* feat: Show the Conversation Pull Request in the Sidebar List
Add a batch lookup route and a sidebar row mark: the state icon with the CI dot, opening the same card beside the row on hover. The row keeps the conversation's own title.
* fix: Address Review Findings on the Sidebar Pull Request Batch and Row Mark
Cap the whole batch with a configurable deadline, make the row mark a keyboard-reachable button that opens the card on focus, refresh rows when the window regains focus through the shared batch, and show a retryable failure state instead of treating a failed first lookup as no pull request.
* fix: Send Sidebar Pull Request Batches One at a Time
Queue every batch request behind the one before it, so an oversized list or an overlapping flush cannot multiply the server's configured lookup concurrency, and bound how long one request may hold the queue so a request that never answers cannot stop the rest.
* fix: Keep the Batch Request Timeout Above the Server Deadline
Default the client's request timeout to just past the longest batchTimeoutSeconds a deployment may configure, so a queued request never starts while the previous request's server-side lookups are still running.
* fix: Keep Batch Lookups Inside the Concurrency Limit and Per-Entry Failures Per Entry
Share one lookup limiter across every batch request so lookups that outlive a request's deadline keep their slot, answer a missing token only for the conversations that needed it, and point the row at its pull request description only while that text is in the page.
* fix: Bound Batch Reads, Scope Concurrency Per Credential, and Gate the Sidebar on the Batch Version
Start the batch deadline when the request arrives and bound the config and lane reads, count lookup slots per credential, advertise a batch route version and ask only a server that has it, take no row space for an absent mark, show a failed refresh over a cached empty answer, keep a malformed response from stalling the request queue, and repair the lane recording tests that still used a mock removed on dev.
* fix: Move Capability Resolution Into the Package and Harden the Batch Route and Queue
Resolve the pull request startup capabilities in packages/api so the route stays wiring only, bound the config and lane reads by the configured deadline, merge identical lanes before taking limiter slots, fall back to the single route when a replica has no batch route, and dispose the batch queue when the signed-in user changes.
* fix: Abort the Batch Fallback With Its Request So the Queue Never Overlaps It
Give the batch fetcher the request's abort signal and pass it to every single-route call, abort a request that times out or is disposed, and stop the fallback workers from starting more calls once it has.
* fix: Forward the Abort Signal to the Batch POST
Pass the request's abort signal through the batch fetcher, the data-provider call and Axios, so a batch request that times out or is disposed is cancelled instead of continuing beside the next one, and settle a request promptly when it is aborted.
* fix: Close the Pull Request Lookup Follow-Ups and Cancel Abandoned Sidebar Rows
* feat: Show Lia on the Welcome Screen
Adds the opt-in Show Lia setting (off by default) and the interface.mascot deployment flag (allowed by default). Lia mounts only on the landing page, loads lazily after first paint, stands on the composer's top edge clear of the greeting, reacts to typing, files and theme changes, and speaks through localized bubbles. She is decorative and hidden from assistive technology.
* fix: Keep Lia's Long-Message Reaction Re-Armed and Pointer Reads Per Frame
Deleting a message back under 280 characters now re-arms the long-message reaction, pointer handling reads layout at most once per animation frame, and the ChatView specs stub Lia like their other children.
* test: Open Settings from the Phone Drawer in the Lia Scenarios
The opt-in scenario no longer resets the stored preference on reload, and both Settings scenarios open the account menu from the mobile drawer on phones.
* fix: Measure Lia's Caret Like the Composer and Label Her When Idle
The caret mirror copies direction, alignment, indentation, wrapping and box sizing, so right-to-left and wrapped text measure like the composer. Hovering Lia between activities now says she is hanging out instead of showing nothing.
* fix: Wave Off Assistants Sends and Keep Lia's Canvas and Bubble Off Welcome Content
* fix: Keep Lia's Speech Bubble Inside Her Stage
* fix: Wrap Long Translations in Lia's Speech Bubble
* fix: Keep Lia's Long-Message Reaction Until It Plays and Pin the Scenario Greeting
* fix: Make Only Lia's Body Clickable So She Fits Beside Long Greetings
* fix: Let Lia Stand Beside the Composer When the Greeting Fills Its Top
* feat: Let Users Drag Lia Around and Keep Her Near the Composer
* test: Cover Picking Lia Up and Setting Her Down on the Composer
* refactor: Keep Lia's Preference Inside Her Feature
* fix: Let Keyboard Users Pet Lia and Move Her With the Arrow Keys
* test: Shift Only the Wall Clock for Lia's Morning Greeting
* test: Measure Lia's Keyboard Walk Both Ways So the Greeting's Width Cannot Pin Her
* fix: Bring Lia Back When a New Chat Opens During Her Farewell
* fix: Start Lia's Reactions on Her Own Clock After a Pause Without Room
* fix: Load Lia Through Stale-Asset Recovery and Skip the Farewell Before She Appears
* feat: Mark the Lia Settings Toggle as Beta
* refactor: Give Sign-In Controls Button, Input and Alert Variants
* fix: Load the Field Stylesheet From SecretInput
* fix: Default the Floating Fields to the Placeholder Sentinel
* test: Cover the Sign-In Control Roles in the Default Theme
* test: Wait for the Floating Label Transition Before Reading It
* test: Read the Floating Label Scale Property
* fix: Keep Button Heights on Theme Roles and Honor Reduced Motion in Floating Fields
* refactor: Clear the Raw-Color and Unknown-Class Design Suppressions
* fix: Keep the Prose Classes style.css Still Reads
* fix: Keep the split-parent Marker the Screenshot Spec Waits On
* chore: Prune the Recorded Suppressions the Rebase Cleared
* fix: Dim a Disabled InputNumber Through Its Nested Input
* test: Cover the Disabled InputNumber Wrapper in the Browser
* test: Tag the InputNumber Scenarios Literally
* fix: Make DataTable Row Estimates Breakpoint- and Leaf-Column-Aware
* test: Cover the Compact Row Estimate Against the Rendered Cell
* test: Tag the Row Estimate Scenarios Literally
* feat: Show the Conversation Pull Request in the Chat Header
A header chip shows the pull request a code conversation opened: the agent picture with a CI status dot, the pull request icon and a truncated title. It opens a card with the number, added and removed lines, a link to GitHub, and state and checks badges. The chip renders nothing while loading, when there is no pull request, or when the first lookup fails.
