transform: scale(panel_width / 960)` to the slide itself with `transform-origin: top left` so the rendered output shrinks from the top-left corner into the wrap. - Cap the scale at 1.0 so small slides don't upscale and get blurry. Streaming + resize: - `MutationObserver` watches the container for slide insertions so streaming renders get scaled on arrival rather than waiting for the entire `previewer.preview` promise to settle. - `ResizeObserver` re-fits all wrapped slides when the iframe resizes (panel drag, window resize). Tests: new "bootstrap wraps + scales each slide" lock in the wrap class, scale computation, observer setup, and native-size caching so a future refactor can't silently re-introduce the overflow. * 🩹 fix: PPTX wrap+scale runs after preview, not during streaming Manual e2e on PR #12934: regenerated PPTX showed "Preview unavailable" in the iframe. Root cause: the MutationObserver I added in the previous commit fired during pptx-preview's render and moved slides out from under the library's references. pptx-preview's async pipeline raised an unhandled rejection, the iframe's window-level listener caught it, and the fallback message replaced the partial render. Fix: drop the MutationObserver. Apply the wrap+scale ONCE in a `finalize` step that runs: - On `previewer.preview().then` (the happy path) - On the 8-second timeout safety net IF the container has children (silent-failure path — pptx-preview emitted slides but never resolved its outer promise) To prevent the user from seeing an unscaled flash while pptx-preview renders into the 960px-wide canvas, the container is set to `visibility: hidden` at init and only revealed inside `finalize` after wrap+scale completes. Resize handling stays via `ResizeObserver` on `document.body`, installed AFTER the wrap pass so it doesn't fire during the wrap itself. Tests: regression assertion now also locks in: - `container.style.visibility = 'hidden' / 'visible'` (the flash- prevention contract) - Absence of MutationObserver (the bug we just removed — must NOT creep back in via a future "let's scale during streaming" idea) * 🩹 fix: PPTX slides fill panel width (drop upscale cap, per-slide scale) Manual e2e on PR #12934: slides rendered correctly but didn't fill the artifact panel — whitespace on either side. Two issues: 1. The scale was capped at `Math.min(1, available / SLIDE_W)`. On panels wider than 960px, the cap clamped the scale to 1.0 and slides rendered at native size with whitespace on the sides instead of stretching. 2. The scale was computed against the constant `SLIDE_W = 960`, but pptx-preview can emit slides whose `offsetWidth` differs from the init param if the source PPTX has a non-16:9 layout. Per-slide division of `available / nativeW` handles that case. Fix: replace `computeScale()` with two helpers — `availableWidth()` returns the panel content-box width and `scaleFor(nativeW)` returns the per-slide scale. No upscale cap. The slide content is rendered by pptx-preview against its 960×540 canvas using vector text / canvas — scaling up to e.g. 1500px doesn't visibly degrade quality. Tests: regression now also asserts: - `availableWidth()` and `scaleFor()` exist by name - The exact scale formula `availableWidth() / (nativeW || SLIDE_W)` - Negative assertion that `Math.min(1, ...)` is NOT present, so a future "let's add an upscale cap" rewrite can't silently re-introduce the whitespace. * 🩹 fix: PPTX preview fills panel height (no white gap below slides) Manual e2e on PR #12934: PPTX preview filled the panel width but left empty space below the last slide. DOCX didn't have this issue because its content (mammoth-rendered HTML) flows naturally and either fits exactly or overflows; PPTX slides are fixed-aspect 16:9 and don't grow with the panel. Two changes: 1. **Body fills the iframe viewport** — `html, body { min-height: 100vh }` plus `body { display: flex; flex-direction: column }` and `#lc-render { flex: 1 0 auto }`. The dark theme bg now fills the iframe even when total slide content is shorter than the panel, so a single-slide deck never reveals a "white below" gap. 2. **Per-slide scale honors viewport height** — `scaleFor(nativeW, nativeH)` now returns `min(width-fit, height-fit)` (largest factor that fits without overflowing either dimension). On a tall artifact panel with a short deck, slides grow up to the full panel height instead of staying at the width-bound size. Existing height-fit was always considered correct conceptually but the previous implementation only used width-fit, leaving half the viewport unused per slide. Tests: regression now also asserts `availableHeight()`, the `Math.min(sw, sh)` formula, and `min-height: 100vh` are in the bootstrap. Negative assertion for the old `Math.min(1, ...)` upscale cap remains. * 🩹 fix: revert body flex on PPTX bootstrap (caused black-screen render) Manual e2e regression on PR #12934: the previous commit added `body { display: flex; flex-direction: column }` plus `#lc-render { flex: 1 0 auto }` to fill the panel height. Side effect: pptx-preview's internal layout assumes block flow on its ancestor elements; making body a flex container caused slides to render as solid-black rectangles (sized correctly, but with no visible content inside). Fix: keep just `html, body { min-height: 100vh }` for the bg-fill effect — that alone gives empty space below short decks the dark theme bg without changing flow. Drop the body-flex and the `#lc-render { flex: 1 0 auto }` directives. The height-aware `scaleFor(nativeW, nativeH)` from the same commit stays — it doesn't interact with pptx-preview's layout, just chooses a per-slide scale. Each slide still grows to fit the viewport contain-style. Negative-assertion added to the regression test: `body { display: flex }` must NOT appear in the bootstrap, so a future "let's flex the body to make height work" rewrite can't silently re-introduce this. (Note: the user also flagged DOCX theming as faint body text; I'm leaving that for now per their note that it may be pre-existing. Not addressed in this commit.) * 🩹 fix: revert PPTX height-fill changes; lock DOCX CDN to light scheme Two fixes for separate manual e2e regressions on PR #12934. **1. PPTX black screen (single slide rendering as solid black).** The previous fix removed `body { display: flex }` thinking that was the sole cause, but the regression persisted. Bisecting against the last known-good commit ( |
||
Yashwanth Alapati
|
3da1d8c961
|
🔍 feat: add Tavily as Search and Scraper Provider (#12581)
* feat: add Tavily integration as search provider and scraper provider * chore:update tavily web search parameters * chore:tavily paramer update * chore:update data-schemas test for tavily * fix: allow Tavily string option modes * fix: align Tavily config options * fix: scope Tavily scraper timeout * fix: use resolved scraper provider timeout * fix: widen Tavily search provider types * fix: harden Tavily web search config * fix: cap Tavily option timeouts --------- Co-authored-by: Danny Avila <danny@librechat.ai> |
||
|
github-actions[bot]
|
9b376178a6
|
🌍 i18n: Update translation.json with latest translations (#12916)
Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> |
||
|
Danny Avila
|
619f28d76d
|
🛡️ fix: Sanitize HTML In Admin Banner And MCP Config Dialog (#12927)
Two `dangerouslySetInnerHTML` sites rendered admin-supplied HTML without sanitization: - `Banner.tsx` rendered `banner.message` directly. - `MCPConfigDialog.tsx` rendered each `customUserVars` description. Wrap both with DOMPurify, allowing only the inline tags needed for formatting (links, emphasis, line breaks). Hardens against compromised admin or yaml supply-chain scenarios. Pattern matches the existing `CustomUserVarsSection.tsx` and `Tooltip.tsx` sanitizer setup. |
||
|
Danny Avila
|
37429e8a3e
|
🚦 feat: Make URL Auto-Submit Configurable (#12929)
`/c/new?prompt=…&submit=true` previously auto-submitted the prompt unconditionally. For deployments where users may receive crafted links from external sources, an authenticated victim's click can trigger an immediate, attacker-controlled prompt against a memory- or tool-enabled model — providing a 1-click vector for prompt-injection exfiltration via markdown image rendering. Add `interface.autoSubmitFromUrl` (default `true` to preserve current behavior). Operators handling sensitive memory/tool data can set it to `false` so URL-supplied prompts only pre-fill the composer; the user must press Send explicitly. |
||
|
Danny Avila
|
756530c2b8
|
🩹 fix: Polish code-execution attachment UX (#12870)
* 🧹 chore: Strip code-execution boilerplate from tool output
The bash executor in `@librechat/agents` appends two kinds of noise to
every successful run:
1. Trailing `Note:` paragraphs — long behavioral hints repeating
rules already in the system prompt ("Files from previous executions
are automatically available...", "Files in 'Available files' are
inputs..."). Re-stating these on every tool call adds ~50 tokens of
waste per call, which compounds across long agent traces.
2. Per-file `| <annotation>` suffixes on every line of `Generated
files:` / `Available files (...):`. The two section headers already
convey the new-vs-known distinction; the per-file annotations are
redundant *and* phrased inconsistently ("downloaded by the user"
vs. "displayed to the user" vs. "known to the user").
Strip both in a small `cleanCodeToolOutput` helper invoked from
`packages/api/src/agents/handlers.ts` for every tool listed in
`CODE_EXECUTION_TOOLS`. Non-code-execution tools pass through
unchanged. The cleaning happens *after* tool resolution but *before*
downstream consumers (model context, SSE forwarding, persistence) see
the content, so subsequent model turns get the lean output.
* 🩹 fix: Polish code-execution attachment rendering
Three rough edges visible in code-interpreter conversations:
1. **Sandbox-internal `.dirkeep` placeholders leak as file chips.** The
bash executor creates `.dirkeep` inside any new directory so the
stateless container preserves the folder across executions. After
`sanitizeArtifactPath`'s `_` prefix and 6-hex collision suffix it
surfaces as `_.dirkeep-<hash>` — a 0-byte chip with no value to the
user, sometimes hiding the real artifact behind it. New
`isInternalSandboxArtifact` helper filters them out of every
routing path (`Attachment`, `AttachmentGroup`, `LogContent`).
2. **The `-<hash>` collision suffix is visible in chip labels.** The
suffix is collision-avoidance machinery; users only need to see the
canonical name. New `displayFilename` strips it for display while
leaving the on-disk `attachment.filename` untouched so downloads
resolve. Applied across `FileContainer`, `ToolArtifactCard`,
`ToolMermaidArtifact`, and `LogContent`'s text-attachment label
path.
3. **0-byte / placeholder files outrank real artifacts in render
order.** Bucket sort by salience (non-empty before empty) sinks
stragglers to the bottom. Stable sort preserves arrival order for
peers.
Added regression tests cover the new helpers, the dirkeep filter
across buckets, and the within-bucket salience ordering.
* 🩹 fix: Don't auto-open artifact panel on history navigation
Navigating to a previous conversation full of code-execution artifacts
would auto-open the side panel and focus the most-recent artifact —
the same code path that fires for fresh streaming artifacts. Users
expect that "auto-open" behavior only when an artifact arrives via
SSE, not when they revisit an old chat.
Two-part gate:
1. `ToolArtifactCard`'s focus effect captures `isSubmitting` at first
render via a ref. A card mounted *during* a stream means a new
artifact arrived → steal panel focus (legacy behavior). A card
mounted while `isSubmitting === false` is part of conversation
history → leave focus alone.
2. `Presentation`'s panel-render condition gains `currentArtifactId
!= null`. With (1) keeping `currentArtifactId` null on history
load, the panel stops rendering at all on navigation — even if
`artifactsVisibility` was left `true` by a prior conversation.
User clicks on a chip to re-open (the click handler is unchanged
and unconditional).
Test seeds `isSubmittingFamily(0)` per case: existing tests opt into
streaming (default `true`) so legacy auto-focus assertions still hold;
new tests for history-load opt into `streaming: false` and verify
no auto-focus + click-to-open still works.
* 🩹 fix: Force panel visible on streaming artifact arrival
The previous commit gated `setCurrentArtifactId` on `isSubmitting` but
left `artifactsVisibility` untouched. When a user had explicitly
closed the panel earlier in the session, a fresh SSE artifact would
set `currentArtifactId` (so the chip read "click to close") but
`Presentation`'s render condition still required `visibility === true`
— net effect: the card claimed to be open, the panel stayed hidden.
Streaming arrivals now also call `setVisible(true)`, which is the
explicit "auto-open when first created" behavior the user asked for.
History mounts (`isSubmitting === false`) still leave both focus and
visibility alone, so navigating to an old conversation does not
re-open the panel.
Two regression tests added: one asserts streaming flips visibility on
even when seeded false, the other asserts history mounts leave a
seeded-false visibility alone.
* 🧹 chore: Tighten code-execution attachment polish per audit feedback
Resolves the eight actionable findings from the comprehensive audit:
- Scope `displayFilename` out of `FileContainer`: opt-in via a new
`displayName` prop. User-uploaded chips (input area, persisted
message files) keep their raw filename, eliminating the false-positive
class where `report-abc123.pdf` was silently rewritten to `report.pdf`.
Code-execution artifact paths in `Attachment.tsx` explicitly compute
the de-suffixed name and pass it through.
- Tighten `TRAILING_NOTES_PATTERN` to anchor on the two known boilerplate
openings (`Files from previous executions`, `Files in "Available files"`),
so a user-authored `Note:` line preceded by a blank line in stdout no
longer gets eaten along with everything after it.
- `ToolMermaidArtifact`: compute `visibleFilename` once and reuse for
title, content, and the download `aria-label` (was using the raw
`attachment.filename` for the aria-label, creating a screen-reader
inconsistency).
- `ToolArtifactCard`: read `isSubmittingFamily(0)` once via a
non-subscribing `useRecoilCallback`, instead of subscribing for the
full lifetime to a value the ref only ever needs at first render.
- Extract `bySalience` and `byEntrySalience` comparators from
`attachmentTypes.ts`, replacing the ten duplicated sort lambdas in
`Attachment.tsx` and `LogContent.tsx`.
- Treat `attachmentSalience({ bytes: undefined })` as neutral (`0`)
rather than empty (`1`); only an explicit `bytes === 0` sinks. Stops
non-code-exec sources (web-search inline results, files where the
schema omits the byte count) from silently sinking past real content.
- Pin the click-history test to the panel-open button by name instead
of relying on `getByRole('button', { pressed: false })`, which
matched by DOM order.
- Add the missing blank line between adjacent `it(...)` blocks.
- Drop the verbose narrating comments in `FileContainer` along with the
removed `displayFilename` import.
Adds three regression tests for the new behavior (FileContainer raw
filename, artifact-context displayName flow, user-authored `Note:` line
preserved through cleanup) and updates the salience test for the new
neutral-undefined semantics.
* 🧹 chore: Drop redundant `@testing-library/jest-dom` import in FileContainer spec
`client/test/setupTests.js` already imports the matchers globally for every
Jest test in the client workspace, so the explicit import here was dead code.
Removing it brings the spec in line with the broader convention used by
`ArtifactRouting.test.tsx`, `LogContent.test.tsx`, and `attachmentTypes.test.ts`.
* 🛡️ fix: Narrow `.dirkeep`/`.gitkeep` filter to the sandbox-specific form
`isInternalSandboxArtifact` was filtering bare `.dirkeep` / `.gitkeep`
along with the post-sanitization form. Bare versions never originate
from the bash executor (the dotfile rewrite + disambiguator step in
`sanitizeArtifactPath` always produces `_.dirkeep-<6 hex>`), so the only
real-world source of a bare `.gitkeep` is project scaffolding the user
uploaded — silently hiding it from every attachment bucket meant the
file disappeared with no way to surface or download it.
Tightening to `^_\.(?:dirkeep|gitkeep)-[0-9a-f]{6}$` keeps the
sandbox-placeholder filter intact while letting user-uploaded markers
render normally. Tests inverted accordingly: bare forms now expected to
render; only the post-sanitization form is filtered.
* 🩹 fix: Address comprehensive-review findings on attachment helpers
Five findings from the latest pass:
- **MAJOR — `displayFilename` false-positive on extensionless 6-hex.**
The previous regex `/-[0-9a-f]{6}(?=\.[^.]+$|$)/` stripped any leaf
ending in `-XXXXXX` regardless of context, so a user-named
`build-a1b2c3` (script-emitted hash artifact, no extension) lost its
tail and rendered as `build`. Split into two narrower patterns:
`COLLISION_SUFFIX_BEFORE_EXT` only matches when followed by an
extension; `SANITIZED_DOTFILE_TRAILING_SUFFIX` only fires when the
leaf starts with `_.` AND ends with `-XXXXXX` — the unambiguous
fingerprint of `sanitizeArtifactPath`'s dotfile rewrite.
- **MINOR — `isInternalSandboxArtifact` filter too aggressive.**
`(file.bytes ?? 0) > 0` treated undefined bytes as zero, falling
through to the regex check. Tightened to `file.bytes !== 0`: only
an *explicit* zero counts as the empty-placeholder shape worth
hiding. Non-code-exec sources without `bytes` populated render
normally now.
- **MINOR — `getValue()` could throw on a degenerate atom state.**
Switched the snapshot read in `ToolArtifactCard` to
`valueMaybe() ?? false` so a transient error / loading state on the
upstream selector doesn't crash card mount. The `false` default is
the right history-fallback (don't auto-open if we can't classify).
- **NIT — `attachmentSalience` / `bySalience` over-broad signature.**
Removed the test-only `{ bytes?: number }` arm; functions now accept
`TAttachment` directly. The internal `bytes` read still goes through
a cast since not every TAttachment branch declares it. Tests updated
to use the existing `baseAttachment(...)` helper.
- **MINOR — Missing regression test for extensionless 6-hex.**
Added `'build-a1b2c3'` and `'out/blob-deadbe'` cases that pin the
preservation behavior, plus an `isInternalSandboxArtifact` test that
asserts undefined-bytes attachments are not filtered.
* 🩹 fix: Make code-file artifacts click-to-open only
Removes mount-time auto-open from `ToolArtifactCard`. Streaming
arrivals no longer hijack the panel — even a freshly-emitted SSE
artifact registers silently in `artifactsState` and waits for the
user to click. Combined with `Presentation`'s
`currentArtifactId != null` render gate, the panel stays closed
across history navigation, page reload, and SSE arrival.
Click is the only path that opens the panel. `handleOpen` is
unchanged: first click focuses + reveals, second click on the same
chip closes.
Dropped:
- `useRecoilCallback` snapshot read of `isSubmittingFamily(0)`
- `mountedDuringStreamRef` ref + lazy-init block
- The whole focus + visibility effect (was effect 3)
- `useRef` import (now unused)
Tests:
- `ArtifactRouting.test.tsx` rewritten to exercise the click path:
registers-on-mount-without-focus, click-to-open-then-close, multi-
card-no-auto-focus, click-when-visibility-was-false. The streaming
state is no longer seeded; both `renderWith` and `renderWithProbe`
collapsed back to plain `RecoilRoot`.
- `LogContent.test.tsx` flips its panel-routing assertions from
`pressed: true` (which asserted auto-focus) to `pressed: false`
with a chip-title check (which asserts the panel card rendered
but stayed unfocused).
* Revert "🩹 fix: Make code-file artifacts click-to-open only"
This reverts commit
|
||
Helge Wiethoff
|
61b9b1daa7
|
🩹 fix(SSE): Treat responseCode === 0 as Transport Failure, Not Server Error (#12834)
* fix(sse): treat responseCode===0 as transport failure, not server error
When a long-running model response (e.g. gpt-5.4 with web_search:true)
takes longer than the browser's idle connection timeout, the SSE transport
drops and sse.js fires an error event with responseCode=0 and e.data set
to the raw response buffer (non-JSON SSE text).
The previous guard `!responseCode` is truthy for both 0 (transport drop)
and undefined (genuine server-sent error event), so the client incorrectly
entered the server-error branch, tried to JSON.parse raw SSE text, logged
"Failed to parse server error", and showed the user a red error banner --
even though the backend continued processing and delivered the final answer
seconds later.
Fix 1 (client): change guard from `!responseCode` to `responseCode == null`
so that only undefined/null (no HTTP status at all) triggers the server-error
parse path. responseCode===0 now correctly falls through to the reconnect path.
Fix 2 (backend): after res.flushHeaders() the response is already committed
as SSE. The fallback branch that wrote res.status(404).json() was an HTTP/SSE
protocol violation. Replace with an SSE-conformant event:error frame + res.end().
* fix(sse): use onError helper on subscribe failure + add regression tests
Replace silent res.end() with onError('Failed to subscribe to stream')
so the client receives a parseable SSE error event instead of a stream
that closes with no signal. The previous res.end() left the UI stuck
in "submitting" state because no error/abort/final event ever fired.
Also adds two missing test cases for the responseCode guard change:
- responseCode === 0 with raw SSE buffer data must NOT call errorHandler
(transport failure should reconnect, not display garbage)
- responseCode == null with JSON error data MUST call errorHandler
(server-sent error events should still surface to the user)
---------
Co-authored-by: Danny Avila <danny@librechat.ai>
|
||
ethanlaj
|
85894c11c7
|
🧜♂️ fix: Preserve Mermaid foreignObject HTML in Sanitized SVG (#12819)
|
||
|
Danny Avila
|
f69e8e26f8
|
🪟 feat: Render Source-Code Artifacts in the Side Panel (#12854)
* 🪟 feat: Render Source-Code Artifacts in the Side Panel (CODE bucket) PR #12832 wired markdown / mermaid / html / .jsx-tsx tool outputs through the side-panel artifact pipeline but explicitly punted on code files: > Everything else (csv, py, json, xls/docx/pptx, …) keeps PR #12829's > inline behaviour — dedicated viewers will land in follow-ups. This adds the code-file viewer. A `simple_graph.py` (and every other common source file) now opens in the side panel alongside markdown, mermaid, html, and react artifacts instead of falling back to the inline `<pre>` rendering. **Design.** New `CODE: 'application/vnd.code'` bucket reuses the static- markdown sandpack template — `useArtifactProps` pre-wraps the source as a fenced code block (` ```python\n...\n``` `) before handing it to `getMarkdownFiles`. The fence carries a `language-<x>` class through `marked`, so a future highlighter swap-in (e.g. drop `highlight.js` into the markdown template) picks up syntax colors automatically. The `react-ts` (sandpack) template's React boot cost is avoided since source files don't need it. **Single source of truth for languages.** New `CODE_EXTENSION_TO_LANGUAGE` map drives BOTH: - `EXTENSION_TO_TOOL_ARTIFACT_TYPE` routing (presence in this map = code file). Adding a new language is one entry. - The fenced-block language hint (exported as `languageForFilename`). Identifiers follow the GitHub / `highlight.js` convention so the future highlighter pickup is automatic. **Scope.** Programming languages + stylesheets + shell + sql/graphql + build files (Dockerfile/Makefile/HCL). Pure data formats (CSV/TSV/JSON/JSONL/NDJSON/XML/YAML/TOML) and config dotfiles (`.env`/`.ini`/`.conf`/`.cfg`) are intentionally NOT routed in this pass — they're better served by dedicated viewers (CSV table view, etc.) or remain inline. Adding them later is a one-entry change in the map. **JSX/TSX kept on the React (sandpack) bucket.** They're React component sources; the existing live-preview should win over the static CODE bucket. Plain `.js`/`.ts` source goes through CODE. **MIME-type fallback.** The codeapi backend serves `text/x-python`, `text/x-typescript`, etc. as `Content-Type` for source files, so a file whose extension was stripped/renamed upstream still routes to CODE via the MIME map. **Empty-text gate.** CODE joins MARKDOWN/PLAIN_TEXT in the empty-text exception (an empty `.py` is still a Python file). HTML/REACT/MERMAID still require text — their viewers (sandpack/mermaid.js) error on empty input. **Files changed:** - `client/src/utils/artifacts.ts` — `CODE` bucket constant, `CODE_EXTENSION_TO_LANGUAGE` map, exported `isCodeExtension` and `languageForFilename` helpers, extension/MIME routing additions, template + dependencies entries, empty-text gate exception, helper hoisting (extensionOf / baseMime moved up so the language map can reference them). - `client/src/hooks/Artifacts/useArtifactProps.ts` — exported `wrapAsFencedCodeBlock`, CODE branch that wraps the source then routes through `getMarkdownFiles`. **Tests (+22):** - 8 parameterized routing cases (.py, .js, .go, .rs, .css, .sh, .sql, .kt) verify the CODE bucket fires. - Extension wins when MIME is generic octet-stream (Python has no magic bytes; common case). - Regression: jsx/tsx STAY on REACT bucket (no live-preview regression). - Regression: data formats (CSV/JSON/YAML/TOML) and config dotfiles (.env/.ini) do NOT route to CODE. - Empty-text exception for CODE (empty Python file is still a Python file). - `useArtifactProps`: CODE → content.md / static template, fenced-block shape, language hint, unknown-extension fallback to raw extension, no-extension empty hint, index.html via markdown template. - `wrapAsFencedCodeBlock`: language hint, empty hint, single-trailing- newline trim, multi-newline preservation, empty-source emit. 87/87 in artifact-impacted tests; 155/155 across the broader artifact suite. No regressions in pre-existing markdown/mermaid/HTML/REACT/text behavior. * 🛡️ fix: Bare-filename routing + adaptive fence delimiter (codex P2 ×2) Two follow-ups from Codex review on the CODE bucket: 1. **Bare-filename routing for extensionless build files (Codex P2).** `Dockerfile`, `Makefile`, `Gemfile`, `Rakefile`, `Vagrantfile`, `Brewfile` have no `.` in their basename — `extensionOf` returns `''` and the extension map can't match, so they fell through to inline rendering despite being in `CODE_EXTENSION_TO_LANGUAGE`. New `bareNameOf` helper returns the lowercased basename for extensionless filenames (returns `''` for files with a `.` so the extension and bare-name paths don't double-match). Both `detectArtifactTypeFromFile` and `languageForFilename` consult it as a second lookup against the same `CODE_EXTENSION_TO_LANGUAGE` map, so adding a new build file is one entry. Path-aware: takes the basename so `proj/Dockerfile` (path-preserving sanitizer output) still routes correctly. Added the four extra Ruby build-script names while I was here. 2. **Adaptive fence delimiter (Codex P2).** A hardcoded ` ``` ` fence breaks when the source contains a line starting with ` ``` ` — for example, a JS file containing a markdown-shaped template literal: const md = ` ``` hello ``` `; CommonMark closes a fence on a line whose backtick run matches-or- exceeds the opener, so `marked` would close the outer fence at the inner `\`\`\`` and the rest of the file would render as markdown — corrupting the artifact and potentially altering formatting / links outside `<code>`. New `longestLeadingBacktickRun(source)` scans for the longest start-of-line backtick run in the payload. Fence length = `max(3, longest + 1)` — strictly more than any internal run, so `marked` can never close the outer fence early. Only escalates when needed; the common case still uses a triple-backtick fence. Inline backticks (mid-line) don't count — they're not fence delimiters. Only column-zero runs trigger escalation, so e.g. a Python file with ` `inline ``` here` ` keeps the 3-fence. +11 regression tests: - 8 parameterized cases: `Dockerfile`/`Makefile`/`Gemfile`/etc. route to CODE via bare-name fallback (case-insensitive on basename). - Path-aware: `proj/Dockerfile` recognized. - No double-match: `dockerfile.dev` (with extension) returns null. - Unknown extensionless files (`README`, `LICENSE`) stay null. - 4-backtick fence when source has ` ``` ` at start-of-line. - 5-backtick fence when source has ` ```` ` at start-of-line. - 3-backtick fence (default) for ordinary code. - Inline backticks don't escalate. - Source starting with backtick run at offset 0. Plus 6 new `languageForFilename` tests covering bare-name fallback and path-awareness. 108/108 in artifact-impacted tests (was 87, +21 tests). No regressions. * 🛡️ fix: Indented fence detection + basename-scoped extensionOf (codex P2/P3) Two follow-ups from the latest Codex review on the CODE bucket: 1. **Indented backtick runs (Codex P2).** `longestLeadingBacktickRun` was scanning `^(`+)` — column 0 only. CommonMark allows fence closers to be indented up to 3 spaces, so a JS source containing an indented `\`\`\`` (e.g. inside a template literal embedded in a class method) would still terminate our outer fence and the remainder would render as markdown. Updated regex to `^ {0,3}(`+)`. Tabs are not allowed in fence indentation (CommonMark expands them to 4 spaces, which is over the 3-space limit), so spaces alone suffice. Backticks indented 4+ spaces are CommonMark "indented code blocks" — they can't terminate a fence, so we correctly don't escalate for them. 2. **`extensionOf` path-laden output (Codex P3).** `extensionOf` took `lastIndexOf('.')` across the FULL path string, so `pkg.v1/Dockerfile` yielded the nonsensical "extension" `v1/dockerfile`. `languageForFilename` returned that as the language hint (broken `language-v1/dockerfile` class on the fenced block), AND the routing's bare-name fallback couldn't fire because the extension lookup returned non-empty. New `basenameOf` helper strips path separators; `extensionOf` and `bareNameOf` both go through it. After the fix: - `pkg.v1/Dockerfile` → `extensionOf` returns `''` → `bareNameOf` returns `dockerfile` → routes to CODE with correct language. - `pkg.v1/main.go` → `extensionOf` returns `go` → routes correctly. - `pkg.v1/script.py` → `extensionOf` returns `py` → routes correctly. +10 regression tests: - 5 parameterized cases covering 1-3 space indent at fence lengths 3, 4, 5 (escalation kicks in correctly). - 4-space indent does NOT escalate (CommonMark indented-code-block territory; can't close a fence). - `pkg.v1/Dockerfile` and `a.b.c/Makefile` route to CODE + `languageForFilename` returns `dockerfile`/`makefile`. - Dotted-directory files (`pkg.v1/main.go`, `a.b.c/script.py`) still route correctly via the basename-scoped extension parse. 