* feat: Place the Pull Request Chip Beside the Model Controls and Move It to the Overflow Menu on Small Screens
On desktop the chip sits with the left header controls and its card opens to the right as an Ariakit hovercard that slides in from the left, on hover or keyboard focus. Small screens offer the same card from the header overflow menu through a dialog, following the dialog-item contract.
* fix: Address Review Findings on Pull Request Polling, Badges and Tooltip Layering
Advertise pullRequestsEnabled in the startup config and skip the lookup without it, keep polling a finished pull request while its checks run, render the card badges with the shared Chip, and portal the GitHub tooltip into the card's own layer.
* fix: Address Second Round Review Findings on the Pull Request Dialog and Chip
Scope the mobile dialog to the conversation it opened for and expose dialog semantics on its menu entry, only use the chat agent's avatar for the routed conversation, and make the hover tests move the pointer the way a browser does so they pass under NODE_ENV=development as CI runs them.
* 🏝️ fix: Apply the No-Workspace Decision Wherever an Agent's Code Route Resolves
A conversation set to run without attached workspaces failed mid-turn with
"Choose an attached workspace before using this agent." when its agent's
default machine was attached and the turn loaded a skill or another tool at
execution time. Request-start initialization honoured the decision, but the
execution-time tool loader re-resolved the agent's attached default and then
demanded a workspace selection the conversation intentionally never made.
The per-agent code rule now lives in one typed function,
resolveAgentCodeExecution (packages/api/src/code/agent.ts): it applies the
conversation's no-workspace opt-out, the run-code grant, the stateful
capability and the stateful environment policy, then resolves the base route.
Request-start initialization, lazy subagent metadata and graph members, the
execution-time and definition tool loaders, and the legacy execute_code
loader all go through it, so an opted-out agent lands on the managed default
route with no code tools at every site. Per-call subagent machine choices
offer and accept no attached machine in such a conversation, and a source
invariant test keeps new resolver call sites on the shared rule.
* fix: Pass the Implicit Stateful Route Rollout Into the Shared Code Rule
The per-agent code rule read the implicit stateful route rollout from the environment itself. It now takes it as a parameter, and the request boundaries (initializeAgent and withRequestCodeInputs) resolve it from the deployment configuration, so the rule stays a pure function of its inputs.
* fix: Carry the Initializer's Code Opt-Out Into Tool Definition Loading
initializeAgent hands the tool loaders a partial agent record (id, tools, provider, model) together with the code route it already resolved. Without the agent's route fields the definition loaders could not see a no-workspace opt-out and recomputed code execution as enabled, so tool classification could advertise the programmatic bash tool that execution then refuses. initializeAgent now passes attachedEnvironmentOptOut through every tool loader, and the shared rule keeps an opt-out its caller already decided.
* feat: Quiet Prompt Editor Boxes, Table Columns and Destructive Buttons for a Theme
* fix: Outrank the Cell Border Color With the Column Rule Share
* fix: Keep the First Table Cell's Outer Edge Under the Inset Share
* fix: Keep the Batched Edit Divider on the Light Border
* fix: Keep Web Search Source Row Separators on the Light Border
* fix: Quiet the Result Switcher Divider and Keep the RTL Table Edge
* fix: Resolve Table Edges by Cell Direction and Keep the Variables Card Opaque in Details
* fix: Draw Table Frame Edges in Logical Terms So Direction Needs No Selector
* fix: Keep Soft Destructive Hover and Pressed Tints Above 4.5:1
* feat: Record Lane Git State and Serve Conversation Pull Requests
A conversation now stores the branch, head and repository its code lane last reported, written only by the server after a finished bash command and excluded from ordinary reads, generic saves, unset fields and imports. GET /api/convos/:conversationId/pull-request finds the pull request for that branch through an injected GitHub source with a bounded cache, answers null for an absence, and maps failures to stable codes without upstream text. The feature is off unless endpoints.agents.pullRequests is enabled with a token reference.
* fix: Address Review Findings on Pull Request Lookup and Lane Writes
Scope the lookup cache, shared in-flight request and rate limit cooldown to the credential, so one tenant's token never serves another. Ask for open pull requests on their own before falling back to the newest closed one, read every page of check runs up to a bound and never report an incomplete rollup as passing, order lane writes per conversation so an older write cannot land last, and name both required token permissions in the example config.
* fix: Address Second Round Review Findings on Pull Request Lookup
Clear the stored lane when the conversation moves to another workspace or leaves attached execution, so the previous workspace's pull request is never served. Fence the lane write on the time it was reported, so an older report from another replica cannot replace a newer one. Honor GitHub's reset window up to an hour instead of ten minutes, and expose the request timeout, whole lookup timeout and check run page limit as endpoints.agents.pullRequests settings with the previous values as defaults.
* fix: Authorize Pull Request Repositories and Fence Lane Writes
Require an administrator allowlist of repositories before the server token is used, since a worker reports its own repository. Match a pull request to the commit the conversation last ran at, so a reused branch name does not return another pull request. Fence the lane write on the workspace the command ran in, so a write queued before a move cannot bring the old lane back, and record a subagent thread's lane on the visible root conversation. Key the lookup cache and shared request by the recorded commit and the lookup policy, and document the worker flag in the example config.
* fix: Order Lane Reports by a Database Sequence and Tighten Head Matching
Replace wall clock ordering of lane reports with a per conversation counter reserved in the database when the command settles, so replica clock skew cannot reject a newer report. Treat a pull request whose head is behind the recorded commit as no match, and expose the candidate and comparison bounds of the history search as settings. Read the app config and the owner scoped lane together, inject the HTTP transport from the route instead of falling back to a global, skip a report identical to the last one written, and name the Contents permission that commit comparison needs.
* fix: Queue Lane Reports by the Visible Conversation and Read Checks for the Current Head
Order lane report reservations by the conversation the write lands on, so two subagent threads of one conversation share a chain, and write every report instead of skipping a repeat, since another recorder may have changed the lane in between. Read a pull request's detail before its check runs and use the head it reports, so line counts, mergeability and checks describe one commit. Treat an expired temporary chat as gone when reserving, writing and reading the lane, and make the shared cache capacity a setting.
* fix: Fence Lane Reports by Attachment Epoch and Place Recorders Before Reporting
Strip dotted descendants of lane fields from saves and imports, partition the pull request cache per credential, resolve the lane write target when the tool is created, and fence reports with a codeAttachmentEpoch that advances only on an owner move or detach.
* fix: Fence Subagent Lane Reports by the Root Epoch and Cover Forks and Moved Heads
Read the visible root's attachment epoch for a subagent report, gate lane recording on pullRequests.enabled, make the credential partition bound a config field, revalidate the detail head against the recorded commit, and find fork pull requests through the recorded commit.
* fix: Keep the Credential Cache Bound Deployment Wide and Budget Head Revalidation
Apply the largest configured credential bound to the shared cache, count cache hits as use, spend the comparison budget on a moved head, and fix the data-provider config default test.
* fix: Cool Down Headerless Secondary Rate Limits
A 403 whose message names a secondary rate limit now starts the credential cooldown for at least a minute instead of being cached as a ten second failure.
Code workers can report the branch and head commit of the lane a command ran in. LibreChat now accepts an optional laneGit key on execute_command results, bounded to a 256 character branch without control characters and a 40 or 64 character lowercase hex head, and rejects anything else. The data-provider gains the pull request types, endpoint, query key and service method, and an endpoints.agents.pullRequests config block that is off by default and takes the token only as an environment variable reference.
* 🔁 fix: Recover Concurrent First SAML, LDAP and Social Logins
SAML, LDAP and social logins looked the user up and then inserted it
with nothing in between, so two first logins for one new identity both
inserted and the unique index failed the second with E11000.
Every strategy now provisions first-login users in packages/api:
createUserOnce inserts through createUserIfAbsent and, on user_exists,
repeats the strategy's own lookup and resolves to the account that won,
once its creator finished provisioning it (its start balance exists when
one is configured), admitted exactly as a found account (a tenant
account resolves its tenant config and that config's email-domain policy
applies). provisionSamlUser, provisionLdapUser and provisionSocialUser
own the create / recover / refresh decision, so a recovered account gets
the same refresh as any account the login finds (claimSamlIdentity, the
LDAP directory overwrite, handleExistingUser), and the CJS strategies
only pass their models and services and map the result to passport.
The lookups take a plain FindUserByFields and the contracts carry
UserRecord and NewUserData. createOpenIDUser now goes through
createUserOnce.
Related to #16766
* ⚡ perf: Read a Recovered Account's Config and Balance Together
After losing the insert race, createUserOnce resolved the tenant config
and only then read the start balance, although the balance check now
depends only on the creation config. Both reads start together.
* 🚪 fix: Let Concurrent First OpenID Logins Continue as the New User
Two requests for the same not-yet-created OpenID identity both missed
findOpenIDUser and both called createUser; the unique email and OpenID
indexes rejected the second insert, and that login failed with E11000.
createUserIfAbsent (data-schemas) inserts a first-login user and reports
a lost race as { ok: false, error: { code: 'user_exists' } }. It
initializes the start balance under the new user's id before inserting
the user, so the account is never visible without it, and removes that
balance when the write fails or a unique index rejects the insert.
createUser shares the insert-only credit (creditStartBalance).
createOpenIDUser (packages/api) creates through it and, on user_exists,
repeats findOpenIDUser with its provider and issuer checks and continues
as the account that won, admitted and refreshed exactly as if the first
lookup had found it: the account's own config (tenant config for a
tenant account) and its email-domain policy apply, and applyOpenIDProfile
writes this callback's claims. It continues only once the winner
finished provisioning the account: when that config sets a start
balance, the balance must exist, which a previous-release creator
credits only after its insert. UserRecord and NewUserData are the plain,
storage-neutral contracts.
Related to #16766
* ↩️ fix: Hold a Recovered Login on the Creation Config's Start Balance
The provisioning barrier read the recovered account's tenant config.
With a tenant start balance the base config does not set, every
concurrent first login in that tenant failed again, because the winner
creates the user under the base config and login balance sync only
initializes the tenant balance afterwards.
The barrier guards against a previous-release creator's $inc landing on
a balance login balance sync already initialized, and that creator only
credits the base config's start balance. createOpenIDUser now holds the
recovered login only when the config new users are created under sets a
start balance; a tenant-only start balance is initialized insert-only by
login balance sync and cannot be added twice.
* fix: Keep Reasoning Out of Every Read Aloud Path
Read aloud spoke the Thoughts section along with the answer. The hover
button flattened think parts into the string it handed to TTS, browser
autoplay and the hooks parsed content without skipping reasoning, and
server streaming spoke the stored text field, which an aborted turn
persists with its reasoning included.