118/118 in artifact-impacted tests (was 108, +10 tests). No regressions. * 🛡️ fix: Comprehensive review polish + MIME-derived language hint (codex P3) Resolves all 8 valid findings from the comprehensive review and the follow-up Codex P3 on the same PR. None are user-visible bugs; the set spans correctness guards, dead-code removal, organization, and test coverage. **Comprehensive review #1 — Remove dead `isCodeExtension` export.** Function was exported with zero callers anywhere in the codebase. **Comprehensive review #2 — Guard the for-loop against silent overwrites.** The `for (ext of CODE_EXTENSION_TO_LANGUAGE)` loop blindly assigned each language extension to the CODE bucket. If a future contributor added `jsx` or `tsx` to the language map (a natural mistake — they ARE source code), the loop would silently overwrite the REACT bucket entries and break the sandpack live-preview with no compile-time or runtime error. Added `if (ext in EXTENSION_TO_TOOL_ARTIFACT_TYPE) continue` so explicit map entries always win. **Comprehensive review #3 — Add `fileToArtifact` end-to-end test for CODE.** Routing was tested via `detectArtifactTypeFromFile`; full Artifact construction (id / type / title / content / messageId / language) for CODE was not. Added 5 new `fileToArtifact` cases. **Comprehensive review #4 — Move pure utilities out of the hook file.** `wrapAsFencedCodeBlock` and `longestLeadingBacktickRun` are pure string transformations with no React dependencies. Moved both to `utils/artifacts.ts`. Test files updated to import from the new location. **Comprehensive review #5 — Correct the MIME-map "mirrors" comment.** Comment claimed the MIME map mirrored `CODE_EXTENSION_TO_LANGUAGE`, but covered ~21 of ~60 entries. Reworded to "best-effort COMMON-CASE list, not an exhaustive mirror" with the rationale (extension routing is primary; MIME is a stripped-filename fallback). **Comprehensive review #6 — Drop `lang ? lang : ''` ternary.** `lang` is typed `string`; the only falsy value is `''`. Removed. (Replaced via the MIME-fallback rewrite of `wrapAsFencedCodeBlock`, where `lang` is now used directly without the ternary.) **Comprehensive review #7 — Avoid double `basenameOf` computation.** `extensionOf(filename)` and `bareNameOf(filename)` both internally called `basenameOf` — when the extension lookup missed, `detectArtifactTypeFromFile` paid for two parses of the same path. Split into private `extensionFromBasename` / `bareNameFromBasename` helpers; the caller computes `basenameOf` once and threads it through. **Comprehensive review #8 — Trim verbose Dockerfile/Makefile comment.** Inline comment block in the language map duplicated `bareNameOf`'s JSDoc. Replaced with a one-line pointer. **Codex P3 — MIME fallback for the CODE language hint.** `detectArtifactTypeFromFile` routes `{ filename: 'noext', type: 'text/x-python' }` to CODE via the MIME bucket map, but then `useArtifactProps` derived the language hint from `artifact.title` ONLY — and `noext` has no extension, so `languageForFilename` returned empty and the fenced block emitted with no `language-` class. The future highlighter swap-in would lose syntax-color metadata for these files. - New `MIME_TO_LANGUAGE` map covering the language MIMEs codeapi actually emits. - `languageForFilename(filename, mime?)` now takes an optional MIME second arg and falls back to it after the extension and bare-name paths. - `fileToArtifact` resolves the language at construction time (using both filename AND `attachment.type`) and stores it on `artifact.language`. The hook reads `artifact.language` directly rather than re-deriving from `title` alone, so the MIME signal survives end-to-end. - Title-derived fallback in the hook covers older callers that don't populate `language`. Tests: +10 cases for the comprehensive review findings (CODE end-to-end via `fileToArtifact`, language storage, non-CODE language un-set). +6 cases for the MIME fallback (`languageForFilename(name, mime)` ordering, MIME parameter stripping, extension/bare-name vs MIME precedence, empty signal). +2 hook tests for `artifact.language` pre-resolved vs title-fallback. 131/131 in directly-impacted files (was 118, +13). 199/199 across the broader artifact suite. No regressions. Pre-existing TypeScript errors in `a11y/`, `Agents/`, `Auth/`, `Mermaid.tsx`, etc. are unrelated to this PR (verified by checking `tsc --noEmit` on origin/dev — same errors). |
||
|
Danny Avila
|
47f65fe39a
|
🪟 feat: Render Code-Execution Text Artifacts as Side-Panel Artifacts (#12832)
* 🪟 feat: Render Code-Execution Text Artifacts as Side-Panel Artifacts Builds on PR #12829 (which populates `text` on code-execution file attachments). When a tool-output file's extension/MIME maps to a viewer we already have, route it through the artifact UI instead of the inline `<pre>`: - text/html, text/htm → existing artifacts side panel (sandpack) - App.jsx / App.tsx → existing artifacts side panel (sandpack) - *.md / *.markdown / *.mdx → existing artifacts side panel (sandpack) - *.mmd / *.mermaid → standalone Mermaid component, inline (no sandpack/react template) The card and the mermaid header both expose a download button so the underlying file is still reachable. Everything else (csv, py, json, xls/docx/pptx, …) keeps PR #12829's inline behaviour — dedicated viewers for csv/docx/xlsx/pptx will land in follow-ups. Backend: `.mmd` and `.mermaid` added to UTF8_TEXT_EXTENSIONS so mermaid sources reach the client with `text` populated. Frontend changes: - `client/src/utils/artifacts.ts` — `TOOL_ARTIFACT_TYPES` constant, `detectArtifactTypeFromFile`, `fileToArtifact` (id is derived from `file_id` so the same artifact across renders dedupes cleanly). - `client/src/components/Chat/Messages/Content/Parts/ToolArtifactCard.tsx` — registers the artifact in `artifactsState`, renders an `ArtifactButton`-style trigger paired with a download button. - `client/src/components/Chat/Messages/Content/Parts/ToolMermaidArtifact.tsx` — wraps the standalone Mermaid component with a filename + download header so the file stays reachable. - `Attachment.tsx` and `LogContent.tsx` — gain panel-artifact and mermaid branches in the routing decision tree, ahead of the existing inline-text fallback. Existing branches untouched. Test coverage: backend extension matrix (mmd/mermaid), frontend predicates (`isPanelArtifact`, `isMermaidArtifact`, `artifactTypeForAttachment`), `fileToArtifact`, and an RTL suite that verifies each type routes to the right component (panel card / mermaid render / inline pre / file chip). * 🩹 fix: Address review on code-artifacts-panel routing - ToolArtifactCard: defer artifact registration to the click handler so rendering a card never side-effects into `artifactsState`. With `artifactsVisibility` defaulting to `true`, eager mount-time registration would surface tool artifacts in the side panel without user intent — now matches ArtifactButton's pattern. Drop the redundant `artifacts` subscription (write-only via useSetRecoilState). - LogContent.tsx: precompute `Artifact`s inside the existing useMemo bucket-sort so each render isn't producing fresh objects. Without this, missing updatedAt/createdAt fields would make `toLastUpdate` return `Date.now()` and churn Recoil state on every parent render. - Attachment.tsx + LogContent.tsx: classify each attachment once via `artifactTypeForAttachment` and branch on the result, instead of calling `isMermaidArtifact` and `isPanelArtifact` back-to-back (each of which internally re-classified). AGENTS.md single-pass rule. - artifacts.ts `detectArtifactTypeFromFile`: strip `;` parameters before the MIME comparison (so `text/html; charset=utf-8` is recognized) and add fallbacks for `application/vnd.react`, `application/vnd.ant.react`, and `application/vnd.mermaid`. - ToolMermaidArtifact: drop the `id` prop entirely when `file_id` is undefined so we never pass an undefined DOM id through to mermaid. - AttachmentGroup: keys derived from `file_id` (not bare index) so add/remove churn doesn't remount stable cards. - Wrappers (PanelArtifact / MermaidArtifact / ToolMermaidArtifact) tightened from `Partial<TAttachment>` to `TAttachment` since the caller always passes a full attachment. - fileToArtifact: drop dead `?? ''` on content (guarded by the preceding type check). - Tests: new click-interaction suite verifying the deferred-registration invariant, click registers + opens panel, and second click toggles closed without losing the registered artifact. * 🧹 chore: Address follow-up review NITs - artifacts.test.ts: regression-pin baseMime() with charset/case variants for text/html, text/markdown, application/vnd.react. - attachmentTypes.ts: drop the now-unused isMermaidArtifact and isPanelArtifact wrappers (the routing collapsed onto a single artifactTypeForAttachment call in the previous commit, so they were only kept alive by their own test). attachmentTypes.test.ts rewritten to exercise artifactTypeForAttachment branches directly. - Attachment.tsx + LogContent.tsx: re-sort the local imports longest-to-shortest per AGENTS.md (~/utils/artifacts is 72 chars and was sitting after a 51-char import). * ✨ feat: Auto-open panel + route txt/docx/odt/pptx through artifacts - artifacts.ts: add `text/plain` to TOOL_ARTIFACT_TYPES so plain-text documents (and the markdown-like ones we don't have rich viewers for yet) can route through the side panel. `useArtifactProps` already dispatches `text/plain` to the markdown-style template, so they render cleanly with no panel-side change. - Extension map gains txt/docx/odt/pptx → text/plain. pptx is wired up speculatively — backend extraction is still deferred, so the routing fires the moment that lands. The MIME map gets the matching office MIME types for symmetry (extension wins, but it's nice to have the fallback when sniffing returns the canonical office MIME). - ToolArtifactCard: register the artifact in `artifactsState` on mount again. With visibility defaulting to `true` and the panel's `useArtifacts` hook auto-selecting the latest artifact, this gives the auto-open behaviour that the legacy streaming artifacts have. Click handler is now just "focus + reveal" (registration already happened); a user who has explicitly closed the panel keeps it closed and uses the click to re-open. - Tests: parameterised row for each new extension; ArtifactRouting invariant flipped from "no register on mount" to "registers on mount so panel can auto-open". Existing TextAttachment test that used `a.txt` switched to `a.csv` since `.txt` now panel-routes. * 🐛 fix: Auto-focus latest tool artifact + self-heal after panel close Two bugs in the previous commit's auto-open behaviour: 1. After closing the side panel, no artifact card could be reopened. `useArtifacts.ts` resets `artifactsState` in its unmount cleanup (line 50), which fires when visibility goes to `false`. The card's mount-only `useEffect` doesn't refire after that wipe, so the subsequent click set `currentArtifactId` to an id that was no longer in `artifactsState`, and `Presentation.tsx` then refused to render the panel because `Object.keys(artifacts).length === 0`. Fix: the registration `useEffect` now has no dependency array, so it self-heals after the wipe (the dedup check keeps it cheap when nothing actually needs writing). 2. Newly-arrived artifacts didn't steal focus from an already-selected one. `useArtifacts`'s fallback auto-select (line 64) only fires when `currentId` is null or no longer in the list — it deliberately protects an existing selection, while the streaming-specific effect that handles legacy focus-stealing is gated on `isSubmitting`. That gate doesn't apply to tool-output artifacts. Fix: a second `useEffect` keyed on `artifact.id` calls `setCurrentArtifactId(artifact.id)` whenever a new card mounts. Cards mount in attachment-array order, so the LAST-mounted card (the newest tool output) wins — matching the legacy "latest auto- opens" UX. Tests: replace the now-stale "no register on mount" assertion with "registers and auto-focuses on mount", flip the toggle test to start from the auto-focused state, and add two regression tests covering the close-then-reopen path and the latest-of-many auto-focus. * ✨ feat: Route pptx through artifact panel with placeholder content Before this commit, pptx files fell through to a plain FileContainer chip even though the extension was wired into the artifact map: backend text extraction is still deferred for pptx, so `attachment.text` came back null/empty and `detectArtifactTypeFromFile`'s strict text check returned null. That meant docx/odt rendered as proper artifact cards while pptx in the same message rendered as a tiny download chip. `detectArtifactTypeFromFile` now allows empty text for the plain-text and markdown buckets, since their viewers (the markdown template) handle empty content gracefully. HTML / React / Mermaid still require real content because sandpack and mermaid.js error on empty input. `fileToArtifact` substitutes a markdown placeholder ("Preview not available yet — click Download to view the file.") when the file routes through the panel without text. The panel renders the placeholder via the markdown template; pptx (and any docx that fails extraction) gets visual parity with its siblings, and the moment backend extraction lands the placeholder is replaced by real content without any frontend change. Tests: split the "no text returns null" assertion into the strict viewers (HTML/React/Mermaid) and the lenient ones (plain-text/markdown); add a fileToArtifact case proving pptx without text gets the placeholder, and another proving real text wins when present. * ✨ feat: Dedup duplicate tool-artifact cards across tool calls + messages Two `ToolArtifactCard` instances for the same file_id (e.g. agent reads back what it just wrote, or the same file is referenced in turns 1 and 5) now collapse to a single chip — the most recent mount wins, the older sibling re-renders to `null`. Implementation: - New `toolArtifactClaim` atomFamily keyed by artifact id. Each card generates a unique component-instance key via `useId()`, claims the slot in a `useLayoutEffect` (synchronous before paint, no flicker), and releases it on unmount only if the claim is still ours. A later card with the same id overwrites the claim → earlier card subscribes via `useRecoilState` and renders `null`. - Family-keyed (per artifact id) so adding/removing a claim for one file never re-renders cards for unrelated files. Addresses the "messages view re-renders frequently" concern: each card subscribes only to its own slice. - `ToolMermaidArtifact` shares the same atom via the new exported `toolArtifactKey()` helper, so the same `.mmd` file can't double- render either. - Latest content always wins for the panel because the eager `setArtifacts` registration is last-write-wins on `artifactsState` by id — independent of which card holds the claim. Updating a file refreshes the panel content even if the chip's visual location doesn't move. Tests: two new cases asserting that duplicate panel and mermaid attachments collapse to a single rendered card. * 🧹 chore: Address comprehensive review on code-artifacts-panel - ToolArtifactCard self-heal now subscribes to a per-id selector (`artifactByIdSelector`) instead of a no-deps `useEffect`. Effect deps are `(artifact, existingEntry, setArtifacts)` so it runs deterministically when the slice transitions to undefined (panel- unmount cleanup) or when artifact content drifts — not on every parent render. Each card subscribes only to its own slice via the selectorFamily, so unrelated state changes don't re-render. - artifacts.ts: localize the empty-content placeholder via a new `fileToArtifact(attachment, options?)` signature. Callers in `Attachment.tsx` (PanelArtifact) and `LogContent.tsx` resolve `com_ui_artifact_preview_pending` from `useLocalize` and thread it in. Default is empty string when no placeholder is supplied. - artifacts.ts: thread `preClassifiedType` through `fileToArtifact` so the routing decision tree's `artifactTypeForAttachment` call is the only classification — previously `fileToArtifact` re-ran `detectArtifactTypeFromFile` after the routing already had the answer. Bucket type updated to `Array<{ attachment, type }>`. - artifacts.ts: drop bare `text/plain` from `MIME_TO_TOOL_ARTIFACT_TYPE`. The extension map handles `.txt` explicitly; routing every unrecognized-extension `text/plain` file (extensionless scripts, `.env`, etc.) through the panel was a wider catch than the PR scope intended. - artifacts.ts: stable `toLastUpdate` fallback of `0` (was `Date.now()`). `useArtifacts` sorts by `lastUpdateTime`, so a fresh timestamp on every call would re-sort entries non-deterministically across renders. - artifacts.ts: drop dead `toolArtifactId = toolArtifactKey` alias. Add `filepath` to the key-derivation fallback chain so two unnamed-and-unidentified files don't collide on the literal `tool-artifact-unknown` key. - ToolArtifactCard import order: package types before local types. - store/artifacts.ts: JSDoc on `toolArtifactClaim` documenting the atomFamily-entries-persist-after-unmount trade-off (entries reset to null on card unmount; total cost is one key + a null per artifact — fine at typical session scale). - Tests: - Updated existing `fileToArtifact` placeholder assertion to use the caller-provided string. - New: panel routing skips re-classification when `preClassifiedType` is provided. - New: bare `text/plain` MIME with unrecognized extension does NOT route through the panel. - New `LogContent.test.tsx` (6 cases) — HTML→panel, mermaid→ inline, CSV→inline `<pre>`, archive→download chip, pptx→ placeholder card, mixed split. - Dedup tests rewritten to use two AttachmentGroups (matching the real per-tool-call render) instead of a same-array duplicate that triggered React's duplicate-key warning. * 🩹 fix: Address codex review + comprehensive review NITs codex (P2): - artifacts.ts: switch placeholder fallback to nullish coalescing. Empty string is now preserved as legitimate content (a 0-byte `.md` or `.txt` is a valid artifact, not "extraction unavailable") — only `null`/`undefined` triggers the deferred-extraction placeholder. - Attachment.tsx: derive React keys via a new `renderKey` helper that combines `file_id` with the array index. Prevents duplicate keys when the same file_id appears twice in one bucket (rare but possible — a tool call writing the same path twice). Without unique keys, React's reconciler could reuse the wrong card instance, undermining the latest-mention dedup. comprehensive review NITs: - Attachment.tsx: hoist `import type { ToolArtifactType }` up into the type-import section per AGENTS.md. - artifacts.ts `fileToArtifact`: defense-in-depth empty-text guard for the `preClassifiedType` path. Mirrors the gate in `detectArtifactTypeFromFile` so a future caller that bypasses classification can't hand sandpack/mermaid an empty buffer. Plain-text and markdown remain tolerated empty. Tests: - New: empty `.md` content passes through unchanged when a placeholder is also supplied. - New: sibling cards with the same file_id in one group render without React key-collision warnings. - Updated existing placeholder test to use `text: null` (the case where the placeholder is actually meant to fire). - Three parameterized cases pinning the new preClassifiedType-with-empty-text safety guard. * 🩹 fix: Address codex P1/P2 review on code-artifacts-panel - P1 (stale artifacts leak across conversations): Add a top-level `useResetArtifactsOnConversationChange` hook in `Presentation.tsx` that wipes `artifactsState` / `currentArtifactId` on every conversation switch, regardless of panel visibility. Without this guard, ToolArtifactCard's self-heal effect would re-register the previous conversation's artifacts after panel close, leaking them into the next conversation's panel on open. - P2 (expiresAt skipped on panel-routed entries): Restore the legacy expiry gate in `LogContent` ahead of panel/mermaid bucket-sort, so expired pptx/html/etc. attachments fall back to the "download expired" message instead of rendering as a clickable artifact card backed by a dead link. Includes regression coverage for both paths. * 🧹 chore: Share renderAttachmentKey across Attachment + LogContent Hoist the per-occurrence React-key helper from `Attachment.tsx` into `attachmentTypes.ts` so `LogContent` can use the same pattern. Apply it to LogContent's panel/mermaid/text/image/nonInline buckets — the prior keys (e.g. `mermaid-${file_id ?? index}`, `file.filepath ?? ...`) would have collided if the same file_id appeared twice in one render, even though that's astronomically rare for a single tool call. Also drops the unused `file_id` field on `MermaidEntry` since the key no longer needs it. * 🩹 fix: Loosen artifacts util input types to match runtime fallbacks `fileToArtifact`, `detectArtifactTypeFromFile`, `toolArtifactKey`, and `toLastUpdate` all read every picked field with a nullish fallback — their inputs were nonetheless typed as required `Pick<TFile, ...>`. That mismatch made every realistic fixture (and several call sites that lack a stable `filepath`) fail typecheck for fields the implementations never strictly need. Wrap the picks in `Partial<>` so the type matches the contract. * 🩹 fix: Gate tool-artifact registration on claim winner When two `ToolArtifactCard` instances mount for the same `artifact.id` with divergent content (a code-execution file overwritten across turns reuses its file_id), both effects subscribe to `existingEntry` through `artifactByIdSelector`. Each card detects the other's write as drift and overwrites it back, ping-ponging `artifactsState` between old and new content and causing render churn / panel flicker. Gate the self-heal registration on `isMyClaim` so only the latest (claim-holding) card writes. The non-winner still subscribes to the slice but short-circuits before calling `setArtifacts`, breaking the loop. Adds a regression test that fails (loop / wrong final content) without the gate. |
||
|
Danny Avila
|
8c073b4400
|
📄 feat: Auto-render Text-Based Code Execution Artifacts Inline (#12829)
* 📄 feat: Auto-render Text-Based Code Execution Artifacts Inline Eagerly extract text content from non-image artifacts produced by code execution tools and render it inline in the message instead of behind a click-to-download file card. Reuses the SkillFiles binary-detection helper and the existing parseDocument dispatcher so docx, xlsx, csv, html, code, and other text-renderable formats land directly under the tool call. PPTX is intentionally classified but not yet extracted — follow-up. * 🌐 chore: Remove unused com_download_expires locale key Removed in en/translation.json so the detect-unused-i18n-keys CI check passes. The only reference was a commented-out localize() call in LogContent.tsx that was deleted in the previous commit. * 🩹 fix: Address PR review on code artifact text extraction - extract.ts: build the temp document path from a randomUUID and pass path.basename(name) as originalname so a malicious artifact name cannot escape os.tmpdir() (P1 traversal flagged by codex/Copilot). - process.js: classify and extract using safeName, not the raw name — defense in depth alongside the temp-path fix. - classify.ts: add a bare-name lookup so extensionless text artifacts (Makefile, Dockerfile, …) classify as utf8-text instead of falling through to other. - Attachment.tsx: wire aria-expanded / aria-controls on the show-all toggle for screen reader support. - LogContent.tsx: restore a download chip (LogLink) on inline-text attachments so users can still pull down the underlying file. - Tests: cover extensionless filenames and the temp-path traversal invariant. * 🩹 fix: Address comprehensive PR review on code artifact extraction - extract.ts: walk back to a UTF-8 code-point boundary before truncating so cuts cannot land mid-multibyte and emit U+FFFD (CJK/emoji concern). truncate() now accepts the original buffer to skip a redundant encode. - extract.ts: add an 8s timeout around parseDocument via Promise.race so a pathological docx/xlsx cannot stall the response path. - process.js: always set `text` (string or null) on the file payload — createFile uses findOneAndUpdate with $set semantics, so omitting the field leaves a stale value behind when an artifact's content changes. - Attachment.tsx: switch the show-all toggle from char-count threshold to a useLayoutEffect ref measurement on scrollHeight, and use overflow-hidden when collapsed (overflow-auto when expanded) so the collapsed box has a single clear interaction model. - Attachment.tsx + LogContent.tsx: lift `isImageAttachment` / `isTextAttachment` into a shared attachmentTypes module. LogContent keeps its looser image check (no width/height required) because the legacy log surface receives attachments without dimensions. - Tests: cover multi-byte boundary, the always-set-text contract on updates, and the new shared predicates. * 🧪 test: Component test for TextAttachment + direct withTimeout coverage - Attachment.tsx: re-order local imports longest-to-shortest per AGENTS.md (attachmentTypes ahead of FileContainer/Image). - extract.ts: export withTimeout so it can be unit-tested directly (it's also used internally — exporting carries no runtime cost). - extract.spec.ts: three small unit tests on withTimeout that cover resolve, propagated rejection, and timeout rejection paths with real timers. - TextAttachment.test.tsx: ten cases for the new React component — text rendering in <pre>, download chip presence/absence, ref-based collapse measurement (with scrollHeight stubbed via prototype), aria-expanded toggle, fall-through to FileAttachment for missing and empty text, and AttachmentGroup routing. * 🩹 fix: Canonicalize document MIME by extension before parseDocument When the classifier puts a file on the document path via its extension (.docx, .xlsx, …) but the buffer sniffer returned a generic value like application/zip or application/octet-stream, we previously forwarded that generic MIME to parseDocument, which dispatches strictly by MIME and silently rejected it — exactly defeating the extension-first classification this PR added. extractDocument now remaps the MIME from the extension (falling back to the original sniffed MIME if the extension is unrecognized, so files that reached the document branch via MIME detection still work). Adds a parameterized test across docx/xlsx/xls/ods/odt against zip/octet sniffs to guard the regression. * 🩹 fix: Reuse existing withTimeout from utils/promise The previous commit's local withTimeout export collided with the already-exported `withTimeout` from `~/utils/promise`, breaking the @librechat/api tsc job (TS2308 ambiguous re-export). Drops the duplicate, imports from `~/utils/promise`, and removes the now-redundant unit tests (the helper has its own coverage in utils/promise.spec.ts). The third argument shifts from a label to the fully-formed timeout error message that the existing helper expects. * 🧹 chore: TextAttachment test polish (NITs) - Use the conventional `import Attachment, { AttachmentGroup }` form rather than `default as Attachment`. - Save the original `scrollHeight` property descriptor and restore it in afterAll, so the prototype patch never leaks past this suite. |
||
|
Danny Avila
|
596f806f60 |
🛡️ fix: Strict Opt-In Skills Activation per Agent (#12823)
* 🛡️ fix: Strict opt-in skills activation per agent
Skills were activating on every agent run that had the capability +
RBAC enabled, regardless of whether the user (ephemeral) or author
(persisted) had opted in. `scopeSkillIds(undefined)` fell through to
"full accessible catalog" whenever `agent.skills` was unset, which is
the default state for any agent created before skills existed and for
every ephemeral agent.
Activation now requires an explicit signal:
- Ephemeral agent → per-conversation skills badge toggle.
- Persisted agent → new `skills_enabled` master switch on the agent
doc, surfaced as a toggle in the Agent Builder skills section.
Enabled + empty/undefined allowlist = full accessible catalog;
enabled + non-empty allowlist = narrow to those ids; disabled (or
undefined) = no skills available, even if an allowlist is set.
Centralised the predicate in `resolveAgentScopedSkillIds` so the
primary-agent path, handoff/discovery, the subagent loop, and both
OpenAI controllers all share one source of truth. Frontend `$`
popover scope mirrors the same logic so the UI never offers skills
the backend would refuse to activate.
* test: mock resolveAgentScopedSkillIds in agent controller specs
* refactor: address review findings on skills opt-in PR
- AgentConfig: associate skills label with toggle via htmlFor for
click/keyboard affordance; simplify Switch handler to Boolean(value).
- skills: mark scopeSkillIds as @internal so runtime callers continue
to route through resolveAgentScopedSkillIds and inherit the activation
predicate (ephemeral toggle, persisted skills_enabled).
* fix(agents): include skills_enabled in agent list projection
Without this field, agents loaded via the list endpoint hydrate into the
client agentsMap with skills_enabled === undefined, causing the `$`
skill popover to hide every skill on a fresh page load even when the
agent was saved with skills_enabled: true.
* fix(skills): fail closed for persisted agents during agentsMap hydration
Returning undefined while the agents map loads let the popover render the
full catalog for a persisted agent before we could read its
skills_enabled flag, so the user could pick a skill the backend would
then refuse for the turn. Match the strict opt-in contract by returning
[] until the map is authoritative.
* refactor(skills): extract skillsHintKey for readability
Replaces the nested ternary in the skills section JSX with a
pre-computed constant so the activation -> hint key mapping reads
top-down.