Add getSpeechText to data-provider as the single source for spoken text:
text parts only for structured content, and the legacy :::thinking block
stripped from plain text, treating a still-open block as reasoning so
streamed speech only ever grows. Route the hover button, browser
autoplay and server streaming through it, and skip reasoning in the TTS
hooks' own content parsing.
* fix: Speak String Message Content Without Its Reasoning
Imported or client-built messages can carry their answer as string
content rather than parts; read it like plain text instead of falling
back to an empty text field.
* fix: Keep Answers Between Repeated Legacy Thinking Blocks
Strip every closed :::thinking block before treating an unterminated one
as reasoning, so a second closed block no longer swallows the answer
that follows it.
* fix: Skip Read Aloud and Autoplay When Only Reasoning Remains
A turn stopped mid-reasoning has no answer to speak once reasoning is
filtered. Gate autoplay on the same speech text it will play, so the
external engine never opens a stream that can only time out, and hide
the Read aloud action when there is nothing to read. Every speech path
reads a finalized message, so speak an unmatched :::thinking marker as
the answer text the UI shows instead of dropping the rest.
* refactor: Move Speech Chunk Polling Into packages/api
The decision of what text to stream as speech lived in the legacy CJS
service. Move the chunk processor to packages/api as
createSpeechChunkProcessor, taking the message cache and lookup from the
caller, and leave streamAudio.js wiring them in. Its tests move with it
onto an in-memory cache instead of module mocks.
* fix: Speak Exactly the Answer the Message View Shows
Share the legacy :::thinking parser between the message view and speech,
so Read Aloud strips the same block the view collapses instead of a
separate pattern drifting from it. Trim structured speech text so a
whitespace-only answer is not speakable, and key autoplay's audio cache
by the spoken text, as manual playback already is, so audio cached with
reasoning before this change is never replayed.
* fix: Fall Back to Text When String Content Is Empty
* feat: Let a Theme Quiet Decorative Borders
* test: Cover the Chrome and Inset Border Roles in the Mock Harness
* fix: Draw the Selected Category Tab Edge at the Chrome Share
* fix: Route Chips, Subtle Icon Buttons and Hover Rings Through the Chrome Share
* fix: Draw the Remaining Inner Card Dividers at the Inset Share
* feat: Add a Focus Role for Rows Inside Content
* test: Cover the Subtle Focus Role in the Mock Harness
* test: Probe the Subtle Focus Role Through Its Theme Variable
* test: Focus the Probe to Read the Subtle Focus Utility
* test: Read the Subtle Focus Ring From Its Utility Property
* fix: Ring the Subagent Progress Disclosure in the Subtle Focus Role
* 📴 fix: Finalize Compaction Snapshots and Move Abort Policy Into packages/api
* 🔕 fix: Inspect Every Part When Finalizing and Reuse the Loaded Row
* 📴 fix: Settle Finalized Compaction Rows and Verify Their Anchor
* 📴 fix: Withhold the Unanchored Abort Final and Settle Marked Snapshots
* 📴 fix: Consolidate the Compaction Abort Policy Into One Injected Operation
* 📴 fix: Settle the Live Compaction Row Past an Anchor-Shaped Collision
* 📴 fix: Lift the Failed-Turn Settlement Into the Compaction Module
* 📴 fix: Keep a Removed Round's Failure and Guard Settled Jobs From Snapshots
* ⚡ perf: Read Only the Anchor Id in the Abort Existence Check
* 📴 fix: Withhold the Abort Final Only When a Row Needed Writing
* test: Cover Stopped Compaction Finalization and the Missing Anchor in E2E
* fix: Withhold a Stopped Compaction That Carries No Anchor Id
* fix: Settle Stopped Compactions, Announce Finalized Rows, and Own the Abort Read in packages/api
* test: Accept the Model Spec Query on the Stopped Reply Conversation URL
* fix: Report the Anchor Read Failure Beside a Withheld Compaction Abort
* fix: Synthesize a Compaction Failure Only for a Round After the Latest Outcome
* fix: Skip the Anchor Read When a Stopped Turn Writes No Row
* fix: Decide the Disconnect Snapshot Gate in packages/api
The last-subscriber partial save now asks resolveDisconnectSnapshotMode for
the settled-job decision instead of branching on the job record in the CJS
controller.
* fix: Leave an Already Settled Compaction Row to the Path That Settled It
* 📴 fix: Persist Cancelled Compactions With Their Identity
* 🧪 test: Build the Cancelled-Compaction Leaf on a Real Conversation
* 🧪 test: Assert the Withheld Controls on the Live Cancelled Compaction
* 📴 fix: Keep Live Disconnect Compaction Rows Free of a Synthesized Failure
* 🧪 test: Cover the Compaction Flag in the Redis Job Round Trip
* 📴 fix: Keep a Stopped Compaction Honest and Its Anchor Untouched
* feat: Layered Surface Roles for the ClickHouse Theme
Adds purpose-named roles for the chat canvas, user message, cards, sidebar rows, settings tabs, menu panels, composer and search pill. Each follows the surface it painted before, so the default theme is unchanged; ClickHouse steps them apart with Click UI tokens.