* refactor(skills): unflatten skillsHintKey to remove nested ternary
|
||
|
Danny Avila
|
523994220d |
🌉 chore: Gate Skills UI by Agent Capability Checks (#12793)
- Updated the Skills component to include a check for `skillsEnabled` from agent capabilities, ensuring skills are only displayed when both permissions and capabilities allow. - Modified the `useHandleKeyUp` hook to prevent triggering skill commands when skills are disabled, enhancing user experience and preventing errors. - Added tests to verify that skills commands do not trigger when skills are disabled, ensuring robust functionality. - Refactored the `useSideNavLinks` hook to incorporate skills capability checks, ensuring consistent access control across the application. |
||
|
Danny Avila
|
d83cb84f59 |
🪆 feat: Subagent configuration in Agent Builder (#12725)
* 🪆 feat: Subagents configuration (isolated-context child agents) Surfaces the new @librechat/agents `SubagentConfig` primitive in the Agent Builder. Subagents let a supervisor delegate a focused subtask to a child graph running in an isolated context window: verbose tool output stays in the child, only a filtered summary returns to the parent. Data model: new `subagents: { enabled, allowSelf, agent_ids }` on Agent, wired through the Zod, Mongoose, and form schemas plus a new `AgentCapabilities.subagents` capability (enabled by default). Backend: `initialize.js` loads explicit subagent configs alongside handoff agents, and drops subagent-only references from the parallel/handoff maps so they don't leak into the supervisor's graph. `run.ts` emits `SubagentConfig[]` on the primary `AgentInputs` — a self-spawn entry when `allowSelf` is enabled plus one entry per configured agent. UI: an "Advanced" panel section with an enable toggle, a self-spawn toggle, and an agent picker (capped at 10). Enabling without adding agents still yields self-spawn; disabling self-spawn with no agents shows a warning. A capability flag gates the whole section. * 🪆 feat: Stream subagent progress to UI (dialog + inline ticker) Pairs with the @librechat/agents SDK change that forwards child-graph events through the parent's handler registry (danny-avila/agents#107): - Self-spawn and explicit subagents can now use event-driven tools, because child `ON_TOOL_EXECUTE` dispatches reach our ToolService via the parent's registered handler. - The same forwarding path wraps the child's run_step / run_step_delta / run_step_completed / message_delta / reasoning_delta dispatches in a new `ON_SUBAGENT_UPDATE` envelope, with start/stop/error bookends. Backend: `callbacks.js` registers an `ON_SUBAGENT_UPDATE` handler that forwards each envelope straight to the SSE stream. Frontend: - `useStepHandler` consumes `ON_SUBAGENT_UPDATE` events and merges them into a per-tool_call Recoil atom (`subagentProgressByToolCallId`). First-seen `subagentRunId` claims the most-recent unclaimed `subagent` tool call in the active response message — a temporal mapping, no SDK wire-format change needed to correlate child runs with parent tool calls. - New `SubagentCall` part component replaces the default `ToolCall` rendering when `toolCall.name === Constants.SUBAGENT`: compact status ticker showing the 3 most recent update labels, clickable to open a dialog with the full activity log + final markdown-rendered result. - Adds `Constants.SUBAGENT`, `StepEvents.ON_SUBAGENT_UPDATE`, and `SubagentUpdateEvent` type in data-provider. Tests: - `packages/api npx jest run-summarization` — 23 pass - `api npx jest initialize` — 16 pass - `npm run build` — clean Dependency note: bumps `@librechat/agents` to `^3.1.67-dev.1` — requires the SDK PR (danny-avila/agents#107) to be merged to dev and published before this PR merges. `ON_SUBAGENT_UPDATE` is absent from dev.0, so the handler registration would be a no-op with the older SDK but would not crash. * 🪆 fix: address Codex review and review audit on subagents Stacks on top of the SDK change in danny-avila/agents#107 (bumped to `^3.1.67-dev.2`). - **P1 (`initialize.js`)**: subagent-only agents were being deleted from both `agentConfigs` AND `agentToolContexts`. The tool-execute handler resolves execution context (agent, tool_resources, skill ACLs) from `agentToolContexts`, so explicit subagents would run without their configured resources and skip action tools. Now only `agentConfigs` is pruned — tool context stays intact. - **P2 (`AgentSubagents.tsx`)**: toggling subagents off set the form field to `undefined`; `removeNullishValues` stripped it from the PATCH, leaving the server copy enabled. Now it persists an explicit `{ enabled: false, ... }` so the update actually clears state. - **Finding 1 (MAJOR)** — `agent_ids` Zod schema gains `.max()` via a new `MAX_SUBAGENTS` export from `data-provider` (shared with the UI cap). Crafted payloads can't trigger hundreds of `processAgent` calls. - **Finding 2 (MAJOR)** — `subagentProgressByToolCallId` atomFamily atoms are now tracked in a ref and reset from `clearStepMaps` via a `useRecoilCallback({ reset })`. No monotonic growth across a session. - **Finding 3 (MAJOR)** — early-arriving `ON_SUBAGENT_UPDATE` events whose parent `tool_call_id` is not yet mapped are now buffered in `pendingSubagentBuffer` (keyed by `subagentRunId`) and replayed in arrival order once correlation completes. Mirrors the existing `pendingDeltaBuffer` pattern. - **Finding 4 (MAJOR)** — switched to deterministic correlation via the new `parentToolCallId` that SDK `3.1.67-dev.2` threads through from `ToolRunnableConfig.toolCall.id`. Temporal fallback now iterates oldest-unclaimed-first (forward), matching tool-call creation order, so concurrent spawns map correctly. - **Finding 6 (MINOR)** — `agent_ids` are deduped on the backend via `new Set(...)` before the load loop. Duplicates no longer produce duplicate `SubagentConfig` entries visible to the LLM. - **Finding 7 (MINOR)** — events array inside each Recoil atom is capped at 200 entries. Long-running subagents no longer replay O(n) spreads on every update; the dialog log still shows the cap window. - **Finding 8 (MINOR)** — documented: subagents are loaded only for the primary agent this release (handoff children get self-spawn but not explicit sub-subagents). In-code comment added so the next maintainer doesn't wonder. - **Finding 9 (NIT)** — removed `{!isSubmitting && null}` dead code and the misleading announce-polite comment in `SubagentCall`. - New `validation.spec.ts` — 9 tests covering the cap on `agent_ids.length` at the subagent schema, agent-create, and agent-update layers. - `run-summarization` — 23 pass, `initialize` — 16 pass, total backend package: 103 pass across touched areas. Findings 5 (component tests) and 10 (micro-allocation) are tracked but deferred; the former needs a Recoil-RenderHook harness that isn't in this PR's scope, and the latter has negligible impact (one `Array.from` per subagent run). * 🧪 test: integration coverage for subagent correlation + backend loading Addresses the follow-up audit on #12725 with real-code tests (no mock handlers, only the existing setMessages/getMessages spies and the standard mongodb-memory-server harness). Six new tests under a dedicated `describe('subagent loading')`: - loads a configured subagent, populates `subagentAgentConfigs`, keeps it out of `agentConfigs` - **P1 regression guard**: drives the real `toolExecuteOptions.loadTools` closure with the subagent id and asserts `loadToolsForExecution` is called with `agent: <subagent>`, `tool_resources`, `actionsEnabled`. If anyone deletes `agentToolContexts` again, this fails. - dedup: three copies of the same id load the agent once - overlap: agent referenced both as handoff target and subagent stays in `agentConfigs` - capability gate: admin disabling `subagents` suppresses loading even when the agent has a config - per-agent disable: `subagents.enabled: false` skips loading entirely Five new tests under `describe('on_subagent_update event')` using a real `RecoilRoot` and a companion `useRecoilCallback` reader so writes from the hook are observable: - deterministic correlation via `parentToolCallId` (happy path with SDK dev.2+) - fallback: oldest-unclaimed tool call wins for concurrent spawns without `parentToolCallId` - early-arrival buffer: updates with no mapping get buffered and replayed once the tool call appears - event cap: 205 updates collapse to 200 retained, oldest dropped - `clearStepMaps` resets tracked atoms back to their null default - F2 — added explicit `// TODO` marker for handoff-subagent-loading extension (matches the comment that referenced it). - F3 — dropped the unnecessary `MAX_SUBAGENTS as MAX_SUBAGENTS_CAP` alias; just import the constant directly. - Bumped `@librechat/agents` to `^3.1.67-dev.3` to pick up the SDK's paired test additions. - `api/server/services/Endpoints/agents/initialize.spec.js` — 22 pass (6 new + 16 existing) - `packages/api/src/agents/validation.spec.ts` + `run-summarization.test.ts` — 103 pass - `client/src/hooks/SSE/__tests__/useStepHandler.spec.ts` — 48 pass (5 new + 43 existing) * 🪆 fix: strip parent run summary + discovered tools from subagent inputs Codex P1 on #12725: `buildSubagentConfigs` reused the shared `buildAgentInput` factory for each explicit child, and that factory always stamps the parent run's `initialSummary` (cross-run conversation summary) and `discoveredTools` (tool names the parent's LLM searched earlier) onto every `AgentInputs` it returns. When subagents were enabled on a conversation that had already been summarized, every child inherited that summary — silently defeating the isolated-context contract and burning extra tokens on unrelated prior chat. Fix in `run.ts`: after `buildAgentInput(child)`, explicitly clear `childInputs.initialSummary` and `childInputs.discoveredTools` before attaching to the `SubagentConfig`. The parent keeps both — that's how the supervisor receives cross-turn context — but the child starts fresh. Paired with danny-avila/agents#107 (bumped to `^3.1.67-dev.4`), which adds the equivalent strip inside `buildChildInputs` to cover the self-spawn path where the SDK clones parent `_sourceInputs` directly and LibreChat never sees the intermediate shape. Belt and suspenders. Regression test (new): - `does NOT leak the parent run initialSummary into an explicit child (Codex P1 regression)` — sets `initialSummary` on the run, enables subagents with an explicit child, asserts the parent still has the summary but `childConfig.agentInputs.initialSummary` is `undefined`. Same for `discoveredTools`. 24 pass. * 🪆 fix: capability gate applies to handoff agents + parallel subagent test ### Codex P2 — handoff agents kept `subagents` after capability disabled The endpoint-level `AgentCapabilities.subagents` gate only cleared `subagents` on `primaryConfig`. Handoff agents loaded into `agentConfigs` retained their persisted `subagents.enabled: true`, and because `run.ts` calls `buildSubagentConfigs` for every agent input, self-spawn would still fire on a handoff target even when the admin had disabled the capability globally. Fix in `initialize.js`: after the subagent loading block, when the capability is off, iterate `agentConfigs.values()` and clear `subagents` + `subagentAgentConfigs` on every loaded config. Regression test: `clears subagents on handoff agents too when capability is disabled (Codex P2 regression)` — seeds a handoff target with its own `subagents.enabled: true`, disables the capability at the endpoint, asserts both primary AND handoff have `subagents` undefined in the client args. 23 init tests pass. ### Parallel subagent correlation — user-requested verification Added `keeps parallel subagent streams independent when events interleave` to `useStepHandler.spec.ts`. Two `subagent` tool calls seeded side by side, 6 interleaved `ON_SUBAGENT_UPDATE` envelopes dispatched (a-start, b-start, a-step, b-step, a-stop, b-step), each carrying its own `parentToolCallId`. Asserts each `tool_call_id`'s Recoil bucket accumulates only its own run's events, statuses reflect each run independently (`call_a` → stop, `call_b` → run_step), no cross-contamination. 49 step-handler tests pass. * 🪆 fix: SubagentCall detects cancelled / errored states (Codex P2) Codex P2 on #12725: the old `running` check only consulted `initialProgress` and the subagent's phase. A user stop, dropped stream, or backend crash before a terminal `stop`/`error` envelope arrived would leave the ticker permanently stuck on "working…". Other *Call components (ToolCall.tsx) already model this via `!isSubmitting && !finished` → cancelled. Mirror that pattern. Re-introduce `isSubmitting` on `SubagentCallProps` (the prop was dropped earlier as 'unused' — that was a bug) and resolve status as a tri-state: - `finished` — initialProgress >= 1, or subagent `stop`/`error` - `cancelled` — `!isSubmitting && !finished` - `running` — neither New locale keys `com_ui_subagent_cancelled` + `com_ui_subagent_errored` swap in the right header text per state. Tests: new `SubagentCall.test.tsx` covers all four states with a real `RecoilRoot` and a `useRecoilCallback` seeder — no mocked store — 5/5 pass. Includes an explicit P2 regression test that simulates the `isSubmitting=false, progress.status='run_step', initialProgress<1` scenario and asserts the cancelled label renders. * 🪆 feat: semantic ticker + aggregated content-part dialog for subagents Two rounds of feedback on #12725: ### Ticker — user-readable lines, not raw event names The old ticker showed \`on_run_step\`, \`on_message_delta\`, etc. — not meaningful to users. Replaced with \`buildSubagentTickerLines\`, a pure helper that walks the \`SubagentUpdateEvent\` stream and emits: - message/reasoning deltas → a single live "Writing: <last 60 chars>" (or "Reasoning: …") line that updates in place as chunks arrive - run_step with tool_calls → "Using calculator(expression=42*58)" for a single call, "Using tool: a, b" for parallel (args dropped when multiple so the line stays short) - run_step_completed → "calculator → 42*58 = 2436" (output truncated to 48 chars; falls back to "Tool X complete" when output is empty) - error → "Error: <message>" - start / stop / run_step_delta → suppressed (too granular / lifecycle-only) Args and output pass through \`summarizeArgs\` / \`summarizeOutput\` which flatten JSON to \`key=value\` pairs and head-truncate long strings so a 200-line tool output never bloats the ticker. ### Dialog — aggregated content parts via leaf renderers \`aggregateSubagentContent\` folds the raw event stream into \`TMessageContentParts[]\` — text/reasoning delta streaks collapse into single \`TEXT\` / \`THINK\` parts, tool calls become \`TOOL_CALL\` parts, and \`run_step\` boundaries correctly break text runs around tool calls. The dialog iterates those parts through a \`SubagentDialogPart\` renderer that delegates to the existing \`Text\`, \`Reasoning\`, and \`ToolCall\` leaf components — the same sub-components \`<Part />\` uses — wrapped in a minimal \`MessageContext\` so reasoning expand state and cursor animation work. Leaf components are used directly rather than importing \`<Part />\` itself to avoid a module cycle (Part → Parts/index → SubagentCall → Part) and to sidestep a hypothetical nested-subagent rendering. ### Tests - \`subagentContent.test.ts\` — 19 pure-function tests covering the aggregator (text concat, reasoning concat, tool call lifecycle, interleaving, phase suppression, late-arriving completions) and the ticker builder (live preview truncation, args/output snippets, parallel-call handling, output truncation, i18n formatter override). - \`SubagentCall.test.tsx\` — 9 component tests: 5 status-resolution (existing) + 2 ticker (semantic text, delta collapse) + 2 dialog (aggregated parts routed to leaf renderers, raw-output fallback). ### Locale keys New: \`com_ui_subagent_ticker_writing\`, \`…_reasoning\`, \`…_error\`, \`…_using\`, \`…_using_with_args\`, \`…_tool_complete\`, \`…_tool_output\`. Preserves i18n at the display layer while the helper stays pure. * chore: drop unused com_ui_subagent_activity_log locale key The dialog no longer renders an "Activity log" section — the new content-parts renderer replaced it. Also tweaks the dialog description copy to match. * 🪆 fix: subagent dialog order, persistence, auto-scroll, width Follow-up pass addressing the four issues observed in real runs against a live subagent-using parent. ### Aggregator ordering (reasoning appearing after text it preceded) Reproducible pattern: LLM emits reasoning → text → tool call in that order, but the dialog rendered text BEFORE reasoning in the content array. Root cause: `aggregateSubagentContent` maintained `currentText` and `currentThink` buffers in parallel and only flushed them at a `run_step` boundary in a fixed (text, think) order, losing the actual arrival order. Fix: when a text chunk arrives, close any open think buffer first (pushes it into the content array right then); symmetric for think → text. Two new regression tests cover the exact reasoning → text → tool_call sequence from the screenshot and the repeated reasoning ↔ text flow across a turn. ### Content persists after completion (markdown not rendering when done) `clearStepMaps` was calling `resetSubagentAtoms()` at stream end, which wiped every `subagentProgressByToolCallId` entry. Once reset, `contentParts.length === 0` and the dialog fell back to rendering the raw `output` string with plain text — hence the literal `##`/`**` in the completed-state screenshot. Stopped resetting; the atoms are bounded per-call (200-event cap) and per-conversation (one per subagent spawn) so growth matches the rest of the conversation state. `resetSubagentAtoms` is kept for a future conversation-switch caller. Also: routed the raw-`output` fallback (older subagent runs recorded before the event forwarder existed) through the same `SubagentDialogPart` → `Text` leaf that content parts use, so its markdown renders the same way. ### Auto-scroll to bottom while running Added a `scrollRef` on the dialog body and a `useEffect` that pins `scrollTop = scrollHeight` while the dialog is open AND the subagent is running. Triggers on `contentParts.length` (new tool calls / part boundaries) and `events.length` (intra-part deltas) so the cursor tracks text streaming. Disabled post-completion so re-opening a finished run doesn't yank to the bottom. ### Wider dialog Went from `max-w-2xl` (42rem / 672px — too cramped on maximized laptop windows) to `w-[min(95vw,64rem)] max-w-[min(95vw,64rem)]`. Narrow on phones, scales up to 64rem on desktop, always leaves a bit of margin from the viewport edge. Bumped `max-h-[65vh]` on the scroll area to give the extra width room to breathe vertically too. ### Tests - `subagentContent.test.ts` — 21 pass (2 new ordering regressions). - `useStepHandler.spec.ts` — 49 pass (1 updated to assert atoms are *preserved* on clearStepMaps). - `SubagentCall.test.tsx` — 9 pass (unchanged; aggregator-level tests cover the ordering). * 🪆 feat: persist subagent_content via SDK createContentAggregator Per-request map of createContentAggregator instances keyed by the parent's tool_call_id. ON_SUBAGENT_UPDATE handler feeds each event into the matching aggregator (phase → GraphEvent mapping); AgentClient harvests contentParts onto the subagent tool_call at message save so the child's reasoning / tool calls / final text survive a page refresh. Reusing the SDK's battle-tested aggregator instead of a bespoke one keeps the persisted shape identical to the parent graph's output and drops ~100 lines of custom aggregation code. * 🪆 fix: incremental subagent aggregation + dialog render parity **Disappearing tool_calls**: the Recoil atom trimmed events to a 200-long rolling window, so verbose subagents could shed the `run_step` that originally created a tool_call part — rebuilding content from the trimmed window then produced only the surviving text/reasoning. Fix: fold each envelope into `contentParts` incrementally in the atom as it arrives (new `foldSubagentEvent` + cursor state). Event trim window now affects only the ticker, never the dialog. **Render parity**: dialog now applies `groupSequentialToolCalls` and renders single parts through `Container` + grouped batches through `ToolCallGroup` — same spacing and "Used N tools" collapsing the main message view uses. **Width**: `min(96vw, 80rem)` — wider on big screens, still responsive. **Labels**: "Subagent: X" is jargon. Named subagents render as `Running "{name}" agent` / `Ran "{name}" agent` (past tense on completion); self-spawns use `Running subtask` / `Ran subtask` since `Running "self" agent` reads badly. * 🪆 polish: subagent dialog parity + agent avatar in header **Labels**: drop "subtask" framing. Self-spawn shows `Running agent` / `Ran agent` (past tense on completion); named subagents stay `Running "X" agent` / `Ran "X" agent`. **Dialog render parity**: stop wrapping every part in `Container`. TEXT keeps its `Container` (gap-3 + `mt-5` sibling margin), THINK and TOOL_CALL render bare so their own wrappers set the full-column width the regular message view gives them — matches the main `<Part>` dispatch. Outer scroll region now uses `px-4 py-3` padding and a `max-w-full flex-grow flex-col gap-0` inner wrapper, mirroring the `MessageParts` container the main conversation uses. **Avatar**: header icon now renders the subagent's configured avatar via `MessageIcon` when `useAgentsMapContext()` has the child agent, falling back to the `Users` SVG (which keeps its running-state pulse). Same icon-left-of-label pattern the tool UI uses. * 🪆 polish: subagent group label, ticker throttle + tail-ellipsis, scroll button **Grouped label**: ToolCallGroup now detects all-subagent batches and labels them "Running N agents" / "Ran N agents" instead of "Used N tools". Mixed batches keep the existing label. The tool-name summary is suppressed for all-subagent groups (every entry dedupes to "subagent", which adds nothing). **Ticker width + tail-ellipsis**: raise the preview cap to 300 chars so wide containers aren't half-empty, and flip the ticker `<li>` to `dir="rtl"` so `text-overflow: ellipsis` clips the *oldest* characters (visually the left edge) — the newest tokens stay pinned to the right regardless of container width. Bidi lays out the Latin text LTR internally, the rtl only affects which side gets the ellipsis. **Throttle**: `useThrottledValue` hook (trailing-edge, 1.2s) smooths the live `Writing: …` preview so tokens no longer strobe past the eye faster than they can be read. Ref-based internals (not `useState`) avoid infinite-update loops when the upstream value is a new-reference each render; `NEGATIVE_INFINITY` sentinel ensures the very first value passes through synchronously so tests and first paint aren't delayed. **Scroll-to-bottom**: dialog tracks `isAtBottom` with a 120px threshold; auto-scroll only engages when the user is already following along, and a persistent jump-to-latest button appears whenever they scroll up — no more fighting the auto-scroll to read back. * 🪆 polish: snappier ticker, prefix-safe labels, agents icon, readable lines **Ticker lines are now incrementally aggregated in the atom** — same pattern as contentParts. The raw-events rolling window is gone; event volume no longer caps what the ticker can display. Verbose subagents that used to drop early tool_call lines out of the window now keep the full 3-line history (using_tool, tool_complete, writing). **Discriminated-union ticker lines** split a constant prefix (e.g. "Writing:") from a tail-truncatable body. The prefix lives in a `shrink-0` span so it never gets clipped when the body overflows; the body uses `dir="rtl"` only on itself — scoped so non-streaming lines (e.g. "Waiting for first update…") can't get their trailing ellipsis flipped by bidi. **Content-aware throttle**: 800ms interval (down from 1200ms), skipped entirely while the live buffer is below 120 chars. Early tokens now appear immediately — no more "Reasoning: I" sitting blank for a full second before the next heartbeat. Once the preview is long enough to fill the container, throttling kicks in at the tighter interval. **Header label** is now a constant verb + optional muted sub-label. Base reads "Running agent" / "Ran agent" / "Cancelled agent" / "Agent errored" for every subagent; named subagents get the configured agent name rendered to the right in secondary text (self-spawns and unresolved names omit it — "Running self agent" is nonsense). **ToolCallGroup** now detects `allSubagents` and swaps `StackedToolIcons` for a single `Users` glyph — otherwise the group header shows a wrench ("tool") icon next to "Ran 5 agents", which reads wrong. * 🪆 feat: delimiter-aware tool labels in ticker + full-width tool lines New shared `parseToolName` helper in `client/src/utils/toolLabels.ts` — single source of truth for splitting `<tool>_mcp_<server>` ids and mapping native tool names (web_search, execute_code, …) to their friendly translation keys. `ToolCallGroup` drops its inline copy and pulls from this helper. Ticker tool lines now use the shared parser + a new `ToolIdentifier` sub-renderer so the live log reads like the main tool UI: - MCP tool → `<server> · <code-badge:tool>` (e.g. "github · `search_code`") - Native → friendly name from `TOOL_FRIENDLY_NAME_KEYS` - Unknown → bare `<code>` badge of the raw id The `using_tool` / `tool_complete` rows now render with a `flex w-full items-baseline gap-1 overflow-hidden` layout matching the writing/reasoning rows — they take the full container width instead of collapsing to content size. Output snippets on `tool_complete` get the same tail-side `dir="rtl"` ellipsis so the newest characters stay flush-right when the container is narrow. Dropped the now-unused template i18n keys (`com_ui_subagent_ticker_using_with_args`, `com_ui_subagent_ticker_tool_complete`, `com_ui_subagent_ticker_tool_output`) in favor of tokens the JSX composes structurally. Only English is touched per the project rule; other locales follow externally. * 🪆 fix: dialog scroll button + auto-scroll during streaming deltas Two race/trigger bugs in the dialog's scroll behavior: **Button never showed**: `addEventListener('scroll', …)` in a `useEffect` ran before Radix's portal had actually committed the scroll container, so `scrollRef.current` was still null — the listener never attached, `isAtBottom` stayed stuck at its initial `true`, and the jump-to-latest button was never rendered. Swap to React's `onScroll` prop on the element itself so the handler wires up as part of DOM commit, not a post-commit effect. **Auto-scroll stalled during text streaming**: the pin-to-bottom effect only re-fired on `contentParts.length` changes. Message/reasoning deltas extend the last TEXT/THINK part's `.text` without changing the array length — so the view would drift up as tokens piled in and never catch back up. Replace the length-dep effect with a `ResizeObserver` on the inner content div; every height change (new part or in-place growth) triggers a scroll-pin when the user is still at the bottom. * 🪆 fix: drop leading ellipsis from ticker body truncatePreview was prepending ... to the tail when the buffer exceeded 300 chars. The component's CSS already produces a left-side ellipsis for overflow via dir=rtl + text-overflow: ellipsis — stacking a data-level ellipsis on top renders a stray dot character right after the Writing: / Reasoning: label (Writing: .Sure!), which looks like a typo to the reader. Data now returns just the last 300 chars when truncating; CSS handles the visual cue whenever the body actually overflows its container. * 🪆 fix: Codex review — subagent isolation + concurrent-safe throttle Three findings from the @codex review pass, all valid: **P1 — buildAgentInput leaks parent discovered-tool state into subagent children.** `buildAgentInput` mutates `agent.toolRegistry` (`overrideDeferLoadingForDiscoveredTools` flips `defer_loading:true→false` on tools the parent previously searched for) and appends those tools' definitions to the returned `toolDefinitions` before the function returns. `buildSubagentConfigs` was clearing the reported `initialSummary` / `discoveredTools` fields on the returned AgentInputs, but that happened post-return — the registry writes and extra tool definitions persisted on the child, silently defeating context isolation and inflating the child's prompt. Fix: `buildAgentInput` now takes an `isSubagent` flag that gates the registry-mutation block and omits `initialSummary` / `discoveredTools` at the source. `buildSubagentConfigs` passes `{ isSubagent: true }` for every explicit child; no post-hoc cleanup needed. **P2 — ToolCallGroup labels a finished subagent group as still running when the child returned no output.** `getToolMeta` computed `hasOutput` as `!!tc.output`, which is `false` for a completed subagent that returned empty text (the UI already has an "empty result" fallback for that case). `allCompleted` would stay `false` and the group header stuck on "Running N agents" forever. Fix: treat `tc.progress === 1` as completion too — progress is the authoritative lifecycle signal, output is just content. **P2 — useThrottledValue schedules `setTimeout` during render.** Discarded renders under Strict Mode / Concurrent rendering would leave orphan timers firing against stale trees. Fix: move `setTimeout` into a `useEffect` keyed on `[value, intervalMs, enabled]`. Render-time still mutates refs (idempotent), but timer scheduling lives post-commit. Cleanup on unmount and on passthrough transitions is preserved. * 🪆 fix: Codex P2 — wipe subagent atoms on conversation switch `clearStepMaps()` intentionally doesn't reset `subagentProgressByToolCallId` so a user can reopen a completed subagent's dialog mid-conversation, but `resetSubagentAtoms` was defined and never exposed / called — so each completed run's aggregated `contentParts` + `tickerState` stayed resident in the `atomFamily` for the whole app session. Unbounded growth across multi-conversation sessions. Expose `resetSubagentAtoms` from `useStepHandler` and fire it from `useEventHandlers` whenever the URL's `conversationId` changes. That's the correct cleanup boundary: historical subagent dialogs rehydrate from persisted `subagent_content` on each `tool_call` at message-save time, so wiping live atoms on switch doesn't lose any viewable history — it just releases per-tool-call state that the old conversation's components no longer subscribe to. * 🪆 fix: Codex round 3 — subagent registry isolation + post-run label Two more valid findings. **P1 — parent-order registry mutation leaks into subagent inputs.** `overrideDeferLoadingForDiscoveredTools` mutates `agent.toolRegistry` in place (the Map *and* the LCTool objects inside it). When an agent appears both as a handoff target (normal graph node) AND an explicit subagent child, a subagent build that ran before the parent's build captures a reference to the same registry — the parent's later mutation leaks through to the child. Fix: for subagent children (`isSubagent`), clone the `toolRegistry` Map and shallow-clone each LCTool inside before returning the inputs. `defer_loading` flips on parent-graph registry mutations can't propagate across the clone boundary. `toolDefinitions` also gets a shallow-copy pass so the same isolation holds for definitions the child carries directly. **P2 — "Running N agents" label stuck after cancel/error.** ToolCallGroup's all-subagent label was gated only on `allCompleted`, which requires every child to have `hasOutput || progress === 1`. A subagent that gets cancelled (stream ends, no `stop` phase, no output) never satisfies that — so even after `isSubmitting` flips false, the header stays on "Running N agents" while each individual card correctly shows "Cancelled agent". Fix: derive a `subagentsDone` flag as `allCompleted || !isSubmitting` and gate the past-tense label on that. Matches the tri-state each SubagentCall card already resolves (finished / cancelled / running). * 🪆 fix: Codex P2 — ACL-check subagents.agent_ids on create/update Codex flagged that `subagents.agent_ids` was accepted as arbitrary strings on the create/update routes while `edges` got a `validateEdgeAgentAccess` pass — so users could save subagent references to agents they can't VIEW. At runtime `initializeClient`'s `processAgent` ACL gate silently drops those, so the persisted configuration and the actual behavior diverged in a way that is difficult to diagnose. Refactor: extract the id-set → unauthorized-ids check into a shared `collectUnauthorizedAgentIds`, wrap it with a dedicated `validateSubagentAccess`, and plumb the same 403-on-failure response the edge path already returns. Applied on both POST /agents and PATCH /agents/:id. * 🪆 fix: Codex round 5 — ACL-disable escape hatch + ticker order Two valid findings. **P1 — can't disable subagents after losing access to a child.** The subagent ACL check ran on every create/update that echoed back the `agent_ids` list, even when the user was explicitly disabling the feature. The UI keeps the list intact when toggling `enabled: false`, so a user who subsequently lost VIEW on any child would be locked in a 403 loop — every edit (including the one that turns subagents off) bounces. Fix: gate the ACL check on `subagents.enabled !== false` at both the POST /agents and PATCH /agents/:id handlers. Empty list stays a no-op. Disabling the feature is always permitted. **P2 — ticker fold merges out-of-order previews across delta switches.** `foldSubagentEventIntoTicker` carried `textLineIdx` / `thinkLineIdx` across a reasoning → text → reasoning transition, so the second reasoning chunk appended to the original reasoning line instead of starting a new chronological one. Fix: close the opposite buffer + cursor when a delta-type switch is detected (same rule the content-parts reducer already applies). Added a regression test. * 🪆 fix: Codex round 6 — preserve mid-stream atoms + honor sequential suppression Two valid findings. **P2 — atom reset fires on initial chat URL assignment.** `useEventHandlers` initialized `lastConversationIdRef` from the URL's current `paramId`, then reset subagent atoms whenever the ref and `paramId` disagreed. For a brand-new conversation the URL stamp goes from `undefined → "abc123"` while the first response is still streaming, which used to drop subagent ticker/content state mid-run and leave dialogs missing earlier updates. Fix: only reset when *both* the old and new IDs are non-null and differ — i.e. a user-initiated switch between two established conversations. The initial assignment passes through without clearing. **P2 — ON_SUBAGENT_UPDATE bypassed `hide_sequential_outputs`.** Every other streaming handler in `callbacks.js` (`ON_RUN_STEP`, `ON_MESSAGE_DELTA`, etc.) gates emission on `checkIfLastAgent` + `metadata?.hide_sequential_outputs`, but the subagent forwarder did an unconditional `emitEvent` — so intermediate agents in a sequential chain were leaking their children's activity to the client even when the chain was configured to suppress intermediates. Fix: accept `metadata` and apply the same `isLastAgent || !hide_sequential_outputs` gate. Aggregation still runs regardless of visibility (persistence + dialog depend on it); only the SSE forward is suppressed. * 🪆 fix: Codex P2 — gate subagent ACL check on endpoint capability `validateSubagentAccess` ran on every create/update where `subagents.enabled !== false`, regardless of the endpoint-level `subagents` capability. When the capability is off at the appConfig level, `initializeClient` already strips the `subagents` block at runtime — so persisted `agent_ids` are inert — but the validation could still 403 on a legacy record whose referenced child is no longer viewable, blocking unrelated edits. Fix: add `isSubagentsCapabilityEnabled(req)` that reads the agents endpoint's capabilities from `req.config` and gate both the create and update ACL checks on it. Capability-off environments can update agents with stale `subagents` data freely; capability-on keeps the full ACL protection. * 🪆 fix: Codex P2 — reset subagent atoms on id→null navigation too Previous guard (both-established) skipped the reset whenever `paramId` became null/undefined, so navigating from an existing chat to a "new chat" route left stale subagent progress resident in the `atomFamily` until the user picked a specific different chat. Swap the both-established check for a one-time flag: skip only the very first `undefined → id` transition (the brand-new-chat URL stamp that happens mid-stream), then reset on any subsequent change — id→id, id→null, null→id-after-reset. If the user started on an established chat the flag is true at mount, so the guard is a no-op and every navigation resets normally. * 🪆 fix: Codex round 9 — subagent persistence gate + handoff children Two valid findings. **P1 — hide_sequential_outputs also gates persistence.** The previous fix gated the SSE forward on `isLastAgent || !hide_sequential_outputs` but still ran the per-tool-call `createContentAggregator` aggregation unconditionally. `finalizeSubagentContent` would then attach the hidden intermediate agent's child reasoning / tool output to the saved message, so a page refresh could reveal activity that was intentionally suppressed live. Move the visibility gate to the top of the handler — hidden agents now skip both aggregation and emission, so "hide_sequential_outputs" is a consistent "don't record" rule for subagent traces. **P2 — handoff agents' explicit subagents were silently dropped.** `initializeClient` only resolved `subagentAgentConfigs` for the primary config, so an agent used via handoff that had its own `subagents.agent_ids` saved in the builder would get self-spawn only; every explicit child was quietly ignored, creating a saved-config / runtime mismatch the user couldn't diagnose. Extract the resolution into a shared `loadSubagentsFor(config)` helper and invoke it for the primary and every handoff agent in `agentConfigs`. The `edgeAgentIds` precomputation stays outside the helper (it's loop-invariant). Capability-off shortcuts return empty early so the existing strip-on-capability-off path still holds. * 🪆 fix: Codex P2 — recursive subagent build for multi-level delegation Previously only the outer `agents[]` loop attached `subagentConfigs` to its inputs, so a child used as a subagent (invoked via the `subagent` tool) lost every explicit spawn target of its own. A user-valid configuration like A → B → C would only run the top layer; B could never actually delegate to C from inside A's run. Recursively build `subagentConfigs` for each child inside `buildSubagentConfigs`, passing the child's freshly-constructed `childInputs` down so its own `subagents.enabled` children get resolved too. Added cycle protection via an `ancestors` Set — a configuration like A → B → A is safely cut off at the second encounter of A rather than recursing forever (the existing `child.id === agent.id` guard already prevents the direct self-loop). * 🪆 fix: Codex P2 — reset subagent atoms on useEventHandlers unmount The effect that resets subagent atoms only fired on `paramId` change, so unmounting the chat container (route change away from /c) never flushed the atoms. `knownSubagentAtomKeys` lives in a ref inside `useStepHandler` — once the hook unmounts the ref is gone, so a subsequent remount can't clean atoms it never registered. Added a second `useEffect` that only runs cleanup on unmount (empty deps aside from the stable `resetSubagentAtoms` callback). Keeps `atomFamily` bounded across full route teardowns too. * 🪆 fix: Codex round 13 — cyclic subagent guard + prefer persisted Two valid findings. **P1 — cyclic subagent ref reloads the primary.** A configuration like `A ↔ B` (B lists A as its own subagent) would send `loadSubagentsFor` down a path that couldn't find A in `agentConfigs` (the primary isn't stored there), so it called `processAgent(A)` a second time. That inserts a fresh config for the primary id, which downstream duplicates in `[primary, ...agentConfigs.values()]` and can replace the primary's tool context with the reloaded copy. Fix: short-circuit when a subagent ref points back at `primaryConfig.id` — reuse the already-loaded primary config. Primary is always an edge id so no pruning bookkeeping needed. **P2 — live atom preferred over canonical persisted trace.** The dialog picked `progress.contentParts` ahead of `persistedContent`, but the Recoil bucket is best-effort — after a disconnect/reconnect it can be stale or partial. The server's `subagent_content` on the `tool_call` is the canonical record refreshed on sync. Preferring live could hide completed tool/reasoning history that was actually persisted. Fix: flip the preference order. Persisted wins when it's non-empty; live covers the mid-stream window (before the parent message saves, persisted is empty) and the older-runs fallback. Updated the test that enforced the old order to lock the new semantics in (separate mid-stream live-fallback assertion kept). * 🪆 fix: Codex P2 — subagent atom reset rule simplified to 'leaving established id' The `hasEstablishedConversationRef` + check for initial undefined→id covered the first navigation but missed the equivalent mid-stream URL stamp when a user goes from an existing chat to a new chat and sends a message there (`id → null → newId`). The null → newId transition was still hitting the reset branch and wiping the in-flight subagent ticker/content for that first turn. Simpler rule: only reset when the PREVIOUS paramId is an established id. Every transition AWAY from an established chat clears (id→id2, id→null, id→undefined); every transition FROM null/undefined passes through (initial mount, new-chat URL stamp mid-stream). Drop the `hasEstablishedConversationRef` machinery in favor of that single condition. * 🪆 fix: Codex P2 — match runtime's strict subagent enable check in ACL Runtime (`initializeClient` + `run.ts`) treats `subagents?.enabled` as a truthy predicate — `undefined`, `null`, missing, and `false` all short-circuit. The ACL gate was using `!== false` which accepted `undefined` / missing as "enabled" and could 403 a payload whose subagent tool would be inert at runtime. Swap both create and update to `enabled === true`. Only a strictly- enabled payload triggers the ACL check; the disable path (`false`) still passes through so a user who lost VIEW on a child can still save the disable edit. * 🪆 fix: Codex P2 — reject missing subagent references with 400 `validateSubagentAccess` collapsed through `collectUnauthorizedAgentIds`, which returns an empty list for ids with no DB record — so typos and references to deleted agents passed validation silently, and `initializeClient` later dropped them at runtime. Saved config would then list spawn targets that the backend never honored, a hard-to- diagnose drift. Refactor the helper into `classifyAgentReferences(ids, …)` which returns `{ missing, unauthorized }` separately. `validateEdgeAgentAccess` keeps its old semantics (missing is intentional — a self-referential `from` names the agent being created). `validateSubagentReferences` surfaces both buckets so the create/update handlers can 400 on missing and 403 on unauthorized with distinct error messages and `agent_ids` lists. * 🪆 polish: tighten subagent dialog grid gap to gap-2 OGDialogContent's grid default is `gap-4`, which renders the title, description, and scroll area as three visually separated panels. Drop to `gap-2` so they read as one block. * 🪆 polish: swap Subagents above Handoffs in Advanced panel Subagents is the more common knob users reach for, so show it first. Handoffs keep the same Controller wiring, just move below. |
||
|
Danny Avila
|
35bf04b26c |
🧰 refactor: Unify code-execution tools (#12767)
* 🛠️ feat: Add registerCodeExecutionTools helper Idempotently registers `bash_tool` + `read_file` in the run's tool registry and tool-definition list via a registry `.has()` dedupe. Sets up the single code-execution tool path shared by: - `initializeAgent` (when an agent has `execute_code` in its tools and the capability is enabled for the run) - `injectSkillCatalog` (when skills are active; unconditional read_file, bash_tool follows `codeEnvAvailable`) Both callers reach the helper in the same initialization sequence, so the second call becomes a no-op and exactly one copy of each tool reaches the LLM — no more double registration for agents that combine `execute_code` capability with active skills. Unit-tested on a fresh run, idempotence (second call, overlap with prior tooldefs, partial overlap), and the no-registry variant. * 🔀 refactor: Route injectSkillCatalog bash_tool + read_file through registerCodeExecutionTools The `skill` tool is still registered inline (it's skill-path-specific), but `bash_tool` + `read_file` now flow through the shared idempotent helper so a prior registration from the execute_code path doesn't produce a duplicate copy later in the same run. Behavior preserved: - `read_file` always registers when any active skill is in scope — manually-primed `disable-model-invocation: true` skills still need it to load `references/*` from storage. - `bash_tool` follows `codeEnvAvailable` exactly as before. Adds a test pinning the cross-call dedupe: when `injectSkillCatalog` runs AFTER `registerCodeExecutionTools` has already seeded the registry + tool definitions with bash_tool/read_file, the resulting `toolDefinitions` still contains exactly one copy of each. * 🪄 feat: Expand `execute_code` tool name into bash_tool + read_file at initialize-time When an agent's `tools` include `execute_code` and the `execute_code` capability is enabled for the run, `initializeAgent` now registers `bash_tool` + `read_file` via `registerCodeExecutionTools` before `injectSkillCatalog`. The legacy `execute_code` tool definition is no longer handed to the LLM — `execute_code` remains on the agent document as a capability-trigger marker, but the runtime expands it into the skill-flavored tool pair. Call ordering matters: the `execute_code` registration runs BEFORE `injectSkillCatalog`, so the skill path's own `registerCodeExecutionTools` call inside `injectSkillCatalog` becomes a no-op via the registry's `.has()` check. Exactly one copy of each tool reaches the LLM whether the agent has: - only `execute_code` (legacy path) - only skills - both No data migration needed — `agent.tools: ['execute_code']` stays in the DB unchanged; the expansion is a runtime operation. Three tests cover the matrix: execute_code + capability on → bash_tool + read_file registered; execute_code + capability off → neither registered; no execute_code + capability on → neither registered. * 🗑️ refactor: Drop CodeExecutionToolDefinition from the builtin registry Removes the legacy `execute_code` entry from `agentToolDefinitions` and the corresponding import. With the initialize-time expansion in place, nothing consults `getToolDefinition('execute_code')` for a tool schema any more — the capability gate still filters on the string `execute_code`, but the actual tool definitions the LLM sees come from `registerCodeExecutionTools` (i.e. `bash_tool` + `read_file`). `loadToolDefinitions` in `packages/api/src/tools/definitions.ts` silently drops `execute_code` when it no longer resolves in the registry — that's the expected path and is now covered by an updated test. No caller of `getToolDefinition('execute_code')` expects a non-undefined result after this change. * 🔌 refactor: Read CODE_API_KEY from env for primeCodeFiles + PTC Finishes the Phase 4 server-env-keyed rollout on the two remaining `loadAuthValues({ authFields: [EnvVar.CODE_API_KEY] })` sites in `ToolService.js`: - `primeCodeFiles` (user-attached file priming on execute_code agents) - Programmatic Tool Calling (`createProgrammaticToolCallingTool`) Both now read `process.env[EnvVar.CODE_API_KEY]` directly, matching `bash_tool`'s pattern. The per-user plugin-auth path is no longer consulted for code-env credentials anywhere in the hot path — the agents library owns the actual tool-call execution and also reads the env var internally. Priming still fires for existing user-file workflows so the legacy `toolContextMap[execute_code]` hint ("files available at /mnt/data/...") stays in the prompt; only the key lookup changed. * 🔧 fix: Type the pre-seeded dedupe-test tools as LCTool CI TypeScript type checks caught `{ parameters: {} }` in the new cross-call dedupe test: `LCTool.parameters` is a `JsonSchemaType`, not `{}`. Use `{ type: 'object', properties: {} }` and type the local registry Map through the parameter-derived shape so the pre-seeded values match what `toolRegistry.set` expects. * 🛡️ fix: Run execute_code expansion before GOOGLE_TOOL_CONFLICT gate Codex review caught a latent regression: the original Phase 8 placement ran `registerCodeExecutionTools` after `hasAgentTools` was computed, so an execute-code-only agent on Google/Vertex with provider-specific `options.tools` populated would no longer trip `GOOGLE_TOOL_CONFLICT` — the legacy `CodeExecutionToolDefinition` used to populate `toolDefinitions` before the guard, but after dropping it from the registry, `toolDefinitions` stayed empty until my expansion ran downstream of the guard. Mixed provider + agent tools would silently flow through to the LLM. Fix moves the `execute_code` expansion to BEFORE `hasAgentTools` computation. `bash_tool` + `read_file` now contribute to the check the same way the legacy `execute_code` def did. Covered by a new test that pins the Google+execute_code+provider-tools scenario — the `rejects.toThrow(/google_tool_conflict/)` path would have silently passed on the prior placement. * 🔗 fix: Thread codeEnvAvailable through handoff sub-agents Round-2 codex review caught the other half of the execute_code expansion gap: `discoverConnectedAgents` omitted `codeEnvAvailable` from its forwarded `initializeAgent` params, so handoff sub-agents with `agent.tools: ['execute_code']` lost the `bash_tool` + `read_file` registration (pre-Phase 8 the legacy `CodeExecutionToolDefinition` would have landed in their `toolDefinitions` via the registry). - Add `codeEnvAvailable?` to `DiscoverConnectedAgentsParams` and forward it verbatim on every sub-agent `initializeAgent` call. - Update the three JS call sites that construct the primary's `codeEnvAvailable` (`services/Endpoints/agents/initialize.js`, `controllers/agents/openai.js`, `controllers/agents/responses.js`) to pass the same flag into `discoverConnectedAgents` — one authoritative source per request. - Two regression tests in `discovery.spec.ts` pin the true/false passthrough so a future refactor that drops the param-forwarding surfaces immediately. Left intentionally unchanged: `packages/api/src/agents/openai/service.ts` (public API helper with no in-repo caller). External consumers of `createAgentChatCompletion` who want code execution should pass a `codeEnvAvailable`-aware `initializeAgent` via `deps` — documenting the full public-API surface is out of scope for this Phase 8 PR. * 🔗 fix: Thread codeEnvAvailable through addedConvo + memory-agent paths Round-3 codex review caught the last two production `initializeAgent` callers missing the Phase-8 capability flag: - `api/server/services/Endpoints/agents/addedConvo.js` (multi-convo parallel agent execution). Added `codeEnvAvailable` to `processAddedConvo`'s destructured params and forwarded it into the per-added-agent `initializeAgent` call. Caller in `api/server/services/Endpoints/agents/initialize.js` passes the same `codeEnvAvailable` it computed for the primary. - `api/server/controllers/agents/client.js` (`useMemory` — memory extraction agent). Computes its own `codeEnvAvailable` from `appConfig?.endpoints?.[EModelEndpoint.agents]?.capabilities` and forwards into `initializeAgent`. Memory agents rarely list `execute_code`, but if one does, pre-Phase 8 they got the legacy `execute_code` tool registered unconditionally — the passthrough restores parity. With this, every production caller of `initializeAgent` explicitly resolves the capability: main chat flow (primary + handoff), OpenAI chat completions (primary + handoff), Responses API (primary + handoff), added convo parallel agents, and memory agents. The one remaining caller, `packages/api/src/agents/openai/service.ts::createAgentChatCompletion`, is a public API helper with no in-repo consumer (external callers must pass a capability-aware `initializeAgent` via `deps`). * 🪤 fix: Remove duplicate appConfig declaration causing TDZ ReferenceError The Responses API controller had TWO `const appConfig = req.config;` bindings inside `createResponse`: one at the top of the function (added by the Phase 4 `bash_tool` decouple) and one inside the try block (added by the polish PR #12760). Because `const` is block-scoped with a temporal dead zone, the inner redeclaration put `appConfig` in TDZ for the entire try block, so any earlier reference inside the try — notably `appConfig?.endpoints?.[EModelEndpoint.agents]?.allowedProviders` at line 348 — threw `ReferenceError: Cannot access 'appConfig' before initialization`. The error was silently swallowed by the outer try/catch, leaving `recordCollectedUsage` unreached and the six `responses.unit.spec.js` token-usage tests failing. Removing the inner redeclaration fixes the six failing tests (verified: 11/11 pass locally post-fix, 0 regressions elsewhere). The outer function-scoped binding already provides `appConfig` to every downstream reference. * 🔗 fix: Thread codeEnvAvailable through the OpenAI chat-completion public API Round-4 codex review (legitimate on the type-safety angle, even though the runtime concern was already covered): the `createAgentChatCompletion` helper defines its own narrower `InitializeAgentParams` interface locally, and the type was missing `codeEnvAvailable`. External consumers who supply a capability-aware `deps.initializeAgent` couldn't route `codeEnvAvailable` through without a type-cast workaround. - Widen the local `InitializeAgentParams` interface to include `codeEnvAvailable?: boolean` (matches the real `packages/api/src/agents/initialize.ts` type). - Derive `codeEnvAvailable` inside `createAgentChatCompletion` from `deps.appConfig?.endpoints?.agents?.capabilities` (the same source the in-repo controllers use) and forward to `deps.initializeAgent`. Uses a string literal `'execute_code'` lookup so this file stays free of a `librechat-data-provider` import — keeping the dependency surface of the public helper minimal. With this, external consumers of `createAgentChatCompletion` who pass `appConfig` with the agents capabilities get `bash_tool` + `read_file` registration automatically; consumers who don't pass `appConfig` retain the existing "explicit opt-in" semantics (the flag stays `undefined`, expansion is skipped). * 🧹 chore: Review-driven polish — observability log, JSDoc DRY, test gaps, no-op allocation Addresses the comprehensive review of PR #12767: - **Finding #1** (MINOR, observability): `initializeAgent` now emits a debug log when an agent lists `execute_code` in its tools but the runtime gate is off (`params.codeEnvAvailable` !== true). The event-driven `loadToolDefinitionsWrapper` path doesn't log capability-disabled warnings, so without this the tool silently vanishes from the LLM's definitions with zero trace. Operators debugging "why isn't code interpreter working?" now get a signal at the initialize layer. - **Finding #5** (NIT, allocation): `registerCodeExecutionTools` now returns the input `toolDefinitions` array by reference on the no-op path (both tools already registered by a prior caller in the same run) instead of allocating a fresh spread array every time. The common dual-call scenario — `initializeAgent` then `injectSkillCatalog` — saves one O(n) copy per request. - **Finding #4** (NIT, DRY): Collapsed the duplicated 6-line JSDoc comment in `openai.js`, `responses.js`, and `addedConvo.js` into either a one-line `@see DiscoverConnectedAgentsParams.codeEnvAvailable` pointer (the two JS call sites) or a compact 3-line block referring back to the canonical source (addedConvo's @param). - **Finding #2** (MINOR, test gap): Added `api/server/services/Endpoints/agents/addedConvo.spec.js` with three cases covering `codeEnvAvailable=true`, `codeEnvAvailable=false`, and omitted (undefined) passthrough. A future refactor that drops the param from destructuring now surfaces here instead of silently regressing multi-convo parallel agents with `execute_code`. - **Finding #3** (MINOR, test gap): Added `api/server/controllers/agents/__tests__/client.memory.spec.js` pinning the capability-flag derivation that `AgentClient::useMemory` uses — six cases covering present/absent/null/undefined config shapes plus an enum-literal pin (`'execute_code'` / `'agents'`). Catches enum renames or config-path shifts that would otherwise silently strip `bash_tool` + `read_file` from memory agents. Finding #7 (jest.mock scoping, confidence 40) left as-is: the reviewer's own risk assessment noted `buildToolSet` doesn't touch the mocked exports, and restructuring a file-level `jest.mock` to `jest.doMock` + dynamic `import()` introduces more complexity than the speculative risk justifies. The existing mock is scoped to the test file and contains the same stubs the adjacent `skills.test.ts` already uses. Finding #6 (PR description commit count) addressed out-of-band via PR description update. All existing tests pass, typecheck clean, lint clean across touched files. New tests: 9 cases across 2 new spec files. * 🧽 refactor: Replace hardcoded 'execute_code' string with AgentCapabilities enum in service.ts Follow-up review (conf 55) caught that `openai/service.ts`'s Phase 8 `codeEnvAvailable` derivation used the literal `'execute_code'` while every in-repo controller uses `AgentCapabilities.execute_code` from `librechat-data-provider`. The file deliberately uses local type interfaces to keep the public API helper's type surface small, but that pattern was never a ban on single-value imports from the data provider — `packages/api` already depends on it. Importing the enum value means a future rename of `AgentCapabilities.execute_code` propagates to this file automatically, matching the in-repo controllers' behavior. Other follow-up findings left as-is per the reviewer's own verdict: - #2 (memory spec mirrors the production expression rather than calling `AgentClient::useMemory` directly): reviewer flagged as "not blocking" / "design-philosophy observation." The test file's JSDoc already explicitly documents the tradeoff and pins the enum literals to catch the most likely drift vector. Standing up `AgentClient` + all its mocks for a one-line regression guard is disproportionate. - #3 (`addedConvo.spec.js` mock signature vs. underlying `loadAddedAgent` arity): reviewer's own confidence 25 noted the mock matches the wrapper's actual call pattern in the production file. Not a real gap. - #4 was self-retracted as a false alarm. * 🗑️ refactor: Fully deprecate CODE_API_KEY — remove all LibreChat-side references The code-execution sandbox no longer authenticates via a per-run `CODE_API_KEY` (frontend or backend). Auth moved server-side into the agents library / sandbox service, so LibreChat drops every reference: **Backend plumbing:** - `api/server/services/Files/Code/crud.js`: `getCodeOutputDownloadStream`, `uploadCodeEnvFile`, `batchUploadCodeEnvFiles` no longer accept `apiKey` or send the `X-API-Key` header. - `api/server/services/Files/Code/process.js`: `processCodeOutput`, `getSessionInfo`, `primeFiles` drop the `apiKey` param throughout. - `api/server/services/ToolService.js`: stop reading `process.env[EnvVar.CODE_API_KEY]` for `primeCodeFiles` and PTC; the agents library handles auth internally. Remove the now-dead `loadAuthValues` + `EnvVar` imports. Drop the misleading "LIBRECHAT_CODE_API_KEY" hint from the bash_tool error log. - `api/server/services/Files/process.js`: remove the `loadAuthValues` call around `uploadCodeEnvFile`. - `api/server/routes/files/files.js`: code-env file download no longer fetches a per-user key. - `api/server/controllers/tools.js`: `execute_code` is no longer a tool that needs verifyToolAuth with `[EnvVar.CODE_API_KEY]` — the endpoint always reports system-authenticated so the client skips the key-entry dialog. `processCodeOutput` called without `apiKey`. - `api/server/controllers/agents/callbacks.js`: `processCodeOutput` invoked without the loadAuthValues round trip, for both LegacyHandler and Responses-API handlers. - `api/app/clients/tools/util/handleTools.js`: `createCodeExecutionTool` called with just `user_id` + files. **packages/api:** - `packages/api/src/agents/skillFiles.ts`: `PrimeSkillFilesParams`, `PrimeInvokedSkillsDeps`, `primeSkillFiles`, `primeInvokedSkills` all drop the `apiKey` param; the gate is purely `codeEnvAvailable`. - `packages/api/src/agents/handlers.ts`: `handleSkillToolCall` drops the `process.env[EnvVar.CODE_API_KEY]` read; skill-file priming is now gated solely on `codeEnvAvailable`. `ToolExecuteOptions` signatures drop apiKey from `batchUploadCodeEnvFiles` and `getSessionInfo`. - `packages/api/src/agents/skillConfigurable.ts`: JSDoc no longer references the env var. - `packages/api/src/tools/classification.ts`: PTC creation no longer gated on `loadAuthValues`; `buildToolClassification` drops the `loadAuthValues` dep entirely (no LibreChat-side callers need it for this path anymore). - `packages/api/src/tools/definitions.ts`: `LoadToolDefinitionsDeps` drops the `loadAuthValues` field. **Frontend:** - Delete `client/src/hooks/Plugins/useAuthCodeTool.ts`, `useCodeApiKeyForm.ts`, and `client/src/components/SidePanel/Agents/Code/ApiKeyDialog.tsx` — the install/revoke dialogs for CODE_API_KEY are fully dead. - `BadgeRowContext.tsx`: drop `codeApiKeyForm` from the context type and provider. `codeInterpreter` toggle treated as always authenticated (sandbox auth is server-side). - `ToolsDropdown.tsx`, `ToolDialogs.tsx`, `CodeInterpreter.tsx`, `RunCode.tsx`, `SidePanel/Agents/Code/Action.tsx` +`Form.tsx`: all API-key dialog trigger refs, "Configure code interpreter" gear buttons, and auth-verification plumbing removed. The "Code Interpreter" toggle is now a plain `AgentCapabilities.execute_code` checkbox — no key-entry gate. - `client/src/locales/en/translation.json`: drop the three `com_ui_librechat_code_api*` keys and `com_ui_add_code_interpreter_api_key`. Other locales are externally automated per CLAUDE.md. **Config:** - `.env.example`: remove the `# LIBRECHAT_CODE_API_KEY=your-key` section and its header. **Tests:** - `crud.spec.js`: assertions flipped to pin "no X-API-Key header" and "no apiKey param". - `skillFiles.spec.ts`: removed env-var save/restore; tests now pin that the batch-upload path is gated solely on `codeEnvAvailable` and that no apiKey is threaded through. - `handlers.spec.ts`: same — just the `codeEnvAvailable` gate pins remain. - `classification.spec.ts`: remove the two tests that asserted `loadAuthValues` was (not) called for PTC. - `definitions.spec.ts`: drop every `loadAuthValues: mockLoadAuthValues` entry from the deps shape. - `process.spec.js`: strip the mock of `EnvVar.CODE_API_KEY`. **Comment hygiene:** - `tools.ts`, `initialize.ts`, `registry/definitions.ts`: shortened stale comment references to "legacy `execute_code` tool" without naming the retired env var. Tests verified: 678 packages/api tests pass, 836 backend api tests pass. Typecheck clean, lint clean. Only remaining CODE_API_KEY mentions in the code are two regression-guard assertions: - `crud.spec.js`: pins "no X-API-Key header" stays absent. - `skillConfigurable.spec.ts`: pins `configurable` never grows a `codeApiKey` field. * 🧹 chore: Remove the last two CODE_API_KEY name mentions in LibreChat Follow-up to the prior full deprecation commit: two tests still named the retired identifier in their regression-guard assertions. - `packages/api/src/agents/skillConfigurable.spec.ts`: drop the "does not inject a codeApiKey key" test. The `codeApiKey` field is gone from the production configurable shape, so an absence-assertion naming it re-introduces the retired identifier in code. - `api/server/services/Files/Code/crud.spec.js`: rename the "without an X-API-Key header" case back to "should request stream response from the correct URL" and drop the `expect(headers).not.toHaveProperty('X-API-Key')` assertion. The surrounding request-shape checks (URL, timeout, responseType) still pin the behavior; the explicit header-absence line was named-after the deprecated contract. Result: `grep -rn "CODE_API_KEY\|codeApiKey\|LIBRECHAT_CODE_API_KEY"` against the LibreChat source tree returns zero hits. The only remaining `X-API-Key` strings in this repo are on unrelated OpenAPI Action + MCP server auth configurations, where the string is user-facing config, not a LibreChat-owned identifier. Tests: 677 packages/api pass (2 pre-existing summarization e2e failures unrelated); 126 api-workspace controller/service tests pass. Typecheck and lint clean. * 🎯 fix: Narrow codeEnvAvailable to per-agent (admin cap AND agent.tools) Before this commit, `codeEnvAvailable` was computed in the three JS controllers as the admin-level capability flag only (`enabledCapabilities.has(AgentCapabilities.execute_code)`) and passed through `initializeAgent` → `injectSkillCatalog` / `primeInvokedSkills` / `enrichWithSkillConfigurable` unchanged. A skills-only agent whose `tools` array didn't include `execute_code` still got `bash_tool` registered (via `injectSkillCatalog`) and skill files re-primed to the sandbox on every turn — wrong, because the agent never opted in to code execution. **Fix:** `initializeAgent` now computes the per-agent effective value once as `params.codeEnvAvailable === true && agent.tools.includes(Tools.execute_code)`, reuses the same boolean for: 1. The `execute_code` → `bash_tool + read_file` expansion gate (previously already consulted `agent.tools`; now shares the single `effectiveCodeEnvAvailable` binding). 2. The `injectSkillCatalog` call (previously got the raw admin flag). 3. The returned `InitializedAgent.codeEnvAvailable` field (new, typed as required boolean). **Controllers (initialize.js, openai.js, responses.js):** store `primaryConfig.codeEnvAvailable` in `agentToolContexts.set(primaryId, ...)`, capture `config.codeEnvAvailable` in every handoff `onAgentInitialized` callback, and read it from the per-agent ctx inside the `toolExecuteOptions.loadTools` runtime closure. The hoisted `const codeEnvAvailable = enabledCapabilities.has(...)` locals in the two OpenAI-compat controllers are gone — they were shadowing the narrowed per-agent value. **primeInvokedSkills:** `handlePrimeInvokedSkills` in `services/Endpoints/agents/initialize.js` now uses `primaryConfig.codeEnvAvailable` (per-agent, narrowed) instead of the raw admin flag. A skills-only primary agent won't re-prime historical skill files to the sandbox even when the admin enabled the capability globally. **Efficiency:** one extra `&&` in `initializeAgent`. No runtime hot-path cost — the `includes()` scan on `agent.tools` was already happening for the `execute_code` expansion gate; it's now just bound to a local. Tool execution closures read `ctx.codeEnvAvailable === true` (property access + strict equality, O(1)). **Ephemeral-agent note:** per-agent narrowing is authoritative for both persisted and ephemeral flows. The ephemeral toggle (`ephemeralAgent.execute_code`) is reconciled into `agent.tools` upstream in `packages/api/src/agents/added.ts`, so `agent.tools.includes('execute_code')` is the single source of truth by the time `initializeAgent` runs. **Tests:** two new regression tests pin the narrowing contract: - `initialize.test.ts` — four-quadrant matrix on `InitializedAgent.codeEnvAvailable` (cap on × agent asks, cap on × doesn't ask, cap off × asks, neither). Catches future refactors that drop either half of the AND. - `skills.test.ts` — `injectSkillCatalog` with `codeEnvAvailable: false` against an active skill catalog must NOT register `bash_tool` even though it still registers `read_file` + `skill`. This is the state a skills-only agent gets post-narrowing. All 191 affected packages/api tests pass + 836 backend api tests pass. Typecheck clean, lint clean. * 🧽 refactor: Comprehensive-review polish — hoist tool defs, pin verifyToolAuth contract, doc appConfig Addresses the comprehensive review of Phase 8. Findings mapped: **#1 (MINOR): `verifyToolAuth` unconditional auth for execute_code** - Added doc comment explicitly stating the deployment contract (admin capability → reachable sandbox; no per-check health probe to keep UI-gate queries O(1)). - New `api/server/controllers/__tests__/tools.verifyToolAuth.spec.js` with 4 regression tests pinning the contract: 1. `authenticated: true` + `SYSTEM_DEFINED` for execute_code. 2. 404 for unknown tool IDs. 3. `loadAuthValues` is never consulted (catches a future revert that would resurface the per-user key-entry dialog). 4. Response `message` is never `USER_PROVIDED`. **#2 (MINOR): `openai/service.ts` undocumented `appConfig` dependency** - Expanded the `ChatCompletionDependencies.appConfig` JSDoc to spell out that omitting it silently disables code execution for agents with `execute_code` in their tools. External consumers of `createAgentChatCompletion` now have the contract documented at the type boundary. **#5 (NIT): `registerCodeExecutionTools` re-allocates tool defs** - Hoisted `READ_FILE_DEF` and `BASH_TOOL_DEF` to module-level `Object.freeze`d constants. The shapes derive entirely from static `@librechat/agents` exports, so a single frozen object per tool is safe to share across every agent init. Eliminates the ~4-property allocations on every call (including the common second-call no-op path). **#6 (NIT): Verbose history-priming comment in initialize.js** - Trimmed the 16-line `handlePrimeInvokedSkills` block to a 5-line summary with `@see InitializedAgent.codeEnvAvailable` pointer. The canonical narrowing explanation lives on the type; the controller comment is just the ACL-vs-capability rationale. **Skipped:** - #3 (memory spec tests a mirror function): reviewer self-dismissed as a design tradeoff; the enum-literal pin already catches the highest-risk drift vector. - #4 (cross-repo contract for `createCodeExecutionTool`): user will explicitly install the latest `@librechat/agents` dev version once the companion PR publishes, so the version pin will be authoritative. - #7 (migration/deprecation note for self-hosters): out of scope per user direction — release notes handle this. Tests verified: 679 packages/api + 840 backend api tests pass. Typecheck + lint clean. * 🔧 chore: Update @librechat/agents version to 3.1.68-dev.1 across package-lock and package.json files This commit updates the version of the `@librechat/agents` package from `3.1.68-dev.0` to `3.1.68-dev.1` in the `package-lock.json` and relevant `package.json` files. This change ensures consistency across the project and incorporates any updates or fixes from the new version. |
||
|
Danny Avila
|
ac913aa886 |
🔐 chore: Skills Permissions Housekeeping, Reachable Admin Dialog + Defaults Tests (#12766)
* 🔐 chore: Skills permissions housekeeping — reachable admin dialog + defaults tests
Phase 9 housekeeping pass. Skills was already gated on `PermissionTypes.SKILLS`
(seeded from `interface.skills`) and `AgentCapabilities.skills` everywhere it
matters, but two smaller parity gaps with Prompts/Memory/MCP remained:
- The skills admin settings dialog had no UI entry point. The only mount was
inside an unused `FilterSkills` component, so admins had no way to reach the
role-permissions editor for skills. Mounted it in `SkillsAccordion` gated on
`SystemRoles.ADMIN`, matching the `PromptsAccordion` pattern.