* fix: Align Sticky Phase Headers, Boot Routes, and Row Fades With the Layer Roles
* fix: Cover Trailing Slash and Bare Dashboard Paths in the Boot Canvas Route
* fix: Match Boot Canvas Routes Case-Insensitively
* test: Expect the User Message Role on Other-Author Turns
* test: Expect the User Message Role on the Subagent Briefing
* feat: Tooltip, error alert and artifact preview theme roles
* test: Cover the Tooltip Theme Roles in the Mock Harness
* fix: Lighten the Dark Document Tile and Leave the Preview Color Scheme Alone
* fix: Let the Artifact Preview Frame Follow the System Scheme and Ink Autofill in the Field Text
* 🧾 perf: Load Tool-Call Previews and Fetch Full Parts on Demand
Long agent conversations shipped every tool output, argument string and
subagent transcript on load, though the cards render collapsed: 12-18 MB
of message JSON per chat, enough to crash iPhone Safari tabs.
A client that adds ?toolPreviews=1 to the conversation load now receives
bounded previews of settled tool calls (start and end of long outputs,
JSON arguments and outputs kept valid with only their longest strings
shortened, exit-status trailers kept whole, subagent transcripts omitted)
with explicit markers and full lengths. Opening a card or a subagent
panel fetches the stored part from a new owner-scoped endpoint,
GET /api/messages/:conversationId/:messageId/parts/:partIndex, cached
under its own React Query key. Clients that do not ask, exports and share
links keep the full payload; bounds are configurable under
toolCallPreviews in librechat.yaml, and a client-authored message that
carries previews is refused rather than saved.
* 🧾 fix: Validate Part Reads in Parallel and Preview Fork Responses
The full-part read now starts beside conversation-access validation
instead of after it, and keeps the active-job fallback the conversation
read has, so a conversation still live before its first save can expand
its previews. Fork, duplicate and shared-link fork responses honor the
preview parameter, so they no longer seed the conversation cache with
the full history. Background-task results whose structure cannot fit a
valid JSON preview go whole, image cards keep their prompts, and the
subagent panel offers a retry when its stored part fails to load.
* 🧾 fix: Key Hydrated Parts by Preview Revision and Call Identity
A fetched part is now cached under the preview's revision and replaces
only the fields the server shortened, so a call whose stored output
changes later (a harvested background result) is fetched again rather
than rolled back. Part lookups carry the run-step and agent identity and
refuse an ambiguous fallback, since provider tool-call ids can repeat
within one response. A subagent that finished without final text has
its transcript omitted like any other settled call, and JSON nested past
a fixed depth falls back to a text preview; a call whose preview fails
is sent whole instead of failing the conversation load.
* 🧾 fix: Release Hydrated Parts With Their History and Bound JSON Work
Full tool-call parts fetched for a conversation are removed when its
message history is released or cleared, so they cannot outlive it under
their own key. The structure-preserving JSON preview now bails out before
cloning when the value has more nodes or strings than could ever fit,
keeps a __proto__ key as ordinary data, and the part endpoint accepts any
safe-integer index instead of an arbitrary ceiling.
* 🧾 fix: Count Property Names Before Cloning a JSON Preview
Property names are never shortened, so the pre-check now adds their
length to its budget and falls back to the text preview before any
cloning when they alone could not fit.
* 🧾 fix: Preview Artifact and Branch Responses and Rehydrate Subagent Prompts
Artifact edits and parallel-response branches merge the message the
server returns into the conversation cache, so both requests now ask
for previews and their routes honor it. The subagent panel reads the
prompt from the stored arguments once they load, and the part endpoint
no longer caps identifier length, so any previewed call can be fetched.
* 🧾 fix: Preview Only Tool Calls That Can Be Fetched Back by Identity
A call without an id has nothing but its position to name it, and the
position can shift on the client, so it is left whole instead of being
previewed. The part endpoint now requires a tool-call id, so an indexed
lookup always confirms the call it returns.
* 🧾 fix: Fetch Parts by Exact Identity and Preview Only Unique Ones
The part endpoint now finds a call by its exact identity (provider id,
run-step id, agent) instead of trusting the client's index, and the
preview pass only shortens calls whose identity is unique in their
message, so every preview names exactly one stored part. Legacy
web-search failures that carry their error phrase mid-output go whole,
since the collapsed card reads that phrase anywhere.
* 🧾 fix: Bound the JSON Fit Walk Before Queueing and Normalize Empty Identities
The structure check now counts an array's or object's children against
the node budget before queueing them, so a very wide value is refused
without building a work list its size. An empty run-step or agent id is
treated as absent by the part lookup, matching how the preview pass and
the client already treat it.
* 🧾 fix: Key Fetched Parts to the Stored Version and Hydrate Search Details
Each preview now carries `previewRevision`, when its message last
changed, and the client keys fetched parts to that plus a fingerprint of
the whole preview, so any change to the stored call is fetched again.
The web-search details card asks for the stored call when it opens, so
its query is never left shortened.
* 🧾 fix: Cache Fetched Parts by Identity and Hold Copy Until Hydrated
The part query key drops the client-side index, which only rides along in the request; the server locates the call by message and exact identity, so a shifted index serves the cached copy instead of downloading it again. Copy controls inside a previewed part are disabled until the stored content arrives, so a slow or failed fetch cannot export shortened text.
* 🧾 fix: Shape Shared Forks With the Recipient's Bounds and Keep JSON Numbers Exact
A shared-link fork resolves the share owner's config for its content checks, but the forked conversation belongs to the requester, so its preview bounds now come from the requester's tenant and role (prepareResultToolCallPreviews). A structure-preserving JSON preview now requires every number in the text to survive a parse and re-serialization as the same value; an integer past 2^53, an out-of-range exponent or an over-precise decimal falls back to the text preview (or the whole output for parsed-output tools) instead of showing a changed value.
* 🏝️ fix: Let a No-Workspace Turn Carry a Remembered Approval Mode
A conversation set to run without attached workspaces failed every turn with
"The selected code approval mode is not permitted by the current policy" when
the reader had picked Accept edits or Full access for a workspace earlier.