- No regression lock on skill permission defaults. `roles.spec.ts` asserted
structural completeness but not the specific shape — a future refactor
could silently flip ADMIN's `USE/CREATE/SHARE/SHARE_PUBLIC` to false or
drop SKILLS from USER defaults without failing. Added explicit Skills
assertions for both roles.
- No lock on `AgentCapabilities.skills` being in `defaultAgentCapabilities`.
Added an assertion in `endpoints.spec.ts`.
* 🩹 fix: Remove duplicate `const appConfig` in Responses createResponse
The Skills polish commit (#12760) added `const appConfig = req.config;` at
line 381 inside the try block of `createResponse`, without noticing that
the earlier drive-by fix (
|
||
|
Danny Avila
|
91cd3f7b7c |
🧽 refactor: Skills polish: precedence-aware body validation, controller drop logs, SkillPills rename (#12760)
Post-merge sanity-review cleanup on top of #12746: - `createSkill` / `updateSkill` now parse SKILL.md body's always-apply status once and reuse it for both validation and derivation (was parsing the same YAML block twice per call). - Body-inline `always-apply:` validation becomes precedence-aware: a caller sending an explicit top-level `alwaysApply` or a structured `frontmatter['always-apply']` no longer gets rejected for a typo in the body — the body value is never consulted at derivation time when a higher-precedence source wins. New tests cover the three relevant interactions (explicit+body-typo, frontmatter+body-typo, body-only typo still rejects). - OpenAI and Responses controllers now emit a `logger.warn` when `injectSkillPrimes` drops always-apply primes to stay under `MAX_PRIMED_SKILLS_PER_TURN`. `injectSkillPrimes` already logs internally; the controller-level warn adds endpoint context so operators can identify which path hit the cap at a glance. Mirrors AgentClient's existing log. - Rename `ManualSkillPills` → `SkillPills` (component + type + file + test + all JSDoc references). The component handles both manual and always-apply pills now; the original name was carried over from the manual-only Phase 3 and misleads new readers. - Drive-by fix: declare `appConfig = req.config` at the top of `createResponse` in `responses.js` — it was used unqualified on lines 381/396, which silently evaluated to `undefined` (via optional chaining) and disabled the skills-capability check on the Responses endpoint. Pre-existing, surfaced by lint on the touched file. |
||
|
Danny Avila
|
dfc3dfa57f |
📍 feat: always-apply frontmatter: auto-prime skills every turn (#12746)
* 🔁 refactor: Rebase always-apply work onto merged structured-frontmatter columns Phase 6 (disable-model-invocation / user-invocable / allowed-tools) landed first on feat/agent-skills. Reconcile this branch with the new mainline: - Thread alwaysApplySkillPrimes through unionPrimeAllowedTools alongside manualSkillPrimes, applying the combined MAX_PRIMED_SKILLS_PER_TURN ceiling before loading tools. - Add `_id` to ResolvedAlwaysApplySkill to match Phase 6's ResolvedManualSkill shape (read_file name-collision protection). - Register 'always-apply' in ALLOWED_FRONTMATTER_KEYS / FRONTMATTER_KIND so Phase 6's validator recognizes it. - Drop frontmatter from the listSkillsByAccess projection; the backfill helper remains as defensive code but its read path is no longer exercised on summary rows (no legacy rows exist — the branch never shipped), saving ~200KB per page. - Retire the corresponding "backfills legacy on summaries" test. - Plumb listAlwaysApplySkills through the JS controllers + endpoint initializer so the always-apply resolver sees a real DB method. * 🧹 fix: Dedupe manual/always-apply overlap, share YAML util, tidy comments Addresses review findings: - Cross-list dedup: when a user $-invokes a skill that is also marked always-apply, the always-apply copy is now dropped so the same SKILL.md body never primes twice in one turn. Manual wins (explicit intent, closer to the user message). Dedup runs in both initializeAgent (so persisted user-bubble pills stay in sync) and injectSkillPrimes (defense-in-depth at splice time). New test cases cover single-overlap, partial-overlap, and dedup-before-cap. - DRY: extract stripYamlTrailingComment to packages/data-schemas/src/utils/yaml.ts; packages/api/src/skills/import.ts now imports the shared helper. Also drop the redundant inner stripYamlTrailingComment call inside parseBooleanScalar — the call site already strips. - Mark injectManualSkillPrimes as @deprecated in favor of injectSkillPrimes (kept for external consumers of @librechat/api). - Document SKILL_TRIGGER_MODEL as forward-looking plumbing for the model-invoked path rather than leaving it as a bare unused export. - Replace the stale "frontmatter is included" comment on listSkillsByAccess with an accurate explanation of why it was intentionally excluded. * 🔒 fix: Include always-apply primes in skillPrimedIdsByName + clear alwaysApply on body opt-out Two bugs flagged by Codex review: P1 (read_file): `manualSkillPrimedIdsByName` only carried manual-invocation primes, so an always-apply skill with `disable-model-invocation: true` was blocked from reading its own bundled files, and same-name collisions could resolve to a different doc than the one whose body got primed. - Rename `buildManualSkillPrimedIdsByName` → `buildSkillPrimedIdsByName` (accepts both manual + always-apply prime arrays). - Rename the configurable field `manualSkillPrimedIdsByName` → `skillPrimedIdsByName` throughout the plumbing (skillConfigurable.ts, handlers.ts, CJS callers, tests). - Overlap resolution: manual wins on the rare edge case where the same name appears in both arrays (upstream dedup should prevent this, but defensive merging treats manual as authoritative). - New tests: (1) gate-relaxation fires for always-apply primes, (2) `_id` pinning works for always-apply same-name collisions. P2 (updateSkill): when a body update had no `always-apply:` key, `extractAlwaysApplyFromBody` returned `absent` and the column was left untouched. A skill that was once `alwaysApply: true` would keep auto-priming even after its SKILL.md no longer declared the flag. - Treat `absent` as a positive "not always-apply" declaration when the body is explicitly submitted; flip the column to `false`. - Explicit top-level `alwaysApply` still wins (three-source precedence unchanged). - New tests: body removes key → false, body has no frontmatter at all → false, explicit + body-without-key → explicit wins. * 🧵 refactor: Collapse duplicate prime types + tighten parse + test hygiene Sanity-check review follow-ups: - Collapse `ResolvedManualSkill` / `ResolvedAlwaysApplySkill` into a single `ResolvedSkillPrime` canonical interface with two backward- compatible type aliases. Both resolvers feed the same pipeline stages (injectSkillPrimes, unionPrimeAllowedTools, buildSkillPrimedIdsByName); the per-source distinction lives on `additional_kwargs.trigger`, not on the resolver output. - Move the `always-apply` branch in `parseFrontmatter` to operate on the raw post-colon text. The outer `unquoteYaml` was fine today because it's idempotent on non-quoted strings, but running it twice (once per line, once after stripping the inline comment) would be fragile if the unquoter ever grows richer YAML-escape handling. - Add the missing `alwaysApplyDedupedFromManual: 0` field to the `injectSkillPrimes` mocks in `openai.spec.js` and `responses.unit.spec.js` so they match the full `InjectSkillPrimesResult` contract. - Insert the blank line between the `unionPrimeAllowedTools` and `resolveAlwaysApplySkills` describe blocks. * 🔧 fix(tsc): Cast mock.calls via `unknown` for strict tuple destructure `getSkillByName.mock.calls[0]` is typed as `[]` by jest's generic default; a direct cast to `[string, ..., ...]` fails TS2352 under `--noEmit` even though the runtime shape matches. Go through `as unknown as [...]` like the earlier test in the same file so CI's type-check step stays green. * 🪢 fix: Propagate skillPrimedIdsByName into handoff agent tool context Handoff agents go through the same `initializeAgent` flow as the primary (with `listAlwaysApplySkills` now plumbed), so they resolve their own `manualSkillPrimes` and `alwaysApplySkillPrimes` — but the `agentToolContexts.set(...)` for handoff agents didn't carry `skillPrimedIdsByName` into the per-agent context. That meant `handleReadFileCall` fell back to the full ACL set + a `prefer*` flag for handoff agents: same-name collisions could resolve to a different doc than the one whose body got primed, and a `disable-model-invocation: true` skill primed via manual `$` or always-apply inside the handoff flow would be blocked from reading its own bundled files. Build the map via `buildSkillPrimedIdsByName(config.manualSkillPrimes, config.alwaysApplySkillPrimes)` for every handoff tool context so `read_file` behaves identically across primary and handoff agents. |
||
|
Danny Avila
|
82173f7b91 |
🛡️ feat: Persist & enforce disable-model-invocation / user-invocable / allowed-tools (#12745)
* 🧬 feat: Persist `disable-model-invocation` / `user-invocable` / `allowed-tools` Adds first-class columns mirroring the three runtime-enforced frontmatter fields, with a `deriveStructuredFrontmatterFields` helper that maps from frontmatter at create/update time and re-syncs (via `$unset`) when fields are removed. `listSkillsByAccess` projection includes them so the Phase 6 catalog filter and popover filter can both read off the summary row. Marks `invocationMode` as @deprecated on `TSkill` and the `InvocationMode` enum — the runtime now reads the persisted pair instead. * 🛡️ feat: Enforce frontmatter at runtime (catalog, skill tool, manual resolver, tool union) Wires the persisted columns into actual runtime behavior across all four invocation paths: - `injectSkillCatalog` excludes `disableModelInvocation: true` skills before catalog formatting — they cost zero context tokens and stay invisible to the model. - `handleSkillToolCall` rejects with a clear error when the model names a skill marked `disable-model-invocation: true` (defends against a stale-cache or hallucinated invocation getting past the catalog filter). - `resolveManualSkills` skips `userInvocable: false` skills with a warn log so an API-direct caller can't bypass the popover-side filter. - `unionPrimeAllowedTools` collects skill-declared `allowed-tools` minus what's already on the agent; `initialize.ts` re-runs `loadTools` for the extras and merges resulting `toolDefinitions` into the agent's effective set for the turn. Tool-name resolution is tolerant — unknown names silently drop with a debug log so cross-ecosystem skills referencing yet-to-be-implemented tools (Claude Code's `edit_file`, etc.) import without breaking. The agent document is never modified; the union is turn-scoped. Helper exports (`unionPrimeAllowedTools`) are structured so Phase 5's always-apply primes flow through the same union (combined `[...manualPrimes, ...alwaysApplyPrimes]`) once the resolver lands. Skill handler wire format gains the three fields so clients can render them on detail / list views. * 🎛️ feat: `$` popover reads `userInvocable` instead of UI-only `invocationMode` Replaces the phase-1 UI-only `invocationMode` check with the persisted `userInvocable` field (mirrors the `user-invocable` frontmatter). Skills authored with `user-invocable: false` no longer surface in the popover; the backend resolver enforces the same rule for defense-in-depth. Default-visible behavior is preserved: skills without an explicit `userInvocable` value (older rows, freshly imported skills that don't declare the field) stay visible — only an explicit `false` hides them. Test fixture updated to reflect the new field. * 🔧 fix: Address Phase 6 review findings Codex P2 + reviewer #1: Single `loadTools` call with the union of `agent.tools + allowed-tools`. The earlier two-call approach dropped `userMCPAuthMap` / `toolContextMap` / `actionsEnabled` from the skill-added pass — an MCP tool gained via `allowed-tools` would be visible to the model but fail at execution without per-user auth context. Resolution of `manualSkillPrimes` is hoisted before `loadTools` so the union can be computed up-front; the dropped-tools debug log now compares loaded vs. requested across the single call. Codex P3 + reviewer #2: `injectSkillCatalog.activeSkillIds` now includes `disable-model-invocation: true` skills. The runtime ACL check in `handleSkillToolCall` previously couldn't reach the explicit "cannot be invoked by the model" rejection because the broader access set excluded those skills. Catalog text and tool registration still gate on the visible subset (zero-context-token guarantee preserved); only the per-user `isActive` filter is a hard exclusion now. Reviewer #1 (try/catch around loadTools, MAJOR): A single bad `allowed-tools` entry from a shared skill could crash the entire turn. Now wrapped — on failure with extras, retry with just `agent.tools` and continue (the dropped-tools debug log surfaces what vanished). If the retry-without-extras still throws, propagate; the agent's own tools are the load-bearing surface. Reviewer #3 (integration tests, MAJOR): Added six tests in `initialize.test.ts` covering the full `allowed-tools` loading path: union pass-through, no-extras short-circuit, agent-baseline dedup, loadTools throw + retry, propagated throw without extras, and the empty-tools edge case. Smaller cleanups bundled in: - Reviewer #4: Moved `logger` import to the package-imports section (was wedged among local imports). - Reviewer #5: Removed unused index on `disableModelInvocation` (filtering happens application-side in `injectSkillCatalog`; index cost write overhead for zero query benefit). - Reviewer #6: Swapped order of `userInvocable` and body checks in `resolveManualSkills` so the more authoritative author-decision reason surfaces first when both apply. - Reviewer #8: Documented the `allowedTools` enforcement gap on the schema + type — model-invoked skills (mid-turn `skill` tool calls) do NOT trigger tool union, since adding tools after the graph starts would require a rebuild. Manual / always-apply (Phase 5) primes are the supported paths. - Reviewer #9: Renamed `dmi` / `ui` / `at` locals to `disableModelInvocationRaw` / `userInvocableRaw` / `allowedToolsRaw` in `deriveStructuredFrontmatterFields`. Reviewer #7 (DRY shared `getSkillByName` return type) deferred — field sets diverge meaningfully across the three call sites (handler needs `body + fileCount`; resolver needs `author + allowedTools + userInvocable`; the InitializeAgentDbMethods contract needs the superset). A `Pick<>`-based consolidation is a follow-up cleanup. * 🔧 fix: Address codex iter 2 — catalog quota + duplicate-name dedup P1: `injectSkillCatalog` cap now counts only model-visible skills, not the merged active set. The previous behavior let a tenant with many `disable-model-invocation: true` rows near the top of the cursor exhaust the 100-slot quota before any invocable skill got scanned — the catalog could end up empty even though invocable skills existed further down the paginated results. `MAX_CATALOG_PAGES` stays the ceiling on scan budget; only `visibleCount` drives the early-exit on quota fill. P2: When an invocable and a `disable-model-invocation: true` skill share a name, drop the disabled doc(s) from `activeSkillIds`. Without this dedup, `getSkillByName` (which sorts by `updatedAt` desc) could pick the disabled doc and every model call to the cataloged name would fail with "cannot be invoked by the model" instead of executing the visible skill. When ONLY a disabled doc exists for a name, it stays in `activeSkillIds` so the explicit-rejection error path still fires for hallucinated invocations. Tests: 3 new cases in `injectSkillCatalog` covering (a) cap counted on visible skills only, (b) same-name collision drops disabled doc, (c) sole-disabled-name case keeps the disabled doc. * 🔒 fix: Apply `disable-model-invocation` gate to read_file too (codex iter 3 P1) `activeSkillIds` is shared between the `skill` and `read_file` handlers. The skill-tool gate was applied last iteration, but `handleReadFileCall` authorized purely on `getSkillByName(..., accessibleIds)` — so a model that learned a hidden skill's name (stale catalog or hallucination) could still read its `SKILL.md` body or bundled files via `read_file`, defeating the contract. Same explicit rejection now fires from both handlers; no change needed to the ACL set itself (disabled docs stay in `activeSkillIds` so the explicit error path keeps firing). Two new tests in `handlers.spec.ts` cover the read_file gate and regression-protect the happy path. * 🔧 fix: Address codex iter 4 — manual-prime exception + legacy frontmatter backfill P1: Scope the `read_file` `disableModelInvocation` gate to AUTONOMOUS model probes only. A user-invoked `$` skill that is also marked `disable-model-invocation: true` had its bundled `references/*` / `scripts/*` files unreadable, leaving the manually-primed skill body referencing files the model couldn't load. Now the handler bypasses the gate when the skill name appears in `manualSkillNames` (the per-turn allowlist threaded from `manualSkillPrimes` → `agentToolContexts` → `enrichWithSkillConfigurable` → `mergedConfigurable`). Defense-in-depth: the bypass is scoped to the specific names in the allowlist; a different disabled skill name is still rejected. P2: Read-time fallback for legacy skills authored before Phase 6 landed the structured columns. `user-invocable: false` / `disable-model-invocation: true` set in `frontmatter` (the validator already accepted those keys) but with no derived column would incorrectly evaluate as "user-invocable / model-allowed" until a save backfilled the columns. New `backfillDerivedFromFrontmatter` helper fills undefined columns from frontmatter at read time in both `getSkillByName` and `listSkillsByAccess` — column wins when both are set, frontmatter fills the gap when only it's set. No DB writes; the next `updateSkill` naturally persists. `listSkillsByAccess` projection expanded to include `frontmatter` (bounded by validator, payload impact small) so summaries can also be backfilled. Sticky-primed disabled skills (ones invoked in prior turns of the same conversation) are not yet in the manual-prime allowlist — same- turn manual invocation is the load-bearing path codex flagged; the sticky-turn case is a known limitation tracked for a follow-up. Tests: 2 new in handlers.spec.ts (manual-prime allows + name-scoped block holds), 3 new in skill.spec.ts (legacy backfill via getSkillByName + listSkillsByAccess + column-wins precedence). * 🔧 fix: Address codex iter 5 — propagate manualSkillNames + keep read_file P1: `enrichWithSkillConfigurable` is also called from `openai.js` and `responses.js` (the OpenAI Responses + completions endpoints). Both were ignoring the new `manualSkillNames` parameter, which meant the manual-prime exception in the `read_file` gate (iter 4) only worked on the agents endpoint. Now all three call sites pass `primaryConfig.manualSkillPrimes?.map(p => p.name)` so manual `$` invocations of disabled skills work consistently across endpoints. P2: When every accessible skill is `disable-model-invocation: true`, the catalog text and `skill` tool are correctly omitted (no model- reachable targets) — but `read_file` and `bash_tool` MUST still be registered. A user manually invoking such a skill gets its SKILL.md body primed into context; if the body references `references/foo.md` or `scripts/run.sh`, those reads need a registered tool. Restructured `injectSkillCatalog` so `skill` registration is gated on `catalogVisibleSkills.length > 0` while `read_file` (always) and `bash_tool` (when codeEnvAvailable) register whenever any active skill is in scope. Tests: existing all-disabled test rewritten to assert read_file IS registered + skill is NOT; new test confirms bash_tool joins it when codeEnvAvailable. * 🔧 fix: Address codex iter 6 — name-collision consistency via preferInvocable P2a (resolveManualSkills): a name collision between an older user-invocable doc and a newer non-user-invocable doc made manual `$` invocation silently no-op. The popover surfaced the older invocable doc; resolver looked it up by name; `getSkillByName` returned the newer non-invocable doc; resolver skipped on `userInvocable: false`. P2b (handler / runtime ACL): with same-name duplicates (e.g. older invocable + newer disabled), the manual prime resolved to one doc while later `read_file` / `skill` execution resolved a different doc through `activeSkillIds`. Model could follow one SKILL.md body while reading files from a different skill. Both root-cause: `getSkillByName` always returned the newest match and let the caller filter, but with collisions the newest can be something the caller didn't want. Fix: extend `getSkillByName` with `options.preferInvocable`. When true, prefer the newest doc satisfying BOTH `userInvocable !== false` AND `disableModelInvocation !== true` (with frontmatter backfill); fall back to the newest match otherwise. Fast path preserved when caller doesn't opt in. Callers passing `preferInvocable: true`: - `resolveManualSkills` — picks the popover-visible invocable doc even when a newer disabled / non-user-invocable duplicate exists. - `handleSkillToolCall` — keeps execution aligned with the catalog; falls back to the disabled doc only when no invocable variant exists (so the explicit "cannot be invoked by the model" gate still fires for the hallucinated-disabled-name case). - `handleReadFileCall` — same alignment, plus the manual-prime exception added in iter 4 still applies. Tests: - 2 new in skill.spec.ts (preferInvocable picks invocable when collision exists; falls back to newest when no clean-invocable exists). - 1 new in skills.test.ts (resolver passes preferInvocable through). - 2 new in handlers.spec.ts (skill tool + read_file pass it). - Existing initialize.test.ts assertion updated for the new option. * 🔧 fix: Address codex iter 7 — split preferInvocable into per-axis flags The previous unified `preferInvocable` filter required BOTH `userInvocable !== false` AND `disableModelInvocation !== true`. That was wrong for the model paths: `userInvocable: false` skills are model-only and remain valid `skill` / `read_file` invocation targets. A duplicate-name scenario where the newer cataloged doc was model- only would let the older user-invocable variant shadow it on every model call. Split the option into two independent axes: - `preferUserInvocable` — for manual paths (`$` popover). Skips docs with `userInvocable: false`. Disable-model-invocation status is irrelevant; iter 4 explicitly supports manual prime of disabled skills. - `preferModelInvocable` — for model paths (`skill` / `read_file` handlers). Skips docs with `disableModelInvocation: true`. User- invocable status is irrelevant; model-only skills are valid here. Both flags fall back to the newest match when no preferred doc exists, so the explicit-rejection error paths still fire correctly in the sole-disabled-name case. Callers updated: - `resolveManualSkills` → `preferUserInvocable: true` - `handleSkillToolCall` / `handleReadFileCall` → `preferModelInvocable: true` Tests: - New spec test for preferModelInvocable not filtering on userInvocable. - Existing preferInvocable test renamed/split to cover the new axes. - New test asserts preferUserInvocable still returns disabled docs (preserves iter 4 manual-disabled support). - Caller tests assert each path passes the right single flag and does NOT pass the wrong one. * 🔧 fix: TypeScript type-check failure in handlers.spec.ts (CI green) `jest.fn(async () => ...)` without explicit args infers an empty tuple for the call signature, so `mock.calls[0][2]` flagged as "Tuple type '[]' has no element at index '2'." Cast to `unknown[]` then narrow to the expected option shape. Behavior unchanged. Caught by the `Type check @librechat/api` CI step (.github/workflows/backend-review.yml). * 🔧 fix: Address codex iter 8 — undefined-result fallback + read_file alignment P1 (loadTools returning undefined): Production loaders (`createToolLoader` in `initialize.js` / `openai.js` / `responses.js`) wrap `loadAgentTools` in try/catch and return `undefined` on failure rather than throwing. Without explicit handling, my iter-1 try/catch only fired for thrown errors — a silent-failure on a skill-added tool would fall through to the empty fallback and silently DROP the agent's baseline tools for the turn (much worse than just losing the extras). Added an `undefined`-result branch that retries with just `agent.tools`, mirroring the throw branch. Test pins both behaviors. P2 (read_file alignment with manual prime): When a skill is in this turn's `manualSkillNames`, the `read_file` handler now uses `preferUserInvocable` instead of `preferModelInvocable`. Same name-collision rule as `resolveManualSkills`, so the doc whose files get read is the same doc whose body got primed. For autonomous probes (skill not in `manualSkillNames`), the handler keeps `preferModelInvocable` to align with the catalog the model saw. Two new tests cover both branches and regression-protect that the wrong flag isn't passed. * 🔧 fix: Address codex iter 9 — pin read_file lookup to primed skill _id P1 (manually-primed disabled IDs were dropped from activeSkillIds): The `executableSkills` dedup in `injectSkillCatalog` correctly drops `disable-model-invocation: true` duplicates when an invocable doc shares the name — but `resolveManualSkills` legitimately primes disabled docs (iter 4 supports manual `$` invocation of disabled skills). When the resolver primed a disabled doc, the read_file handler couldn't find it in the (deduped) `activeSkillIds` and either resolved a different same-name skill or returned not-found. Fix: `ResolvedManualSkill` now carries `_id`; the legacy `initialize.js` / `openai.js` / `responses.js` controllers build a `manualSkillPrimedIdsByName` map and `enrichWithSkillConfigurable` passes it into `mergedConfigurable`. `handleReadFileCall` now pins its lookup's `accessibleIds` to `[primedId]` whenever the requested skill is in the map. The constrained set guarantees the lookup returns the EXACT doc the resolver primed — body/files come from the same source even when same-name duplicates exist or the dedup removed the prime's id from `activeSkillIds`. Autonomous read_file probes (skill not in the manual-primed map) keep the full ACL set + `preferModelInvocable` so they align with the catalog the model saw and the disabled-only case still fires the explicit-rejection gate. Test fixture changes flow from `_id` becoming required on `ResolvedManualSkill`. `buildSkillPrimeContentParts` / `injectManualSkillPrimes` widen their param types to `Pick<...>` because they only read `name` / `body` and shouldn't force test literals to invent placeholder ids. * 🧹 fix: Address independent reviewer findings (DRY + types + tests + docs) Sanity-pass review surfaced 7 findings; addressed 6 (the 7th — DRY on inline `getSkillByName` return types — is acknowledged tech debt deferred to a follow-up). #1 [MAJOR, DRY]: The 4-line `manualSkillPrimedIdsByName` map construction was duplicated across 4 CJS call sites (openai.js, responses.js x2, initialize.js). Extracted `buildManualSkillPrimedIdsByName` helper in `skillDeps.js`; all four sites now call the helper. If `ResolvedManualSkill` ever renames `_id` or gains identifying fields, only the helper changes. #2 [MINOR, type safety]: `handleReadFileCall` was casting a hex string to `Types.ObjectId[]` via `as unknown as`, relying on mongoose's auto-cast in `$in` queries. Replaced with `new Types.ObjectId(...)` so any future consumer comparing with `.equals()` / `===` gets the correct value type. Imported `Types` as a value (was type-only). #5 [MINOR, test gap]: Added a test for the worst-case silent-failure path — both the union and base-only `loadTools` calls return undefined. The agent gets no tools but the turn doesn't crash hard; pinning that contract. #4 [MINOR, performance]: Added a TODO on the `listSkillsByAccess` projection noting the `frontmatter` field can be dropped once a write migration backfills all pre-Phase-6 skills' columns. ~2KB/skill × 100/page is wasted bandwidth post-backfill. #6 [NIT, docs]: `backfillDerivedFromFrontmatter` JSDoc said "Pure" right before "mutates its undefined fields in place". Replaced with "Side-effect-free w.r.t. the DB (no writes), but mutates its argument in place" which describes both halves accurately. #7 [NIT, test determinism]: Replaced `await new Promise(r => setTimeout(r, 5))` in two same-name collision tests with explicit `updateOne` setting `updatedAt: new Date(Date.now() - 1000)` on the older doc. Removes the wall-clock race on fast CI runners. The pagination test (line 480) still uses setTimeout — that test is pre-existing and order is incidental, not load-bearing. Existing test fixtures updated to use valid 24-char hex ObjectIds (required by the iter-9 test that constructs a real `ObjectId`). #3 [MINOR, deferred]: Inline `getSkillByName` return type duplicated across `handlers.ts`, `initialize.ts`, `skills.ts`. Reviewer acknowledged this as deferred; field sets diverge across call sites (handler needs `fileCount`, resolver needs `author`/`allowedTools`). A `Pick<>`-based consolidation is a clean follow-up. |
||
|
Danny Avila
|
539c4c7e4d |
🎬 feat: Prime Manually-Invoked Skills via $ Popover (#12709)