With no attached target, the server validated the requested mode against an
empty policy that only allows ask. A turn with no attached target now accepts
any known mode, resolves it to ask for the run, and keeps the conversation's
stored mode instead of overwriting or unsetting it. Unknown values are still
rejected, and any attached target, including a subagent's, keeps the strict
check.
The composer treated an explicit no-workspace chat as still running on the
agent's default machine, so it kept offering and submitting the remembered
mode. It now offers no attached mode there, and the submit path sends ask
whenever the turn resolves to run without a workspace.
* fix: Leave an Unset Approval Mode Unset on a No-Workspace Turn
A turn without an attached target seeded the conversation's approval mode
from the request when the row stored none. A current client sends the gated
ask there, so the first no-workspace turn of a new chat wrote ask and
replaced the remembered pick the composer would otherwise keep offering.
Such a turn now only carries forward a mode the conversation already stores;
a stored mode still comes only from a turn that validated it against its
attached targets.
* fix: Gate the Approval Menu on the Resolved Workspace Mode and the Saved Conversation
The composer hid attached approval modes only for a chat that recorded a
no-workspace decision, while useCodeWorkspace also resolves one for a saved
chat with no recorded decision and for a draft still choosing. The approval
menu, the queued-turn snapshot, and the submit path now read the mode the
composer resolved, so the control matches what the turn sends.
A no-workspace turn kept the mode of the row admission loaded, but a turn
sent with overrideConvoId is saved under another conversation; the stored
mode is now read only when that row is the conversation being saved.
* fix: Preserve the Saved Conversation's Approval Mode Through a No-Workspace Turn
A no-workspace turn chose what to record from the row admission loaded, but
a turn sent with overrideConvoId is saved under another conversation while
its unsets are still computed from that loaded row, so either the source
row's mode was copied onto the target or the target's own mode was unset.
Such a turn now records no mode and declares codeApprovalMode preserved:
the conversation write skips it when computing unsets, so the stored mode of
whichever conversation the turn is saved under stays as it is, and a chat
that stores none keeps offering the remembered pick. Turns with attached
targets record exactly the mode they validated, as before.
interface.runningChatRename now defaults to true, so a running chat or a placeholder-title chat can be renamed without a librechat.yaml override. Set it to false during a rolling upgrade from a version without title ownership.
* 🔱 feat: Simplify Steering Controls to Three Modes
* Preserve Admitted Artifacts During Interrupts
* Preserve Result Admission Across Interrupts
* Gate Interrupt for Provider-Hosted Tools
* Match Interrupt Hints to Availability
---------
Co-authored-by: Lia <lia@librechat.ai>
* 💳 feat: Show Balance in the Context Gauge as Credits, Currency, or Percent
Move the balance out of the account menu into the composer's context gauge,
below the context window, and add `balance.display` (credits | currency |
percent, default credits) to choose how it reads. Currency reuses
`interface.currency` at 1,000,000 credits per USD; percent shows only the
share of the period's allotment spent.
Settings > Account > Billing renders the same summary from the same hook, so
both surfaces show identical figures, bar and refill timing in every mode.
* 🧭 fix: Normalize the Balance Display Mode and Reuse a Fresh Reading on Hover
Admin overrides reach the client without schema validation, so an unknown
`balance.display` now reads as credits instead of being typed as a mode it is
not. Hovering the gauge mounts the balance reader each time the card opens; a
30 s stale window stops every hover from refetching, while the explicit
refetch after each turn still bypasses it.
* 🩹 fix: Restore Gauge Popover Fixes and Tighten Balance States
The popover extraction was written from a pre-sync read of TokenUsage, so it
dropped dev's viewport caps and scrolling, the hideOnInteractOutside close fix,
the Tailwind v4 outline-hidden rename, and the card's max-w-full; restore them.
Refill copy now says when a refill becomes available rather than promising a
deposit, since one lands only once the interval has passed and a charge would
exhaust the balance. An empty balance reads as fully used even without an
allotment, percent mode without one says usage is unavailable instead of
leaving the figure blank, the balance-only ring takes the balance tone (amber
until empty) and is a meter only once it has a value.
* 🪙 fix: Read a Non-Positive Refill Amount as No Auto-Refill in Settings
The server never refills when the amount is not positive, and the balance
summary already says so; Settings showed its load error instead.
* 💸 fix: Show the Insufficient-Funds Error in the Balance Display Mode
The token-balance error still printed raw credit figures, so `percent` leaked
the numbers it hides and `currency` mixed units with the gauge. It now reads
the same display mode and formatter as the balance summary: credits as before,
money in currency mode, no figures in percent mode.
The auto-refill settings row also renders its own loading and failure states,
since settings search can show it without the balance row beside it.
* 🎚️ fix: Keep Fractional Credits and Drop the Moved Popover's Suppression
Credit amounts keep up to two decimals, so an insufficient-funds message no
longer reads 100.4 against 100.3 as 100 against 100; whole amounts still print
without decimals.
The design-rule backlog may not follow code into a new file, so the extracted
popover uses the standard `transition` utility instead of the arbitrary
`transition-[opacity,transform]` (which also misses Tailwind v4's separate
`translate` and `scale` properties) and its suppression is removed. The
composer geometry spec now reads the classes from `Popover.tsx`, where the
gauge button lives.
* 🧭 feat: Let a Parent Choose Its Subagent's Code Machine Per Call
A parent agent can now pass optional `machine` / `workspace` arguments on a
subagent call to pick where the child runs, using the host-argument
extension point added to @librechat/agents.
- Each lazy child declares enums built only from machines it may reach:
admitted in the conversation's sealed codeWorkspaces, on its default or
allowlist, in the requesting principal's tenant-scoped list, and served by
a ready worker. A child the conversation pinned to a machine offers only
that machine; a child with no reachable machine declares nothing.