* 🎬 feat: Prime Manually-Invoked Skills via $ Popover Lands the backend for manual skill invocation, making the $ popover deterministically prime SKILL.md before the LLM turn instead of leaving the model to discover the skill via the catalog. Flow: popover drains pendingManualSkillsByConvoId on submit, attaches names to the ask payload, controllers forward to initializeAgent, and initialize resolves each name to its body (ACL + active-state filtered, reusing the same rules as catalog injection). AgentClient splices the primes as meta HumanMessages before the user's current message. - Extract primeManualSkill / resolveManualSkills in packages/api/src/agents/skills.ts and reuse primeManualSkill inside handleSkillToolCall for a single shape source. - Thread manualSkills + getSkillByName through InitializeAgentParams / DbMethods and all three initializeAgent call sites (initialize.js, responses.js, openai.js). - Splice HumanMessage primes in client.js chatCompletion after formatAgentMessages, shifting indexTokenCountMap so hydrate still fills fresh positions correctly. - Carry isMeta / source / skillName in additional_kwargs for downstream filtering. * 🛡️ fix: Scope manual skill primes to single-agent + cap resolver input Two follow-ups to the Phase 3 priming path flagged in Codex review. Multi-agent runs: skipping the splice when agentConfigs is non-empty. `initialMessages` is shared across every agent in `createRun`, so splicing a skill body there would bypass Phase 1's per-agent `scopeSkillIds` contract — a handoff / added-convo agent with a different skill scope would see content its configuration excludes. Warn + skip is the minimal correct behavior; lifting this to per-agent initial state is a follow-up. Input bounding: `resolveManualSkills` now truncates to `MAX_MANUAL_SKILLS` (10) after dedup, with a warn listing the dropped tail. Controllers only validate `Array.isArray(req.body.manualSkills)`, so a crafted payload could otherwise fan out into an unbounded `Promise.all` of concurrent `getSkillByName` DB lookups. Cap lives in the resolver so every caller (including future `always-apply` in Phase 5) inherits it. * 🧪 refactor: Testable Helpers + Payload Validation for Manual Skill Primes Follow-ups from the comprehensive review. No behavior change for the happy path — these are architectural and defensive improvements that shrink the JS surface in /api, tighten the request-body contract, and cover the delicate splice logic with proper unit tests. - Extract `injectManualSkillPrimes` into packages/api/src/agents/skills.ts so the message-array splice and `indexTokenCountMap` shift are unit- testable in TS. client.js now calls the helper. Tests pin the `>=` vs `>` boundary condition — a regression here would silently corrupt token accounting for every message after the insertion point. - Extract `extractManualSkills(body)` and use in all three controllers (initialize.js, responses.js, openai.js). Replaces copy-pasted `Array.isArray(...) ? ... : undefined` with a helper that also filters non-string / empty elements — closes a type-safety gap where a crafted payload like `{"manualSkills": [123, {"$gt":""}]}` would otherwise reach `getSkillByName` and waste DB round-trips. - Rename `primeManualSkill` → `buildSkillPrimeMessage`. The helper serves three invocation modes (`$` popover, `always-apply`, model-invoked); the old name misled readers coming from `handleSkillToolCall`. - Add `loadable.state === 'hasValue'` guard in `drainPendingManualSkills` — defensive, since the atom has a synchronous `[]` default, but the previous `.contents` cast would have been unsound under loading/error. - Document why `resolveManualSkills` honors the active-state filter even for explicit `$` selections (Phase 2 popover filter + API-direct hardening). - Remove stray `void Types;` in initialize.test.ts — `Types` is already consumed elsewhere in that test. * 🔖 refactor: Single source for the skill-message source marker Export `SKILL_MESSAGE_SOURCE = 'skill'` and use it in both construction paths that stamp skill-primed messages — `buildSkillPrimeMessage` (for the model-invoked tool path) and `injectManualSkillPrimes` (for the user-invoked splice path). Downstream filtering and telemetry read this marker, so the two paths must agree; keeping the literal in one place removes the risk of them drifting when Phase 5's `always-apply` adds a third caller. * ♻️ refactor: Drop Multi-Agent Guard + Review Polish - Remove the multi-agent skip in `AgentClient.chatCompletion`. Leaking primes to handoff / added-convo agents via shared `initialMessages` is the agents SDK's concern to scope; this layer should just inject and let the graph handle agent-scoped state. The guard was well-intended but produced a silent-drop UX where `$skill` in a multi-agent run did nothing. - Bound the `[resolveManualSkills] Truncating ...` warn output to the first 5 dropped names plus a count suffix. A malicious payload of 1000 names was previously spilling all ~990 names into the log line. - Remove dead `?? []` from the `hasValue`-guarded loadable read in `drainPendingManualSkills` — the atom always yields a string[] when resolved, so the nullish fallback was unreachable. - Reorder skills.ts imports to follow the style guide: value imports shortest-to-longest (`data-schemas` → `langchain/core/messages` → multi-line `@librechat/agents`), type imports longest-to-shortest. * 🧠 fix: Strip Skill Primes from Memory Window + Unbreak CI Mocks Two fixes after the last push. CI unbreak: `responses.unit.spec.js` and `openai.spec.js` mock `@librechat/api` and the mock didn't expose the new `extractManualSkills` symbol, so every test in those files crashed before reaching the `recordCollectedUsage` assertion. Added `extractManualSkills: jest.fn()` returning `undefined` to both mocks; the controllers now no-op on manualSkills as the tests expect. Codex P2: `runMemory` passes `messages` straight through to the memory processor, so after the splice in `injectManualSkillPrimes`, SKILL.md bodies ride along as if they were real user chat. That pollutes memory extraction with synthetic instruction content and crowds out real turns from the window. - Export `isSkillPrimeMessage(msg)` from `packages/api/src/agents/skills.ts` — a predicate keyed on the shared `SKILL_MESSAGE_SOURCE` marker. - Filter `chatMessages = messages.filter(m => !isSkillPrimeMessage(m))` at the top of `runMemory` before the window-sizing logic. Keeps the primes visible to the LLM (they still ride in `initialMessages`) but invisible to the memory layer. - 5 new tests for the predicate covering marker-present, plain messages, different source, non-object inputs, and array filter integration. * 📜 feat: Show Skill-Loaded Cards for Manually-Invoked Skills The $ popover was priming SKILL.md bodies into the turn but leaving no visible trace on the assistant response — from the user's view it looked like the `$name ` cosmetic text did nothing. Now each manually-invoked skill renders the same "Skill X loaded" tool-call card that model-invoked skills already produce via PR #12684's SkillCall renderer. Approach: post-run prepend to `this.contentParts`. The aggregator owns per-step indices during the run, so pre-seeding collides; waiting until `await runAgents(...)` returns lets the graph settle before synthetic parts slot in at the front. - Export `buildSkillPrimeContentParts(primes, { runId })` from `packages/api/src/agents/skills.ts`. Returns completed tool_call parts (`progress: 1`, args JSON-encoded with `{skillName}`, output matching the model-invoked path's wording) that the existing `SkillCall.tsx` renderer draws identically. - In `AgentClient.chatCompletion`, prepend the built parts to `this.contentParts` immediately after `await runAgents`. Persistence and the final-event reconcile come for free — `sendCompletion` already reads `this.contentParts` verbatim. - Card ordering: skills appear first in the assistant message, reflecting that priming ran before the LLM's turn. Live-during-streaming cards are a separate follow-up — the graph's index-based aggregator makes that a bigger lift and this change delivers the core UX win without fighting the stream ordering. 6 new unit tests covering part shape, args JSON contract, output text, unique IDs, empty input, and startOffset ID differentiation. * ⚡ feat: Emit Optimistic Skill Cards + Wire Primes in OpenAI/Responses Two follow-ups from testing. Optimistic card emit: the main chat path was only showing "Skill X loaded" cards at final-reconcile time, so the user saw nothing happen until the stream finished. Now emit synthetic ON_RUN_STEP + ON_RUN_STEP_COMPLETED events right before `runAgents` starts — same pattern the MCP OAuth flow uses in `ToolService` — so the cards appear immediately. The graph's content at index 0 may overwrite them during streaming, but the post-run `contentParts` prepend (unchanged) restores them on final reconcile. OpenAI + Responses parity: both controllers were resolving `manualSkillPrimes` via `initializeAgent` but never injecting them into `formattedMessages` before the run. Manual invocation silently did nothing on `/v1/chat/completions` and the Responses API path. Now both call `injectManualSkillPrimes` on the formatted messages so the model sees SKILL.md bodies on every path. LibreChat-style card SSE events don't apply to these OpenAI-shaped responses, so the live-emit is chat-path-only. - Export `buildSkillPrimeStepEvents(primes, { runId })` from `packages/api/src/agents/skills.ts`. Uses `Constants.USE_PRELIM_RESPONSE_MESSAGE_ID` by default so the frontend maps events to the in-flight preliminary response message, matching the OAuth emitter. - In `AgentClient.chatCompletion`, emit via `sendEvent` (or `GenerationJobManager.emitChunk` in resumable mode) after `injectManualSkillPrimes` runs, before the LLM turn begins. - Wire `injectManualSkillPrimes` into `openai.js` + `responses.js` after `formatAgentMessages`. Refactored the destructure to `let` on `indexTokenCountMap` so the injector's returned map is usable. - 8 new unit tests covering the step-event builder: pair cardinality, default/custom runId, TOOL_CALLS shape + JSON args, progress:1 on completion, index ordering, stepId/toolCallId pairing, empty input. * 🎯 fix: Route Skill Prime Events to the Real Response + Sparse-Array Offset Two bugs in the optimistic-card emit from the last pass. 1. Wrong runId. The events used `USE_PRELIM_RESPONSE_MESSAGE_ID` (the MCP OAuth pattern), but OAuth emits DURING tool loading — before the real response messageId exists. By the time skill priming fires, the graph is about to emit with `this.responseMessageId`, so the PRELIM runId orphaned every card onto the client's placeholder response entry in `messageMap`, separate from the one the LLM's events were building. Net effect: cards never rendered mid-stream. Now passing `this.responseMessageId` — the same ID `createRun` receives — so synthetic and real steps land on the same `messageMap` entry. 2. Index 0 collision. With the runId fixed, card-at-0 would have hit `updateContent`'s type-mismatch guard when the LLM's text delta arrived at the same index, suppressing the whole text stream. New `SKILL_PRIME_INDEX_OFFSET` = 100 placed on both the live SSE emit and the server-side `contentParts` assignment. Sparse array during streaming renders as `[llm_text, ..., card]` (skip-holes via `Array#filter` / `Array#map`). `filterMalformedContentParts` from `sendCompletion` compacts to dense `[text, card]` before persistence, so streaming UI and saved message agree on order — no finalize reorder jank. Post-run switches from `contentParts.unshift` to `contentParts[OFFSET + i] = part` to mirror the live placement. - Add `startIndex` option to `buildSkillPrimeStepEvents` with `SKILL_PRIME_INDEX_OFFSET` default. Export the constant from `@librechat/api` so `client.js` can reuse it for the post-run splice. - Update the existing index-ordering test to the new default and add a new test for the explicit `startIndex` override. * 🎗️ feat: Replace \$skill-name Text with Pills on the User Message The `$skill-name ` cosmetic text the popover was inserting into the textarea had two problems: it lingered in the user message forever (the card is a more meaningful marker), and it implied that free-form text invocation like \"\$foo help me\" should work — which it doesn't, and supporting it would mean another parsing layer nobody asked for. Dropped the textarea insertion. Visual confirmation after submit now comes from a compact `ManualSkillPills` row on the user bubble that self-extinguishes once the backend's live skill-card stream (`buildSkillPrimeStepEvents` from the last commit) populates the sibling assistant response. Multiple skills render as multiple pills — the atom was already a string array, so multi-select works for free. - `SkillsCommand.tsx`: select handler no longer writes to the textarea. Still drops the trigger `$` via `removeCharIfLast`, still pushes to `pendingManualSkillsByConvoId`, still flips `ephemeralAgent.skills`. - `families.ts`: new `attachedSkillsByMessageId` atomFamily keyed by user messageId. `useChatFunctions.ask` writes the drained skill list here on every fresh submit (regenerate/continue/edit still skip). - `ManualSkillPills.tsx` renders pills conditionally: hidden when the message isn't a user message, when no skills are attached, or when the sibling assistant response already carries a `skill` tool_call content part (the live card took over). Reads messages via React Query so we don't re-render on every message-state keystroke. - `Container.tsx` mounts the pills above the user message text, parallel to the existing `Files` slot. - Updated the SkillsCommand select-flow spec to assert the textarea is cleared of `$` instead of populated with `\$name `. 5 new tests for `ManualSkillPills` covering empty state, non-user message guard, multi-skill rendering, the skill-card hide condition, and the text-only-content-doesn't-hide case. * 🎛️ feat: Manual Skills as Persisted Message Field + Compose-Time Chips Three problems with the previous pass: 1. Cards rendered BELOW the LLM text on the assistant message (and stayed there on reload) because the sparse index-100 offset put them after the model's content. Now back to `unshift` — cards at the top, same as before the live-emit detour. 2. Pills on the user message disappeared the moment the live card arrived, so users barely saw them. The live-emit channel also added meaningful complexity and relied on a per-message Recoil atom that had no clean cleanup story. 3. No visual cue at all during new-chat compose — the `$name ` text was removed, the submitted-message pills weren't there yet, and the popover closes after selection. User had no way to see what they'd queued up before sending. New architecture: `manualSkills` is a first-class field on `TMessage`, persisted by the backend on the user message. `ManualSkillPills` reads straight from `message.manualSkills` — no atom, no sibling-lookup — so pills survive reload, show in history, and stay for the lifetime of the message. Compose-time chips above the textarea read the existing `pendingManualSkillsByConvoId` atom and let users × skills out before submitting. Backend reverts: - `client.js`: dropped the `ON_RUN_STEP` live-emit loop, restored `this.contentParts.unshift(...primeParts)` so cards sit at the top of the persisted assistant response. - `skills.ts`: removed `buildSkillPrimeStepEvents` and `SKILL_PRIME_INDEX_OFFSET` (both unused now). `GraphEvents`, `StepTypes`, and `Constants` imports went with them. Removed 8 tests. Field persistence: - `tMessageSchema` gains `manualSkills: z.array(z.string()).optional()`. - Mongoose message schema gains `manualSkills: { type: [String] }` with matching `IMessage` TS field. - `BaseClient.js` reads `req.body.manualSkills` on user-message save, filters to non-empty strings, pins onto `userMessage` before `saveMessageToDatabase`. Mirrors the existing `files` pattern right above it. Runtime resolution still reads top-level `req.body.manualSkills` — persistence and resolution are separate concerns. Frontend: - `useChatFunctions.ask` sets `currentMsg.manualSkills` directly; the drained atom value goes onto the message, not a separate atom. Removed the `attachSkillsToMessage` Recoil callback. - `ManualSkillPills`: pure render of `message.manualSkills`. No more `useQueryClient`, no sibling scan, no atom read. Loses the auto-hide-when-card-arrives behavior — pills stay on the user bubble, cards live on the assistant bubble, both are informative. - Dropped the `attachedSkillsByMessageId` atomFamily and its export. - New `PendingManualSkillsChips` above the textarea reads the compose-time atom and renders chips with × to remove. Mounted in `ChatForm` right after `TextareaHeader`. Naturally hides on submit when the atom drains. Tests: updated `ManualSkillPills` suite to the new field-based reads (5 passing). New `PendingManualSkillsChips` suite covering empty state, multi-chip render, single × removal, and full-clear (4 passing). Backend suite trimmed to 89 (was 97) from the step-events test removal — no regressions on the remaining helpers. * 🧪 feat: Assistant-Side Skill-Loading Chips + Pill Padding Two small UX fixes on top of the field-on-message architecture. Pill padding: bumped the user-side `ManualSkillPills` from `py-0.5` to `py-1` on each chip and added `py-0.5` to the wrapper so the row breathes a little without feeling tall. Mid-stream indicator: new `InvokingSkillsIndicator` mirrors the parent user message's `manualSkills` onto the assistant bubble as transient "Running X" chips while the real card is in flight. Renders above `ContentParts` in `MessageParts`. Hides itself when the assistant's own `content` grows a `skill` tool_call — the authoritative card from `buildSkillPrimeContentParts.unshift` is showing, so the placeholder steps aside. No SSE emit, no aggregator injection, no index collision with the LLM's streaming content: just a render slot keyed off the parent's field. Why not stream the cards live: whichever content index we'd choose either blocks the LLM's text stream (`updateContent` type-mismatch at index 0) or lands below the response after sparse compaction (index 100+). Mirroring the parent field sidesteps the aggregator entirely and gives the user an immediate "skill is loading" signal that naturally gives way to the real card at finalize. Covers the gap the user flagged: pills on the user message said "I asked for these" but nothing on the assistant side said "we're working on it" until the stream finished. 5 new tests for the indicator: user-msg guard, missing parent-field guard, multi-chip render, hides-on-card-landing, orphan-parent guard. * 🔁 fix: Indicator Visibility + Carry Manual Skills Through Regenerate/Edit Two bugs. Indicator never rendered: `InvokingSkillsIndicator` looked up the parent user message via `queryClient.getQueryData([QueryKeys.messages, convoId])`, but on a new chat the React Query cache is keyed by `"new"` (the URL `paramId`) until the server assigns a real conversation ID — while `message.conversationId` on the assistant message is already the server ID. Lookup missed, `skills.length === 0`, nothing rendered. Switched to `useChatContext().getMessages()`, which reads from the same `paramId` the rest of the UI uses, so new-chat and existing-chat cases both resolve to the correct message list. Regenerate / save-and-submit dropped manual skills: the compose-time `pendingManualSkillsByConvoId` atom is drained on the first submit, so replaying that turn later found an empty atom and sent `manualSkills: []`. The pills were still on the user bubble, so from the user's point of view the model was running primed — but the backend saw nothing and produced an unprimed response. - Added `overrideManualSkills?: string[]` to `TOptions`. Callers with a reference message pass its persisted `manualSkills`; `useChatFunctions.ask` uses the override verbatim when present, otherwise falls back to the existing drain-or-empty logic. - `regenerate` in `useChatFunctions` passes `parentMessage.manualSkills` — the user message being regenerated has the field persisted by the backend, so the second turn primes the same skills as the first. - `EditMessage.resubmitMessage` covers both edit branches: - User-message save-and-submit: forwards the edited message's own `manualSkills` so the new sibling turn primes identically. - Assistant-response edit: forwards the parent user message's `manualSkills` for the same reason. Indicator test suite converted from `@tanstack/react-query` harness to a jest-mocked `useChatContext().getMessages()`. 6 tests (was 5), added a cache-miss case. * 🧭 fix: Drive Mid-Stream Skill Chips from Submission Atom, Not Message Lookup Message-ID-keyed lookups kept racing the stream: the user message flips from its client-side intermediate UUID to the server-assigned ID mid-run, conversation IDs flip from the URL `paramId="new"` to the real convo ID on brand-new chats, and the React Query cache splits briefly between the two. Previous attempts — direct `queryClient.getQueryData` and then `useChatContext().getMessages()` — each missed a different window. `TSubmission.manualSkills` is already populated at `ask()` time and the submission atom (`store.submissionByIndex(index)`) is the single stable anchor across the whole lifecycle: set once at submit, lives through every SSE event, cleared when the stream ends. No ID lookups, no cache timing. - `InvokingSkillsIndicator` now reads `submissionByIndex(index)` via Recoil. Shows chips when: • the message is assistant-side, • a submission is in flight with non-empty `manualSkills`, • the assistant's `parentMessageId` matches `submission.userMessage.messageId` (so chips appear only on the bubble for the current turn, never on siblings), • the assistant's own content doesn't yet carry a `skill` tool_call (real card takes over from the server's post-run `contentParts.unshift`). - Drops the `useChatContext().getMessages()` dependency and the `useQueryClient` dependency before that. No more lookups by conversationId or messageId. Test suite now mocks `useChatContext` to supply `index: 0` and seeds the `submissionByIndex(0)` atom via Recoil initializer. 6 cases cover user-side, no-submission history, empty `manualSkills`, multi-chip render, hides-on-card-landing, and wrong-turn guard. * 🌱 fix: Seed Response manualSkills in createdHandler, Indicator Becomes Pure The mid-stream indicator kept getting wired off state I don't own: first `queryClient.getQueryData` (raced the new-chat paramId flip), then `useChatContext().getMessages()` (same cache, same race), then `useRecoilValue(submissionByIndex)` (pulled every message into the submission subscription — re-renders all indicators on any submission change, exactly the "limit hooks in rendering" concern). Cleanest path is the one the user pointed at: the submission owns the data, `useSSE` / `useEventHandlers` owns the save points, so seed the field ONTO the response message at the save site and let the indicator be a pure prop-read. - `createdHandler` now writes `manualSkills` onto the initial response from `submission.manualSkills` at the moment the placeholder enters the messages array. The field rides through the normal mutation pipeline via spreads (`useStepHandler` response creation, `updateContent` result returns) — no special handling needed. - `InvokingSkillsIndicator` drops the Recoil / context / queryClient reads. Pure function of `message`: if assistant, has `manualSkills`, and `content` hasn't grown a `skill` tool_call yet, render chips. Only `useLocalize` left, which was already unavoidable for the i18n string. - Renders decouple: no single state change (`submissionByIndex` flip, React Query cache update) forces every indicator in the message list to re-render anymore. Only the message whose prop changed re-runs. Finalize story unchanged: server's `responseMessage` doesn't carry the frontend-only `manualSkills` field, so `finalHandler`'s replacement drops it — but by then the real `skill` tool_call is in `content` and the indicator's content-scan hides itself anyway. Test suite back to pure prop mocks: 7 cases covering user-guard, no-seed, multi-chip render, skill-card-hide, non-skill-tool-call-keeps, text-only-keeps, and missing message. * 🪞 fix: Render Skill Indicator Inside ContentParts, Adjacent to Parts The indicator still wasn't showing because even though MessageParts mounted it as a sibling of ContentParts, ContentParts is a `memo`'d component that owns the only rendering path that refreshes in lockstep with content deltas. Mounting above it put the indicator one layer further out — reachable, but not exercised on the same render cycle that processes the streaming `message` prop. Moved the indicator into ContentParts itself, rendered at the top of both the sequential and parallel branches. Reads the `message` prop (newly threaded through as an optional prop alongside `content`), so: - Same render cycle as Parts — updates from the SSE pipeline flow through the same pathway. - Lives outside the `content.map`, so delta-driven content reshuffles never wipe it. - Still a pure prop-read inside the indicator itself (no Recoil, queryClient, context hooks). The only dep is `useLocalize`. Thread: - `ContentPartsProps` gains `message?: TMessage`. - `MessageParts` passes `message={message}` through, drops its own indicator mount + import. - `ContentParts` renders `<InvokingSkillsIndicator message={message} />` in both the parallel-content and sequential-content branches, right under `MemoryArtifacts` and before the empty-cursor / parts map. Companion data flow (unchanged): `createdHandler` seeds `initialResponse.manualSkills` from `submission.manualSkills`; the field rides through `useStepHandler` via spreads; indicator hides on `skill` tool_call landing in `content`. * 🔎 refactor: Narrow Skill Components to Scalar skills Prop, Kill Memo Churn Passing the full `message` object into presentational components busts `React.memo` shallow comparisons every time the message reference changes for unrelated reasons. Swap to scalar `skills?: string[]` throughout: - `InvokingSkillsIndicator`: props-only (`skills?: string[]`); visibility logic (user-vs-assistant, skill tool_call arrival) now lives in the caller so this stays pure presentational. - `ManualSkillPills`: props-only (`skills?: string[]`). - `ContentParts`: takes `manualSkills?: string[]` scalar, computes `showInvokingSkills` once per render from `manualSkills` + content scan for the `skill` tool_call, then mounts the indicator with `skills=` prop in both parallel and sequential branches. - `MessageParts`: passes `manualSkills={message.manualSkills}` through to `ContentParts`. - `Container`: passes `skills={message.manualSkills}` to `ManualSkillPills`. - Tests updated to exercise the narrowed prop surface. * 📜 feat: Mid-Stream Skill Cards via SkillCall, Drop Custom Indicator Instead of a separate `InvokingSkillsIndicator` chip component, render pending skill placeholders through the existing `SkillCall` renderer — same component the backend's finalized prime part uses. The loading visual (`progress < 1` + empty output → pulsing "Running X") and the completed visual ("Ran X") now come from one source of truth. `ContentParts` computes `pendingSkillNames` from `manualSkills` minus any `skill` tool_call already in `content` (dedupe by `args.skillName` since the synthetic's id differs from the real one). Those names render through a separate slot ABOVE the Parts iteration — not prepended to the content array, which would shift React keys on every downstream streaming text / tool part and force unmount/remount mid-stream. When the real prime `tool_call` lands at finalize (backend unshifts to content[0..]), `collectExistingSkillNames` picks it up, the pending set empties, and the real part takes over rendering in the Parts iteration. Layout is identical either way because primes are always at the top of content. - `InvokingSkillsIndicator.tsx` + test deleted (no longer referenced) - `ContentParts.tsx` renders `<SkillCall .../>` directly for pending names, mirrors `Part.tsx`'s usage of the same component - `createdHandler` doc comment updated to reflect the new flow * ✂️ fix: Render Interim Skill Cards From manualSkills Only, Leave Content Untouched Previous revision read `content` to de-dupe pending cards against real `skill` tool_calls, so any optimistic skill part streamed from the backend would race our placeholder off the screen mid-turn — exactly the "getting overridden" symptom. Now: interim `SkillCall` cards are driven purely by the response message's `manualSkills` field. `content` is never inspected here, so no backend delta can pull the cards down. The field is now seeded directly onto the assistant placeholder in `useChatFunctions` (not only in `createdHandler`) so the cards appear from the first render, before the `created` SSE event round-trip. Lifecycle: - `useChatFunctions` puts `manualSkills` on the freshly-minted `initialResponse` — cards render the instant the placeholder lands. - `createdHandler` keeps its own re-seed (idempotent; safe) so a regenerate / save-and-submit flow that hits that path still works. - `useStepHandler` spread operations preserve the field through every content update. - `finalHandler` replaces the message with the server-backed `responseMessage` (no `manualSkills`) — cards