- On use, a request-scoped routing session in packages/api re-validates the
choice against the current request and live worker status and refuses an
invalid one with a typed, model-visible error instead of falling back. A
routed child is pinned to the chosen machine at every routing site,
overriding parent inheritance.
- Omitted arguments keep inherit-by-default routing, extended so a child
follows the machine a per-call-routed parent runs on (passing through
agents without code). A subagent keeps one machine for the whole request.
- A child whose default workspace is unavailable stays guarded unless it has
per-call choices; then it stays callable and its description asks the
parent to pick a listed machine.
- Routed children keep their tool context per execution for tool loading,
provisioning and artifact handling.
- Paused-approval bindings, the attached-machine policy hook, skill-file
priming and tool-output references account for per-call choices.
The declarations are gated on the installed SDK exporting
SubagentHostArgumentError, so this change is inert until LibreChat depends on
an @librechat/agents release that includes it.
* 🧭 fix: Reserve Default Routes Early and Route a Routed Subagent's Graph Members
- Reserve a subagent's default machine when it is placed, as
initializeAgent will resolve it (including parent inheritance), so a
concurrent call cannot claim another machine during initialization.
- Graph members spawned by a per-call-routed subagent follow its machine
when they may, unless they already initialized on their own route this
request.
* 🧭 fix: Settle Concurrent Inherited Routes and Register Routed Graph Members
- Re-read a subagent's inherited route after resolving targets, so two
concurrent calls from differently routed parents settle on one machine.
- Attribute a workspace-only conflict to the workspace argument.
- Register graph members a routed subagent initialized, so each gets the
attached-machine policy of the machine it actually runs on.
* 🧭 fix: Re-read Inherited Routes After Resolving Subagent Targets
- Re-read a subagent's inherited route after the target resolution await,
so two concurrent omitted-machine calls from differently routed parents
settle on one machine.
- Attribute a workspace-only conflict to the workspace argument.
* 🧪 test: Prove a Routed Graph Member Takes Its Machine's Policy Through createRun
- Register a resolved child's graph members inline in the lazy resolver
instead of through an exported generic helper whose test did not
typecheck.
- Drive the check through createRun's real PreToolUse hook: a member
routed to a deny machine flips from its metadata machine's allow to deny
once its parent resolves.
* 🧭 fix: Hold a Lazily Loaded Graph Member's Default Machine While It Initializes
- Pass a graph member's code flags when it initializes on its default
route, so the routing session reserves that machine before the
asynchronous initialization instead of only after it, and a concurrent
call cannot place the same agent on another machine meanwhile.
* 🧭 fix: Validate the Approval Mode Against Every Machine a Subagent May Use
- Collect each agent's default attached route and each attached machine a
parent may route it to per call, and validate the requested approval
mode against that set in createRun and both client checks.
- A child whose default is unavailable but whose per-call machine permits
accept edits no longer fails run creation; full access still requires
every such machine to permit it.
* 🧭 fix: Admit, Prime, and Cancel Per-Call Subagent Routes Conservatively
- Build scheduled pause admission from every attached machine a parent
may route an agent to, so a permissive default does not hide a
per-call machine that asks.
- Walk lazy graph-member metadata when collecting execution routes, so
historical skill files are primed for every machine such a member may
run on.
- Stop placing a canceled call after its target resolution resolves, so
it never claims a machine for an execution that will not initialize.
* 🧭 fix: Release Unused Route Claims and Scope Per-Call Machines to Their Run
- Hold a subagent's route claim per placement: a failed or canceled
initialization gives the machine back unless another execution still
holds it, and a claim is permanent once a child on it attaches.
- Fold an agent's default route into its per-call choice digest, so each
agent contributes at most one paused-approval binding target.
- Use only an agent's route policies for pause admission when it has
them, so a missing default does not make a decided route look asking.
- Drop a routable child's provisional attached-machine membership when it
resolves off attached machines.
- Offer and accept no attached machine per call while run files are
shared, since shared children must run on managed code.
* 🧭 fix: Give Back a Route When Its Placement Verification Fails
- Release a placement's hold when attach finds the child initialized on a
different machine than the one it was routed to, so a retry that
follows the unavailable-machine guidance may name another machine.
* 🧭 fix: Commit a Lazy Child's Routes Only When Its Whole Resolution Succeeds
- Keep the routes a lazy child and its routed graph members hold pending
through the child's entire resolution (settleExecution), committing them
on success and giving them back on failure or cancellation unless
another execution still holds them.
- Placements outside any call (shared graph members) still commit when
they attach.
* 🧭 fix: Drop a Failed Lazy Resolution's Config from the Graph Cache
- When a lazy child's graph resolution fails or is canceled after its
config was cached for graph reuse, remove that entry with its route, so
a later graph never reuses a config for a machine the child gave back.
* 🧭 fix: Follow a Routed Agent Through Self-Spawned Runs and Scope Routed Member Configs
- A call whose parent run was never placed (an allowSelf spawn reuses its
routed inputs) inherits the routed claim of its parent agent, which keeps
one machine per request.
- Never share a graph member initialized for a routed parent's execution
through the graph-member cache, so a failed resolution leaves no config
behind for a route it gave back.
- Race lazy target discovery with the request signal.
* 🧭 fix: Pass a Machine On Through Self-Spawned Runs of Agents Without Code
- Record the machine each agent without stateful code passes on, and let a
call from that agent's unplaced run (an allowSelf spawn) follow it when
the agent passed on exactly one machine this request.