disappear, and the real `skill` tool_call part in `content` takes over. ContentParts changes: - Drop `collectExistingSkillNames` / `parseJsonField` dedupe path. - `renderPendingSkills` reads only `manualSkills` + `isCreatedByUser`. - Simpler control flow — one boolean (`hasPendingSkills`) gates the early return, one function renders. * 🩹 fix: Codex Review Resolutions — Localization, Guards, Tests, Docs Addresses seven findings from comprehensive code review: Finding 1 (MAJOR) — Document sticky re-priming as intentional - `buildSkillPrimeContentParts`: expanded doc comment explaining synthetic `skill` tool_calls persist and get re-primed on every subsequent turn via `extractInvokedSkillsFromPayload` (shape parity with model-invoked skills). This matches the UX: the assistant skill card is a visible, persistent signal that the skill is active for the conversation. Not a bug — called out explicitly so future maintainers don't mistake it for one. Finding 2 (MAJOR) — Add ContentParts render tests - New `ContentParts.test.tsx` with 7 cases covering the interim skill card logic: assistant-only rendering, user-message suppression, undefined-content safety, parallel+sequential branch integration, progress<1 (pending) state. Child components mocked so the test exercises only the branching and prop wiring ContentParts owns. Finding 3 (MINOR) — Localize hardcoded aria-labels - Added `com_ui_skills_manual_invoked` + `com_ui_skills_queued` keys. - Reused existing `com_ui_remove_skill_var` for the remove-button aria-label. - `PendingManualSkillsChips` and `ManualSkillPills` now call `useLocalize()`. Test mocks updated to the label-echo pattern. Finding 4 (MINOR) — Max-length guard in `extractManualSkills` - New `MAX_SKILL_NAME_LENGTH = 200` constant and filter. Blocks a crafted payload like `{ manualSkills: ['a'.repeat(100000)] }` from reaching `getSkillByName` / Mongo's query planner. Finding 5 (NIT) — `BaseClient.js` comment contradicted itself - Rewrote to call the filter what it is: defense-in-depth on top of Mongoose schema validation, not a redundant second layer. Finding 6 (NIT) — `ManualSkillPills` now wrapped in `React.memo` - Consistent with peer components (`PendingManualSkillsChips`, `ContentParts`). Rendered inside `Container`, which re-renders on every content update, so the memo is a real cycle savings. Finding 7 (NIT) — Redundant guard in `ContentParts.renderPendingSkills` - Collapsed the duplicate null-check by computing `pendingSkills` as a `useMemo`'d array (`[]` when not applicable), and mapping directly. `hasPendingSkills` now derives from the array length — one source of truth, no redundant gate inside the render function. * 🔧 fix: Update ParallelContent to Handle Optional Content Prop Modified the `ParallelContentRendererProps` to make the `content` prop optional, ensuring safer access within the component. Adjusted the calculation of `lastContentIdx` to handle cases where `content` may be undefined, preventing potential runtime errors. This change enhances the robustness of the component when dealing with varying message structures. * 🎯 fix: Thread manualSkills Through ContentRender — The Real Renderer This is why the interim skill cards never appeared across many rounds of iteration: `ContentRender.tsx` (the memo'd renderer used by most paths, including the agents endpoint) was calling `ContentParts` without the `manualSkills` prop. Only `MessageParts.tsx` had it wired up — and that's not the component that actually renders the assistant response in production. Two fixes: 1. Pass `manualSkills={msg.manualSkills}` to the `ContentParts` call. 2. Extend the `areContentRenderPropsEqual` memo comparator to include `manualSkills.length`, otherwise a message update that adds the field (seeded by `useChatFunctions` on the initialResponse) would be bailed out by the memo and never re-render. Verified the two ContentParts call sites are now consistent; Container usages for `ManualSkillPills` on the user side were already correct. * 🧹 polish: Address Audit Follow-Up (F1/F3/F6) F1 — Clarify sticky re-priming opt-out path. The previous comment said "regenerate without the pick" as one opt-out, but `useChatFunctions.regenerate` forwards the original picks via `overrideManualSkills`, so regeneration alone keeps the skill sticky. Updated to: edit the originating message to remove the pills and resubmit, or start a new conversation. F3 — Add DOM-order assertions to the parallel + sequential tests. The two "alongside" tests verified both elements existed but didn't pin the ordering contract. Both now use `compareDocumentPosition` to assert the pending SkillCall precedes the real content, matching the backend semantic (`contentParts.unshift(...primeParts)` puts primes at the top). F6 — Fix package import order in PendingManualSkillsChips. `recoil` (58 chars) was listed before `lucide-react` (45 chars) which violates the "shortest to longest after react" rule in AGENTS.md. Swapped order; no behavior change. F2 / F4 / F5 from the audit were confirmed as non-issues (React-safe empty map, cosmetic test-mock artifact, accepted memo tradeoff) and require no change. * ✨ feat: Dedicated PendingSkillCall + Running→Ran Transition on Real Content UX polish on the interim skill card now that it's actually rendering: 1. New `PendingSkillCall` component (mirrors `SkillCall` visually but drops the expand affordance). `SkillCall`'s underlying `ProgressText` always renders a chevron + clickable button when any input is present, which on a card with empty output points at nothing — misleading cursor:pointer and a no-op toggle. The pending variant has only the icon + label, no button wrapper, no chevron. 2. "Running X" → "Ran X" transition when real content lands. `ContentParts` computes `hasRealContent` (any non-text part, or a text part with non-empty content — placeholder empty-text parts don't count) and passes `loaded={hasRealContent}` to `PendingSkillCall`. Matches what users see for model-invoked skills as they finish priming: pulsing shimmer → static icon. 3. Cleanup: - Dropped direct `SkillCall` import from `ContentParts` (replaced by `PendingSkillCall`). `SkillCall` is still used by `Part` for real `skill` tool_call content parts — no behavior change there. - Removed the now-redundant explicit `manualSkills` assignment in `createdHandler`. `useChatFunctions` seeds the field on `initialResponse` at construction, so the `...submission.initialResponse` spread already carries it through — the re-assignment was defensive belt-and-suspenders doing the same work twice. Comment rewritten to describe the actual lifecycle. Tests updated to the new component (12/12 pass): two new cases pin the loaded-state transition (unloaded when content has no real parts, flips to loaded once a non-empty text part lands). |
||
|
Danny Avila
|
3f1bde4831 |
🪆 feat: Compose Agent Scope and Active-State Filters in $ Popover (#12708)
* feat: compose per-agent scope and per-user active-state filters in $ popover
Stack two runtime-truth filters on top of the existing `isUserInvocable`
check so the `$` skill popover matches what will actually be available at
turn time:
- Per-agent skill scope from `agent.skills` (resolved via useChatContext
+ useAgentsMapContext), mirroring backend `scopeSkillIds` semantics:
`undefined`/`null` → no scope, `[]` → empty, non-empty → intersection.
- Per-user ownership-aware active state via `useSkillActiveState().isActive`.
The filters are composed in a pure, exported `filterSkillsForPopover`
helper so the agent-scope ∩ active ∩ invocable matrix can be unit-tested
without rendering the component. Short-circuits on cheapest check first
(agent scope → active → invocation mode).
Backend still enforces both filters at runtime; this PR is a UX mirror so
users do not see popover entries that would be filtered out by the time
the LLM turn begins.
* refactor: thread agentId into SkillsCommand as a prop
Drop the direct `useChatContext()` call inside SkillsCommand in favor of
receiving `agentId` from ChatForm. The parent already subscribes to the
conversation via its single useChatContext call, and SkillsCommand is
wrapped in React.memo — threading the id as a prop means the popover only
re-renders when agent_id actually changes instead of on every unrelated
conversation-shape mutation. Mirrors the pattern AttachFileChat already
uses for `conversation` / `agentId`.
No behavior change; filter semantics and test coverage are identical.
* fix: fail closed when agent skill scope cannot be resolved
Previously, if `conversation.agent_id` was set but the agents map had no
entry for it (hydration pending, query failure, or missing VIEW access),
the popover treated the scope as `undefined` and showed the full ACL
catalog. That leaks options the backend will reject at turn time, the
opposite of what this phase is meant to do.
Distinguish the unresolved cases ("map undefined" and "agent not in map")
from the intentionally-unconfigured case ("agent exists, no `skills`
field") and return `[]` for the former, preserving the backend semantics
of `scopeSkillIds` only for the latter. Adds two tests covering both
fail-closed branches.
* fix: surface agent.skills in list projection and treat ephemeral ids as unscoped
Two holes flagged on the earlier fail-closed commit:
1. The list-agents projection in `getListAgentsByAccess` omitted the
`skills` field, so `agentsMap[agentId].skills` was always undefined
and the popover fell back to the full ACL catalog for every scoped
agent — the opposite of this phase's intent. Add `skills: 1` to the
projection and lock it in with a new test.
2. Conversations can carry ephemeral agent ids (e.g. `ephemeral` after
switching off the agents endpoint) that intentionally don't live in
the agents map. The prior fail-closed branch blanked the popover in
those cases. Treat anything that doesn't start with `agent_` as
unscoped via the existing `isEphemeralAgent` helper so the popover
shows the full ACL-visible catalog, matching how no-agent convos
already behave.
Frontend: 18 tests pass (adds one ephemeral-id case).
Backend: 10 getListAgentsByAccess tests pass (adds one projection case).
* refactor: pass through hydration race, only fail closed when map is authoritative
Split the two "cannot resolve scope" cases that were previously collapsed
onto the same fail-closed branch:
- `agentsMap === undefined` means the agents list query has not settled
yet. Return `undefined` (full catalog). The map typically hydrates well
before the first `$` open, and the backend still scopes at turn time —
blanking the popover during a sub-second race produces worse UX with
no security benefit.
- `agentsMap` is populated but the agent is absent means the agent was
deleted or the user's VIEW access was revoked mid-session. That is
authoritative missing state, so keep the fail-closed behavior — the
full catalog would be misleading.
Updates the associated test case to assert the hydration-race path now
shows the catalog, and rewrites the in-code comment to distinguish the
two branches.
* refactor: drop `as string` in agent scope memo by tightening the guard
`isEphemeralAgent` returns true for null/undefined so the original code
was runtime-safe, but its signature returns plain `boolean` rather than
a type predicate, so TypeScript never narrowed `agentId` to `string` and
the subsequent map lookup required an `as string` assertion. Split the
guard into `!agentId || isEphemeralAgent(agentId)` so the narrowing falls
out naturally and the assertion can be removed.
No behavior change; 18 tests pass.
|
||
|
Danny Avila
|
9225a279eb |
🎚️ feat: Per-User Skill Active/Inactive Toggle with Ownership-Aware Defaults (#12692)
* feat: per-user skill active/inactive toggle with ownership-aware defaults - Add `skillStates` map (Record<string, boolean>) to user schema for per-user active/inactive overrides on skills - Add `defaultActiveOnShare` to interface.skills config (default: false) so admins can control whether shared skills auto-activate - Add GET/POST /api/user/settings/skills/active endpoints with validation - Add React Query hooks with optimistic mutations for skill states - Add useSkillActiveState hook with ownership-aware resolution: owned skills default active, shared skills default inactive - Add toggle switch UI to SkillListItem and SkillDetail components - Filter inactive skills in injectSkillCatalog before agent injection - Add localization keys for active/inactive labels * fix: use Record instead of Map for IUser.skillStates Mongoose .lean() flattens Map to a plain object, causing type incompatibility with IUser in methods that return lean documents. * fix: address review findings for skill active states - Fail-closed when userId is absent: filter rejects all shared skills instead of passing them through unfiltered (Codex P1) - Validate Mongoose Map key characters (reject . and $) in controller to return 400 instead of a 500 from schema validation (Codex P2) - Block toggle while initial skill states query is loading to prevent overwriting server-side overrides with an empty snapshot (Codex P2) - Extract shared SkillToggle component, eliminating duplicate toggle markup in SkillListItem and SkillDetail (Finding #3) - Move skill state query/mutation hooks from Favorites.ts to Skills/queries.ts per feature-directory convention (Finding #4) - Fix hardcoded English aria-label in SkillListItem by passing the localized string from the parent SkillList (Finding #5) - Fix inline arrow in SkillList render loop: pass stable callback reference so SkillListItem memo() is not invalidated (Finding #1) - Extract toRecord() helper in controller to DRY the Map-to-Object conversion (Finding #6) - Remove Promise.resolve wrapping synchronous config read (Finding #8) - Remove unused TUpdateSkillStatesRequest type (Finding #12) * fix: forward tabIndex on SkillToggle to preserve list keyboard nav The original inline toggle had tabIndex={-1} so the row itself remained the sole tab target. The extraction into SkillToggle dropped this prop, making every list toggle a tab stop. Add an optional tabIndex prop and pass -1 from SkillListItem. * fix: plumb skillStates to all agent entry points, isolate toggle keydown - Add skillStates/defaultActiveOnShare loading to openai.js and responses.js controllers so shared-skill activation is respected across all agent entry points, not just initialize.js (Codex P1) - Stop keydown propagation on SkillToggle so Enter/Space does not bubble to the parent row's navigation handler (Codex P2) * fix: paginate catalog fetch and serialize toggle writes - Paginate listSkillsByAccess (up to 10 pages of 100) until the active catalog quota is filled, so inactive shared skills in recent positions do not starve active owned skills past the first page (Codex P1) - Extend listSkillsByAccess interface with cursor/has_more/after for catalog pagination - Serialize skill-state writes via a ref queue: one in-flight request at a time, with the latest desired state sent when the previous one settles. Prevents last-response-wins races where an older request overwrites newer toggles (Codex P2) * fix: share write queue across hook instances, block toggle on fetch error - Move the write queue from a per-instance useRef to a module-scoped object so every mount of useSkillActiveState (SkillList, SkillDetail, etc.) serializes against the same in-flight slot. Prior per-instance queues allowed two components to race full-map POSTs (Codex P1) - Extend the toggle guard beyond isLoading: also block when isError is true or data is undefined. Prevents a failed GET from seeding a toggle with an empty baseline that would wipe server-side overrides on the next successful POST (Codex P1) * fix: stale closure, orphan cleanup, and cap-error UX - Read toggle baseline from React Query cache via queryClient.getQueryData instead of the captured skillStates closure. The closure can be stale between onMutate's setQueryData and the next render, so rapid successive toggles would build on old state and drop earlier changes (Codex P1) - Surface the MAX_SKILL_STATES_EXCEEDED error code with a specific toast key (com_ui_skill_states_limit) so users understand the 200-cap rather than seeing a generic error - Prune orphaned entries (skillIds whose Skill doc no longer exists) on both GET and POST in SkillStatesController. Self-heals over time without needing cascade-delete hooks or a migration job. Uses one indexed Skill._id query per request * test: pin skill active-state precedence with unit tests Extract the active-state resolution logic from a closure inside injectSkillCatalog into an exported resolveSkillActive helper, then cover every branch of the precedence matrix: - Fails closed when userId is absent (even with defaultActiveOnShare=true) - Explicit override wins over ownership and config (both true and false) - Owned skills default to active when no override is set - Shared skills default to defaultActiveOnShare value - Undefined skillStates behaves identically to an empty object - defaultActiveOnShare defaults to false when omitted - Owned skills ignore defaultActiveOnShare entirely Closes Finding #2 from the pre-rebase comprehensive review. Mirrors the existing scopeSkillIds test style; injectSkillCatalog now calls resolveSkillActive instead of inlining the closure. * refactor: limit skill active toggle to detail header, drop label - Remove the per-row toggle from SkillListItem and the active-state plumbing (hook call, isSkillEnabled/onToggleEnabled/toggleAriaLabel props) from SkillList. The detail view is now the single place to change a skill's active state - Drop dim/muted styling for inactive skills in the sidebar: without a control there, the visual indication has nowhere to land - Resize SkillToggle to match neighbor buttons: outer h-9 container, h-6 w-11 track with size-5 knob, no label span. The 'Active' / 'Inactive' text that accompanied the detail-view toggle is removed - Remove the now-unused label prop and tabIndex prop (the tabIndex existed only for the list-row context) from SkillToggle. Drop the onKeyDown stopPropagation for the same reason - Remove now-orphaned com_ui_skill_active / com_ui_skill_inactive translation keys * style: shrink SkillToggle track to h-5 w-9 with size-4 knob Container stays at h-9 to match neighbor button heights. The toggle track itself drops from h-6 w-11 to h-5 w-9, with a size-4 knob travelling 1.125rem on activation. Visually lighter inside the row. * fix: remove redundant skillStates entries that match the resolved default When a toggle lands on the ownership/config default, delete the key from the map instead of persisting `{id: defaultValue}`. Without this, a user toggling a skill off and back on would leave `{id: true}` for an owned skill (whose default is already true), silently consuming a slot against the 200-entry cap. Repeated round-trip toggles could exhaust the quota with zero meaningful overrides (Codex P2). Preserves the exceptions-list invariant that the runtime-resolution design depends on. * fix: prune before enforcing skill-state cap; reject non-ObjectId keys Reorder the update controller so pruneOrphans runs before the 200-cap check. Without this, a user near the cap with some orphaned entries (skills deleted since their last GET) could send a payload that would pass after pruning but gets rejected by the raw-size check first. Add a sanity cap on raw payload size (2 * MAX_SKILL_STATES) so abusive inputs do not reach the DB query, and enforce the real cap on the pruned result instead. Harden pruneOrphans: the earlier early-return path could pass non-ObjectId keys through unchanged. Now only valid ObjectIds are returned, and the Skill-model-unavailable fallback filters by format. Also add isValidObjectIdString validation at the input boundary so malformed (but otherwise non-Mongo-unsafe) keys never reach persistence (Codex P2 x2). * fix: enforce active filter at execute time, prune revoked shares, scope queue per user P1: injectSkillCatalog now returns activeSkillIds (the filtered set that appears in the catalog). initializeAgent uses that set as the stored accessibleSkillIds on the initialized agent, so getSkillByName at runtime cannot resolve a deactivated skill — even if the LLM hallucinates a name or the user invokes by direct-invocation shorthand. Previously the executor authorized against the full ACL set, bypassing the active-state guarantee (Codex P1). P2: pruneOrphans now checks user access via findAccessibleResources in addition to skill existence. When a share is revoked, the user's skillStates entry for that skill had no cleanup path and silently consumed the 200-cap. Self-heals on both GET and POST. One extra ACL query per settings read/write; scoped to a single user so no N-user amplification (Codex P2). P2: the write queue moves from a single module-scoped object to a Map keyed by userId. Logout/login in the same tab can no longer flush the previous user's pending snapshot under the new session's auth. Each userId gets its own pending/inFlight slot; the in-flight request retains its original auth via the cookie already attached when sent, so the race window closes (Codex P2). * refactor: extract skillStates helpers to packages/api; add tests; polish Address the remaining valid findings from the comprehensive review: - Extract toRecord, loadSkillStates, validateSkillStatesPayload, and pruneOrphanSkillStates into packages/api/src/skills/skillStates.ts as TypeScript. The controller in /api shrinks to a ~90-line thin wrapper that builds live dependency adapters for Mongoose + the permission service (Review #2 DRY, #3 workspace boundary) - Replace the triplicated 12-line skillStates loading block in initialize.js, openai.js, and responses.js with a single call to loadSkillStates from @librechat/api. One helper, three sites - Swap console.error for the project logger in the controller (Review #7) - Remove the redundant INVALID_KEY_PATTERN regex: a valid ObjectId cannot contain . or $, so isValidObjectIdString already covers it (Review #11) - Parameterize the 200-cap error toast with {{0}} interpolation driven by the error response's `limit` field, so future changes to MAX_SKILL_STATES update the UI message automatically (Review #12) - Add 24 unit tests for the new skillStates helpers (toRecord, resolveDefaultActiveOnShare, loadSkillStates, validateSkillStates- Payload, pruneOrphanSkillStates) covering success paths, malformed input, cap boundaries, and parallel-query behavior (Review #4) - Add 10 tests for injectSkillCatalog pagination covering empty accessible set, missing listSkillsByAccess, single-page filter, owned-vs-shared defaults, explicit-override precedence, multi-page collection, MAX_CATALOG_PAGES safety cap, early termination on has_more=false, additional_instructions injection, and fail-closed without userId (Review #5) Total test count: 60 (was 26 on this surface). * fix: rename skillStates ValidationError to avoid barrel-export collision packages/api/src/types/error.ts already exports a ValidationError (MongooseError extension). Re-exporting a different shape from skills/skillStates.ts through the skills barrel caused TS2308 in CI because the root index re-exports both. Rename to SkillStatesValidationError to keep the exports disjoint. * refactor: tighten tests and absorb caller guard into loadSkillStates Address the followup review findings: - Add optional `accessibleSkillIds` param to loadSkillStates so the helper short-circuits to defaults when no skills are accessible. All three controllers drop the residual 7-line conditional wrapper in favor of a single destructured call (Review #2) - Remove the unreachable `typeof key !== 'string'` check from validateSkillStatesPayload: Object.entries always yields string keys per the JS spec (Review #3) - Replace the two `as unknown as` agent casts in the injectSkillCatalog tests with a `makeAgent()` factory typed directly as the function's parameter shape (Review #4) - Tighten the MAX_CATALOG_PAGES assertion from `toBeLessThanOrEqual(11)` to `toHaveBeenCalledTimes(10)` — the loop deterministically makes exactly 10 page fetches before hitting the cap (Review #1) - Rewrite the parallel-execution test for pruneOrphanSkillStates using deferred promises instead of microtask-order assertions. The test now inspects `toHaveBeenCalledTimes(1)` on both mocks after a single Promise.resolve() yield, pinning Promise.all usage without relying on push-order into a shared array (Review #5) - Evict stale writeQueue entries on user change via a module-scoped `lastSeenUserId` sentinel. When a different user's toggle is the first one after a logout/login, the previous user's queue entry is deleted. Keeps the Map bounded without adding hook-instance effect cleanup (Review #6) * fix(test): mock loadSkillStates in openai and responses controller specs The prior refactor replaced the inline 12-line skillStates loading block with a call to loadSkillStates from @librechat/api. Both controller spec files mock @librechat/api as a flat object, so any new named import from that package is undefined in the test env. Calling `await loadSkillStates(...)` threw before recordCollectedUsage ran, surfacing as "undefined is not iterable" on the test's array destructure of `mockRecordCollectedUsage.mock.calls[0]`. Add the missing mock to both spec files alongside the existing scopeSkillIds stub. * fix: abandon stale skillStates write queues on user switch Close the cross-session leak window where an in-flight flush loop still holds a reference to a previous user's queue: it could fire its next mutateAsync under the new session's auth cookies and persist the stale snapshot to the new user's document (Codex P1). Add an `abandoned` flag on `WriteQueue`. Three mechanisms cooperate: - `getWriteQueue` marks every non-active queue abandoned when the user differs from the last-seen identity (pre-existing eviction site, now more aggressive). - A `useEffect` on `userId` calls the same abandonment pass on every render with a new active identity, covering the window between logout/login and the new user's first toggle (when `getWriteQueue` would otherwise not fire). - The flush loop checks `!queue.abandoned` in its while condition so the second and later iterations exit without firing another `mutateAsync` after the session changes. The first iteration's in-flight request (already dispatched under the original user's cookies) still runs to completion or failure on its own — only the subsequent iterations, which are the dangerous ones, are blocked. |
||
|
Danny Avila
|
3e064c2f2b |
🎯 feat: Per-Agent Skill Selection in Builder and Runtime Scoping (#12689)
* feat: per-agent skill selection in builder and runtime scoping
Wire skills persistence on the Agent model and enable the skills
section in the agents builder panel. At runtime, scope the skill
catalog to only the skills configured on each agent (intersected
with user ACL). When no skills are configured, the full user catalog
is used as the default. The ephemeral chat toggle overrides per-agent
scoping to provide the full catalog.
* fix: add scopeSkillIds to @librechat/api mock in responses unit test
The test mocks @librechat/api but was missing the newly imported
scopeSkillIds, causing createResponse to throw before reaching the
assertions. Added a passthrough mock that returns the input array.
* fix: scope primeInvokedSkills by agent's configured skills
primeInvokedSkills was receiving the full unscoped accessibleSkillIds,
bypassing the per-agent skill scoping applied to initializeAgent. This
allowed previously invoked skills from message history to be resolved
and primed even when excluded from the agent's configured skill set.
Apply the same scopeSkillIds filtering to match the initializeAgent
calls, so skill resolution is consistent across catalog injection
and history priming.
* fix: preserve agent skills through form reset and union prime scope
Two related bugs in the per-agent skill selection flow:
1. resetAgentForm dropped the persisted skills array because the generic
fall-through at the end of the loop excludes object/array values.
Combined with composeAgentUpdatePayload always emitting skills, this
caused any save of a previously-configured agent to silently overwrite
skills with an empty array. Add an explicit case for skills mirroring
the agent_ids handling.
2. primeInvokedSkills processes the full conversation payload, including
prior handoff-agent invocations. Scoping it to only primaryAgent.skills
meant a skill invoked by a handoff agent in a prior turn could not be
resolved when the current primary agent had a different scope, leaving
message history reconstruction incomplete. Union the per-agent scoped
accessibleSkillIds across primary plus all loaded handoff agents so
any skill any active agent could invoke is resolvable from history.
* fix: mark inline skill removals as dirty
The inline X button on the skills list called setValue without
shouldDirty: true, so removing a skill via this control did not
mark the skills field as dirty in react-hook-form state. When a
user removed a skill with the X button and also staged an avatar
upload in the same save, isAvatarUploadOnlyDirty returned true and
onSubmit short-circuited to avatar-only upload, silently dropping
the PATCH that would persist the skill removal.
The dialog path (SkillSelectDialog) already passes shouldDirty: true
on add/remove; this aligns the inline control with that behavior.
* fix: restore full ACL scope for primeInvokedSkills history reconstruction
Reverting the earlier scoping of primeInvokedSkills to the active-agent
union. That change conflated runtime invocation scoping (which correctly
gates what the model can call now) with history reconstruction (which
restores bodies the model already saw in prior turns).
Per-agent scoping still applies at:
- Catalog injection (injectSkillCatalog via initializeAgent)
- Runtime invocation (handleSkillToolCall via enrichWithSkillConfigurable,
using each agent's scoped accessibleSkillIds in agentToolContexts)
History priming is a read of past context, not a grant of new capability.
Scoping it causes historical skill bodies to vanish from formatAgentMessages
when an agent's skills list is edited mid-conversation or when the ephemeral
toggle flips, which breaks message reconstruction and drops code-env file
continuity for /mnt/data/{skillName}/ references. The user's ACL-accessible
set is the correct and sufficient gate for history reconstruction.
* fix: close openai.js skill gap and pin undefined vs [] semantics
Three related gaps surfaced in review:
1. api/server/controllers/agents/openai.js was a third skill resolution
site alongside responses.js and initialize.js, but still used the old
activation gate (required ephemeralAgent.skills === true) and never
passed accessibleSkillIds through scopeSkillIds. Per-agent scoping
silently did not apply on this route. Mirror the same pattern used
in responses.js so all three routes behave identically.
2. scopeSkillIds previously collapsed undefined and [] into the same
"full catalog" fallback, making it impossible for a user to express
"this agent has no skills." Tighten the semantics before any data
is written under the old behavior:
- undefined / null = not configured, full catalog
- [] = explicitly none, returns []
- non-empty = intersection with ACL-accessible set
Update defaultAgentFormValues.skills from [] to undefined so a brand
new agent whose skills UI was never touched does not accidentally
persist "explicit none" on first save (removeNullishValues strips
undefined from the payload server side).
3. Add direct unit tests for scopeSkillIds covering all five cases
(undefined, null, empty, disjoint, overlap, exact match, empty
accessible set). 16 tests total in skills.test.ts pass.
* fix: add scopeSkillIds to @librechat/api mock in openai unit test
Same pattern as the earlier responses.unit.spec.js fix: the test mocks
@librechat/api with an explicit object, so each newly imported symbol
must be added to the mock. Without scopeSkillIds, OpenAIChatCompletion
controller throws on destructuring before reaching recordCollectedUsage,
causing the token usage assertions to fail.
|
||
|
Danny Avila
|
9b4ae068b2 |
💲 feat: Manual Skill Invocation via $ Command Popover (UI only) (#12690)
* feat: add $ command popover for manual skill invocation
Adds a new `$` command trigger in the chat textarea that opens a
searchable popover listing user-invocable skills. Selecting a skill
inserts `$skill-name` into the message and enables the skills badge
on the ephemeral agent. Follows the same patterns as `@` mentions
and `/` prompts.
* test: update useHandleKeyUp tests for $ skill command
Add showSkillsPopoverFamily, dollarCommand, and SKILLS permission
to the store/recoil/access mocks. Include test cases for the $
trigger, toggle gating, and permission gating.
* fix: address review findings in SkillsCommand
- Exclude auto-mode skills from popover (isUserInvocable now
returns false for InvocationMode.auto)
- Rewrite JSDoc to match the corrected filter logic
- Guard ArrowUp/ArrowDown against NaN when matches is empty
- Replace useRecoilState with useSetRecoilState for ephemeralAgent
to avoid subscribing to unrelated agent state changes
- Move skills guard into setter callback and drop ephemeralAgent
from handleSelect dependency array
- Add e.preventDefault() for Tab key alongside Enter
- Render error state (com_ui_skills_load_error) on query failure
- Render empty state (com_ui_skills_empty) when no matches
- Hoist ScrollText icon to module-level constant
- Export isUserInvocable for testability
* fix: address follow-up review findings
- Differentiate empty-catalog ("No skills yet") from no-match
search ("No skills found") using existing com_ui_no_skills_found
- Clamp activeIndex when matches shrink from search filtering to
prevent silent Enter/Tab failures
- Add early return after Escape handler to skip redundant checks
- Reorder package imports shortest-to-longest after react
- Add fast-typing test case for $ command ("$sk" at position 3)
* fix: gate $ command on assistants endpoint, fix Tab on empty matches
- Block $ popover on assistants/azureAssistants endpoints where
ephemeralAgent is not sent in the submission payload
- Allow Tab to close the popover when matches is empty instead of
trapping keyboard focus
- Add endpoint gating tests for $ on both assistants endpoints
* fix: reset skills popover on assistants switch + paginate skills query
- Close $ skills popover when endpoint switches to assistants or
azureAssistants, mirroring the existing + command reset
- Switch SkillsCommand from a single-page list query to the
cursor-paginated useSkillsInfiniteQuery and auto-fetch all pages
so client-side search covers the full catalog instead of only
the first 100 entries
- Show the spinner during background page fetches and suppress the
empty-state copy until paging completes
- Add test for popover reset on endpoint switch
* fix: prevent currency hijack and form submit on $ command
- Reject the $ trigger when fast-typed text after $ does not start
with a lowercase letter, so currency input like $100 or $5.99
no longer opens the skills popover and clears the textarea
- preventDefault() on Enter when the popover has no matches so
the surrounding chat form does not submit when the user dismisses
the popover via Enter
- Add tests for $100 and $5.99 currency inputs
* fix: defer $ popover to second keystroke and stop pagination on errors
- Defer opening the skills popover until a letter follows the $
character. Bare $ no longer triggers the popover, so starting a
message with $100, $5.99, or $EUR is fully preserved (the
textarea is not cleared by useInitPopoverInput on the first
keystroke). $a, $skill, $my-skill still open as expected.
- Add a sticky paginationBlockedRef circuit breaker on the auto
fetchNextPage effect so a transient page request error cannot
spin into an unbounded retry loop when isError flips back to
false on the next attempt.
- Update tests: bare $ no longer triggers, $a does, currency cases
remain blocked.
* feat: scaffold structured manual-skill channel for follow-up PR
Add a per-conversation pendingManualSkillsByConvoId atom family
that SkillsCommand appends to on selection. This is the writer
half of the structured channel that will let a follow-up PR
deterministically prime SKILL.md as a meta user message before
the LLM turn (mirroring Claude Code's `/skill` invocation), so
the backend never has to regex-parse `$name` out of user text.
The submit pipeline does not yet read this atom; the textual
`$skill-name ` insertion remains the authoritative signal until
the follow-up wires the read + reset on submit. Reset is already
wired into useClearStates so the atom does not leak across
conversation switches.
* test: cover SkillsCommand selection-flow contract
Add a component test that locks in the contract the follow-up
manualSkills PR has to honor when a user picks a skill in the $
popover:
- Pushes the skill name onto pendingManualSkillsByConvoId (the
per-conversation structured channel), with dedup
- Flips ephemeralAgent.skills to true via the callback-form setter
- Inserts $skill-name into the textarea as cosmetic confirmation
- Closes the popover via setShowSkillsPopover(false)
- Renders nothing when the popover atom is false
Mocks recoil setters and the skills query so the test exercises
the real component logic without spinning up the full provider
stack.
* revert: drop $ defer-until-letter guard for sibling consistency
Bring $ in line with @, /, and +: open the popover on the bare
trigger character. The earlier defer guard kept currency input
like \$100 from clobbering the textarea, but it broke the natural
"type \$ to browse skills" UX and was inconsistent with the other
trigger chars, none of which gate on the follow-up character
(/path/to/file, @username, +1 all clear the textarea the same way).
The currency hijack remains recoverable via Escape.
Drop the corresponding paste-protection tests for \$100 and \$5.99,
restore the bare-\$ trigger test.
|
||
|
Danny Avila
|
3b820415ad |
🎭 feat: Custom UI Renderers for Skill Tool Calls (#12684)
* feat: Custom UI renderers for skill, read_file, and bash_tool Add specialized tool call components for the three skill tools, replacing the generic ToolCall fallback with contextual UI. * fix: Address review findings for skill tool UI renderers - Fix Codex P2: read skillName (camelCase) matching agent pipeline - Fix Codex P2: remove error regex from ReadFileCall to avoid false positives on normal file content containing "Error:" tokens - Extract useToolCallState hook to eliminate ~60% boilerplate duplication across SkillCall, ReadFileCall, and BashCall - Extract parseJsonField utility with consistent escaped-char-aware regex fallback, shared by all three components - Gate SkillCall bordered card on hasOutput to prevent empty card when expanded before output arrives - Skip highlightAuto for plaintext lang to avoid expensive auto-detection on files with unknown extensions - Expand LANG_MAP with php, cs, kt, swift, scss, less, lua, r; add FILENAME_MAP for Makefile and Dockerfile - Export langFromPath for testability - Add unit tests for parseJsonField, langFromPath, and ToolIcon skill type branches * refactor: Redesign BashCall as minimal terminal widget Replace the ExecuteCode-clone pattern with a purpose-built terminal UI: $ prompt prefix, dark background command zone, icon-only copy button, and raw monospace output. Drops useLazyHighlight, CodeWindowHeader, Stdout, and the "Output" label in favor of a cleaner two-zone layout that feels native to the terminal. * fix: parseJsonField unescape ordering and ReadFileCall empty card Replace the sequential .replace() chain in parseJsonField's regex fallback with a single-pass /\(.)/g replacement. The old chain processed \n before \, so \n (JSON-escaped literal backslash + n) was incorrectly decoded as a newline instead of \n. Gate ReadFileCall's bordered card on hasOutput (matching SkillCall's pattern) so the card does not render as an empty rounded box during streaming before output arrives. Add regression tests for \n decoding and unknown escape sequences. * fix: Followup review fixes - Refactor ExecuteCode to use shared useToolCallState hook, eliminating the last copy of the inline state machine - Escape regex metacharacters in parseJsonField to prevent injection from field names containing ., +, (, etc. - Fix contradictory test description in langFromPath tests * fix: Surface tool failure state in skill tool renderers Add error detection to useToolCallState via the shared isError check so tool calls that complete with an error prefix show a "failed" suffix instead of a success label. Prevents misleading users when read_file, skill, or bash_tool returns an error (e.g. file not found, skill not accessible). Matches the error handling pattern already used by the generic ToolCall component. * feat: Add bash syntax highlighting to BashCall command zone Reuse the shared useLazyHighlight singleton (already loaded by ReadFileCall and ExecuteCode) to highlight the command with bash grammar. Falls back to plain text while lowlight is loading. * fix: Align BashCall scrollbar to span full card width Move max-h/overflow-auto from the inner pre to the outer container so the scrollbar spans the full width like the output zone. Float the copy button with sticky positioning so it stays visible while scrolling long commands. * feat: Use GNU Bash icon for bash_tool progress header and ToolIcon Replace the generic SquareTerminal lucide icon with the GNU Bash logo (already in the project via LangIcon/langIconPaths) for both the BashCall progress header and the ToolIcon stacked icon mapping. * fix: Render raw content while highlighter loads, preserve command text on copy - ReadFileCall: fall back to raw output when useLazyHighlight returns null, preventing a blank code panel on first render before lowlight finishes its dynamic import - BashCall: drop .trim() from the copy handler so the clipboard receives exactly what's displayed (WYSIWYG copy) * fix: Alphabetize new translation keys within en/translation.json Relocate read_file, skill_finished, and skill_running into their correct alphabetical positions within the overall key list. * fix: Surface error state in ExecuteCode, fix BashCall import order - ExecuteCode now uses hasError from useToolCallState to show the "failed" suffix on failed code executions, matching the three new renderers - Reorder BashCall local imports to longest-to-shortest per project style |
||
|
Danny Avila
|
64ec5f18b8 |
⚙️ feat: Skill runtime integration: catalog, tools, execution, file priming (#12649)
* feat: Skill runtime integration — catalog injection, tool registration, execute handler
Wires the @librechat/agents SkillTool primitive into LibreChat's agent runtime:
**Enums:**
- Add `skills` to AgentCapabilities + defaultAgentCapabilities
**Data layer:**
- Add `getSkillByName(name, accessibleIds)` — compound query that
combines name lookup + ACL check in one findOne
**Agent initialization (packages/api/src/agents/initialize.ts):**
- Accept `accessibleSkillIds` param and `listSkillsByAccess` db method
- Query accessible skills, format catalog via `formatSkillCatalog()`,
append to `additional_instructions` (appears in agent system prompt)
- Register `SkillToolDefinition` + `createSkillTool()` when catalog
is non-empty (tool appears in model's tool list)
- Store `accessibleSkillIds` and `skillCount` on InitializedAgent
**Execute handler (packages/api/src/agents/handlers.ts):**
- Add `getSkillByName` to `ToolExecuteOptions`
- `handleSkillToolCall()` intercepts `Constants.SKILL_TOOL`:
extracts skillName, loads body from DB with ACL check,
substitutes $ARGUMENTS, returns ToolExecuteResult with
injectedMessages (skill body as isMeta user message)
**Caller wiring:**
- initialize.js: query skill IDs via findAccessibleResources,
pass to initializeAgent + store on agentToolContexts,
add getSkillByName to toolExecuteOptions,
pass accessibleSkillIds through loadTools configurable
- openai.js + responses.js: same pattern for their flows
Requires @librechat/agents >= 3.1.65 (PR #91 exports).
* feat: Skills toggle in tools menu + backend capability gating
Frontend:
- Add skills?: boolean to TEphemeralAgent type
- Add LAST_SKILLS_TOGGLE_ to LocalStorageKeys for persistence
- Add skillsEnabled to useAgentCapabilities hook
- Add skills useToolToggle to BadgeRowContext with localStorage init
- New Skills.tsx badge component (Scroll icon, cyan theme,
permission-gated via PermissionTypes.SKILLS)
- Add skills entry to ToolsDropdown with toggle + pin
- Render Skills badge in BadgeRow ephemeral section
Backend:
- Extract injectSkillCatalog() into packages/api/src/agents/skills.ts
(reduces initializeAgent module size, reusable helper)
- initializeAgent delegates to helper instead of inline block
- Capability-gate the findAccessibleResources query:
- Agents endpoint: checks AgentCapabilities.skills in admin config
- OpenAI/Responses controllers: checks ephemeralAgent.skills toggle
- ACL query runs once per run, result shared across all agents
* refactor: remove createSkillTool() instance from injectSkillCatalog
SkillTool is event-driven only. The tool definition in toolDefinitions
is sufficient for the LLM to see the tool schema. No tool instance is
needed since the host handler intercepts via ON_TOOL_EXECUTE before
tool.invoke() is ever called.
Removes tools from InjectSkillCatalogParams/Result, drops the
createSkillTool import.
* feat: skill file priming, bash tool, and invoked skills state
Multi-file skill support:
- New primeSkillFiles() helper (packages/api/src/agents/skillFiles.ts)
uploads skill files + SKILL.md body to code execution environment
- handleSkillToolCall primes files on invocation when skill.fileCount > 0,
returns session info as artifact so ToolNode stores the session
- Skill-primed files available to subsequent bash/code tool calls
Bash tool auto-registration:
- BashExecutionToolDefinition added alongside SkillToolDefinition when
skills are enabled, giving the model a bash tool for running scripts
Conversation state:
- Add invokedSkillIds field to conversation schema (Mongoose + Zod)
- handleSkillToolCall updates conversation with $addToSet on success
- Enables re-priming skill files on subsequent runs (future)
Dependency wiring:
- Pass listSkillFiles, getStrategyFunctions, uploadCodeEnvFile,
updateConversation through ToolExecuteOptions
- Pass req and codeApiKey through mergedConfigurable
- All three controller entry points wired (initialize.js, openai.js,
responses.js)
* fix: load bash_tool instance in loadToolsForExecution, remove file listing
- Add createBashExecutionTool to loadToolsForExecution alongside PTC/ToolSearch
pattern: loads CODE_API_KEY, creates bash tool instance on demand
- Add BASH_TOOL and SKILL_TOOL to specialToolNames set so they don't go
through the generic loadTools path (bash is created here, skill is
intercepted in handler before tool.invoke)
- Remove file name listing from skill content text — it's the skill
author's responsibility to disclose files in SKILL.md, not the framework
* feat: batch upload for skill files, replace sequential uploads
- Add batchUploadCodeEnvFiles() to crud.js: single POST to /upload/batch
with all files in one multipart request, returns shared session_id
- Rewrite primeSkillFiles to collect all streams (SKILL.md + bundled files)
then do one batch upload instead of N sequential uploads
- Replace uploadCodeEnvFile with batchUploadCodeEnvFiles across all callers
(handlers.ts, initialize.js, openai.js, responses.js)
* refactor: remove invokedSkillIds from conversation schema
Skills aren't re-loaded between runs, so conversation-level state for
invoked skills doesn't help. Skill state will live on messages instead
(like tool_search discoveredTools and summaries), enabling in-place
re-injection on follow-up runs.
Removes invokedSkillIds from: convo Mongoose schema, IConversation
interface, Zod schema, ToolExecuteOptions.updateConversation, and
all three caller wiring points.
* feat: smart skill file re-priming with session freshness checking
Schema:
- Add codeEnvIdentifier field to ISkillFile (type + Mongoose schema)
- Add updateSkillFileCodeEnvIds batch method (uses tenantSafeBulkWrite)
- Export checkIfActive from Code/process.js
Extraction:
- Add extractInvokedSkillsFromHistory() to run.ts — scans message
history for AIMessage tool_calls where name === 'skill', extracts
skillName args. Follows same pattern as extractDiscoveredToolsFromHistory.
Smart re-priming in primeSkillFiles:
- Before batch uploading, checks if existing codeEnvIdentifiers are
still active via getSessionInfo + checkIfActive (23h threshold)
- If session is still active, returns cached references (zero uploads)
- If stale or missing, batch-uploads everything and persists new
identifiers on SkillFile documents (fire-and-forget)
- Single session check covers all files (batch shares one session_id)
Wiring:
- Pass getSessionInfo, checkIfActive, updateSkillFileCodeEnvIds
through ToolExecuteOptions and all three controller entry points
* feat: wire skill file re-priming at run start via initialSessions
Flow:
1. initialize.js creates primeInvokedSkills callback with all deps
2. client.js calls it with message history before createRun
3. extractInvokedSkillsFromHistory scans for skill tool calls
4. For each invoked skill with files, primeSkillFiles uploads/checks
5. Returns initialSessions map passed to createRun
6. createRun passes initialSessions to Run.create (via RunConfig)
7. Run constructor seeds Graph.sessions, making skill files available
to subsequent bash/code tool calls via ToolNode session injection
Requires @librechat/agents with initialSessions on RunConfig (PR #94).
* refactor: use CODE_EXECUTION_TOOLS set for code tool checks
Import CODE_EXECUTION_TOOLS from @librechat/agents and replace inline
constant checks in handlers.ts and callbacks.js. Fixes missing bash
tool coverage in the session context injection (handlers.ts) and code
output processing (callbacks.js).
* refactor: move primeInvokedSkills to packages/api, add skill body re-injection
Moves primeInvokedSkills from an inline closure in initialize.js (with
dynamic requires) to a proper exported function in packages/api
skillFiles.ts with explicit typed dependencies.
Key changes:
- primeInvokedSkills now returns both initialSessions (for file priming)
AND injectedMessages (skill bodies for context continuity)
- createRun accepts invokedSkillMessages and appends skill bodies to
systemContent so the model retains skill instructions across runs
- initialize.js calls the packaged function with all deps passed explicitly
- client.js passes both initialSessions and injectedMessages to createRun
* fix: move dynamic requires to top-level module imports
Move primeInvokedSkills, getStrategyFunctions, batchUploadCodeEnvFiles,
getSessionInfo, and checkIfActive from inline requires to top-level
module requires where they belong.
* refactor: skill body reconstruction via formatAgentMessages, not systemContent
Replaces the lazy systemContent approach with proper message-level
reconstruction:
SDK (formatAgentMessages):
- New invokedSkillBodies param (Map<string, string>)
- Reconstructs HumanMessages after skill ToolMessages at the correct
position in the message sequence, matching where ToolNode originally
injected them
LibreChat:
- extractInvokedSkillsFromPayload replaces extractInvokedSkillsFromHistory
(works with raw TPayload before formatAgentMessages, not BaseMessage[])
- primeInvokedSkills now takes payload instead of messages, returns
skillBodies Map instead of injectedMessages
- client.js calls primeInvokedSkills BEFORE formatAgentMessages, passes
skillBodies through as the 4th param
- Removed invokedSkillMessages from createRun (no more systemContent hack)
- Single-pass: skill detection happens inside formatAgentMessages' existing
tool_call processing loop, zero extra message iterations
* refactor: rename skillBodies to skills for consistency with SDK param
* refactor: move auth loading into primeInvokedSkills, pass loadAuthValues as dep
The payload/accessibleSkillIds guard and CODE_API_KEY loading now live
inside primeInvokedSkills (packages/api) rather than in the CJS caller.
initialize.js passes loadAuthValues as a dependency and the callback
is only created when skillsCapabilityEnabled.
* feat: ReadFile tool + conditional bash registration + skill path namespacing
ReadFile tool (read_file):
- General-purpose file reader, event-driven (ON_TOOL_EXECUTE)
- Schema: { file_path: string } — "{skillName}/{path}" convention
- handleReadFileCall: resolves skill name from path, ACL check, reads
from DB cache or storage, binary detection, size limits (256KB),
lazy caching (512KB), line numbers in output
- SKILL.md special case: reads skill.body directly
- Dispatched alongside SKILL_TOOL in createToolExecuteHandler
- Added to specialToolNames in ToolService
Conditional tool registration:
- ReadFile + SkillTool: always registered when skills enabled
- BashTool: only registered when codeEnvAvailable === true
- codeEnvAvailable passed through InitializeAgentParams from caller
Skill file path namespacing:
- primeSkillFiles now uploads as "{skillName}/SKILL.md" and
"{skillName}/{relativePath}" instead of flat names
- Prevents file collisions when multiple skills are invoked
Wiring:
- getSkillFileByPath + updateSkillFileContent passed through
ToolExecuteOptions in all three callers
* feat: return images/PDFs as artifacts from read_file, tighten caching
Binary artifact support:
- Images (png, jpeg, gif, webp) returned as base64 in artifact.content
with type: 'image_url', processed by existing callback attachment flow
- PDFs returned as base64 artifact similarly
- Binary size limit: 10MB (MAX_BINARY_BYTES)
- Other binary files still return metadata + bash fallback
Caching:
- Text cached only on first read (file.content == null check)
- Binary flag cached only on first detection (file.isBinary == null)
- Skill files are immutable; no redundant cache writes
Registration:
- ReadFileToolDefinition now includes responseFormat: 'content_and_artifact'
* chore: update @librechat/agents to version 3.1.66-dev.0 and add peer dependencies in package-lock.json and package.json files
* fix: resolve review findings #1,#2,#4,#5,#6,#10,#13
Critical:
- #1: primeInvokedSkills now accumulates files across all skills into
one session entry instead of overwriting. Parallel processing via
Promise.allSettled.
- #2: codeEnvAvailable now computed and passed in openai.js and
responses.js (was missing, bash tool never registered in those flows)
Major:
- #4: relativePath in updateSkillFileCodeEnvIds now strips the
{skillName}/ prefix to match SkillFile documents. SKILL.md filter
uses endsWith instead of exact match.
- #5: File priming guarded on apiKey being non-empty (skip when not
configured instead of failing with auth error)
- #6: Skills processed in parallel via Promise.allSettled instead of
sequential for-of loop
Minor:
- #10: Use top-level imports in initialize.js instead of inline requires
- #13: Log warning when skill catalog reaches the 100-skill limit
* fix: resolve followup review findings N1,N2,N4
N1 (CRITICAL): Wire skill deps into responses.js non-streaming path.
Was completely missing getSkillByName, file strategy functions, etc.
N2 (MAJOR): Single batch upload for ALL skills' files. Resolves skills
in parallel (Phase 1), then collects all file streams across skills
and does ONE batchUploadCodeEnvFiles call (Phase 2). All files share
one session_id, eliminating cross-session isolation issues.
N4 (MINOR): Move inline require() to top-level in openai.js and
responses.js, consistent with initialize.js.
* fix: add mocks for new file strategy imports in controller tests
* fix: restore session freshness check, parallelize file lookups, add warnings
R1: Re-add session freshness check before batch upload. Checks any
existing codeEnvIdentifier via getSessionInfo + checkIfActive. If the
session is still active (23h window), returns cached file references
with zero re-uploads.
R2: listSkillFiles calls parallelized via Promise.all (were sequential
in the for-of loop).
R3: Log warning when skill record lookup fails during identifier
persistence (was a silent empty-string fallback).
* fix: guard freshness cache on single-session consistency
* fix: multi-session freshness check (code env handles mixed sessions natively)
The code execution environment fetches each file by its own
{session_id, fileId} pair independently — no single-session
requirement. Removed the sessionIds.size === 1 guard.
Now checks ALL distinct sessions for freshness. If every session
is still active (23h window), returns cached references with per-file
session_ids preserved. If any session expired, falls through to
re-upload everything in a single batch.
* perf: parallelize session freshness checks via Promise.all
* fix: add optional chaining for session info retrieval in primeInvokedSkills
Updated the primeInvokedSkills function to use optional chaining for getSessionInfo and checkIfActive methods, ensuring safer access and preventing potential runtime errors when these methods are undefined.
* fix: address review findings #1-#9 + Codex P1/P2 + session probe
Critical:
- #1/Codex P1: Add codeApiKey loading to openai.js and responses.js
loadTools configurable (was missing, file priming broken in 2/3 paths)
- Codex P1: Fix cached file name prefix in primeSkillFiles cache path
(was sf.relativePath, now ${skill.name}/${sf.relativePath})
Major:
- Codex P2: Honor ephemeral skills toggle in agents endpoint
(check ephemeralAgent?.skills !== false alongside admin capability)
- #4: Early size check using file.bytes from DB before streaming
(prevents full-file buffer for oversized files)
Minor:
- #5: Replace Record<string, any> with Record<string, boolean | string>
- #6: Localize Pin/Unpin aria-labels with com_ui_pin/com_ui_unpin
- #8: Parallelize stream acquisition in primeSkillFiles via
Promise.allSettled
- #9: Log warning for partial batch upload failures with filenames
Performance:
- Session probe optimization: getSessionInfo now hits per-object
endpoint (GET /sessions/{sid}/objects/{fid}) instead of listing
entire session (GET /files/{sid}?detail=summary). O(1) stat vs
O(N) list + linear scan.
* refactor: extract shared skill wiring helper + add unit tests
DRY (#3):
- New skillDeps.js exports getSkillToolDeps() with all 9 skill-related
deps (getSkillByName, listSkillFiles, getStrategyFunctions, etc.)
- Replaces 5 identical copy-paste blocks across initialize.js, openai.js,
responses.js (streaming + non-streaming paths)
- One place to maintain when skill deps change
Tests (#2):
- 8 unit tests for extractInvokedSkillsFromPayload covering:
string args, object args, missing skill tool_calls, non-assistant
messages, malformed JSON, empty skillName, empty payload, dedup
* fix: remove @jest/globals import, use global jest env
* fix: resolve round 2 review findings R2-1 through R2-7
R2-1 (toggle semantics): openai.js + responses.js now check admin
capability (AgentCapabilities.skills) alongside ephemeral toggle.
Aligns with initialize.js.
R2-2 (swallowed error): primeInvokedSkills now logs
updateSkillFileCodeEnvIds failures (was .catch(() => {}))
R2-4 (test cast): Record<string, string> → Record<string, unknown>
R2-5 (DRY regression): Extract enrichWithSkillConfigurable() into
skillDeps.js. Replaces 4 identical loadAuthValues blocks.
Each loadTools callback is now a one-liner. JSDoc added (R2-6).
R2-7 (sequential streams): primeInvokedSkills now uses
Promise.allSettled for parallel stream acquisition.
* fix: require explicit skills toggle + treat partial cache as miss
- initialize.js: change ephemeralSkillsToggle !== false to === true
(unset toggle no longer enables skills)
- primeSkillFiles cache: require ALL files to have codeEnvIdentifier
before using cache (partial persistence = cache miss = re-upload)
- primeInvokedSkills cache: same check (allFilesWithIds.length must
equal total file count)
* fix: pass entity_id=skillId on batch upload, eliminates per-user cache thrashing
primeSkillFiles now passes entity_id: skill._id.toString() to
batchUploadCodeEnvFiles. This scopes the code env session to the
skill, not the user. All users sharing a skill share the same
uploaded files — no more cache thrashing from overwriting each
other's codeEnvIdentifier.
The stored codeEnvIdentifier now includes ?entity_id= suffix so
freshness checks pass the entity_id through to the per-object
stat endpoint. Both primeSkillFiles and primeInvokedSkills
store consistent identifier formats.
* fix: pass entity_id on multi-skill batch upload, consistent identifier format
* Revert "fix: pass entity_id on multi-skill batch upload, consistent identifier format"
This reverts commit
|