* 🧭 fix: Keep a Pass-Through Route Only When Its Resolution Succeeds
- Record the machine an agent without code passes on with the rest of its
call's routing state: kept when the resolution deciding it succeeds,
dropped when it fails, so a self-spawn never follows a failed attempt.
* 🧭 fix: Settle a Lazy Child's Routing State at the SDK Handoff
- Wrap the run's whole lazy resolution, through the inputs handed to the
SDK, in the descriptor's settle hook, so a cancel after the host
resolver returns still gives the call's routes back.
- Roll back the per-agent tool context a pending placement set (restoring
the entry it replaced) when its resolution fails, so a retry on another
machine seeds its own context.
- Drop a failed call's shared config from the graph cache through the
same settle hook.
* 🧭 fix: Keep Per-Agent Tool Contexts on Live Placements and Queue Target Probes
- Track every pending placement that uses an agent's per-agent tool
context: a committed one sets it, a failed one hands it to a surviving
placement, and only the last failure restores the earlier entry.
- Let an unplaced run of an agent (an allowSelf spawn) defer to the routed
run of the same agent that encloses it when looking up tool contexts.
- Queue live target probes at eight in flight, below the shared status
poller's ceiling, so concurrent descriptors never misreport a healthy
machine as unavailable.
* 🧭 refactor: Load Routed Graph Members Through a packages/api Service
- Move the routed graph-member load (shared reuse, skip list, view check,
execution identity, failure policy) into createRoutedGraphMemberLoader;
the initializer only injects its dependencies.
- Thread the request and call signals into target resolution so queued
probes are skipped once canceled.
- Probe nothing in shared-file runs, where every per-call target is
attached and unusable.
- Retire pending tool-context ownership when a committed placement
installs the per-agent entry.
* 🧭 fix: Share a Lazy Child's Config with Later Graphs Only After It Commits
- Publish a default-placed lazy child's config to the shared graph cache
from the settle hook's commit, so no concurrent graph can borrow a config
whose route may still be given back.
- Resolve a child's own appearance in its graph to the child config itself.
* 🧬 fix: Default Subagents to Their Parent's Code Machine and Workspace
A subagent resolved its code machine from its own saved fields only, so
when a chat routed the parent to an allowlisted machine the subagent fell
back to its own default and whatever workspace the conversation had
selected there, or failed the parent's whole turn when none existed.
Subagents now inherit the parent's resolved attached machine when it is
their own default or on their allowlist where per-chat choice applies,
the principal can use it, the sealed decision already selected it, and no
selection names the subagent as an explicit owner. Inheritance is derived
per request from the sealed decision and current agent configuration and
fed to every routing site; it is never persisted, so existing chats and
older clients get it without a decision change. The composer mirrors it.
A subagent whose machine has no usable workspace is advertised to the
parent as unavailable instead of failing the turn; a locked decision
stays fatal.
* fix: Inherit Across Every Spawning Path, Graph Members and Deferred Code Tools
A subagent now inherits only when every agent that can spawn it, at any
depth, runs on the same machine; spawn cycles keep their own routes.
Subagent graph members inherit like explicit subagents. The deferred
execute_code loader passes inheritance to its fallback resolver. Subagent
reads and VIEW checks move to a packages/api loader shared by inheritance,
lazy descriptors and graph members. The composer lets a draft's subagent
follow its parent's machine before a workspace is picked, and keeps an
inheriting subagent's former selection without asking for a move.
* fix: Prune Spawn Back-Edges, Gate the Composer on an Inheritance Capability
Spawning parents follow the run's path semantics, so a back-edge the
descriptor tree prunes no longer blocks inheritance, and agents that spawn
each other follow their single outside parent machine when it routes them
all there. Only admitted subagents count toward the walk's node limit.
Graph members re-read and re-check VIEW when they initialize. The composer
echoes an inheriting subagent's former selection without gating on that
machine, and mirrors inheritance only when the API advertises
codeWorkspaceInheritanceVersion; approval modes keep intersecting every
machine an agent could use.
* fix: Settle Each Mutually Spawning Group on Its Own
Independent spawn cycles no longer merge into one group, so a pair under
one parent follows that parent even when another pair follows a different
machine. Also types the data-provider spec for its ES5 target.
* fix: Bound the Inheritance Walk, Normalize Parallel Roots, Move Subagent Availability to TS
The walk reads in batches no larger than the remaining node budget, stops
once the limit is reached, and runs through the run's subagent load limiter.
Roots are keyed by saved agent ID, so a parallel root such as id____1 can
never receive an inheritance entry. Resolving a lazy subagent's code route,
reporting it unavailable, and guarding its descriptor now live in
packages/api (resolveSubagentCodeAvailability, guardUnavailableSubagent).
* fix: Keep Missing Subagent Loads Parallel Near the Node Limit
Batches never shrink below eight IDs, so missing or unviewable subagents
near the limit still load in parallel under the injected limiter, and at
most a handful of reads land past the limit.
`mistral-large-2512` (and `mistral-large-latest`, which now points to it)
had no dedicated entry and fell back to the generic `mistral-large` key,
billing at Mistral Large 2 rates ($2 / $6), 4x the actual price.
- Add `mistral-large-2512` and `mistral-large-3` at $0.50 / $1.50
- Point the generic `mistral-large` fallback to the current model's price
- Pin `mistral-large-2411` to its $2 / $6 rate
- Drop the stale Sonnet 5 "revert after 2026-08-31" comments: Anthropic
made the $2 / $10 pricing standard, so the scheduled increase won't happen
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* 🔗 feat(agents): use LibreChat prompt groups as agent instructions
Agents can link a prompt group revision or Production as their instructions. The link resolves at the start of each turn through a cached resolver. The cache clears on promote and delete. Editors without VIEW on the group see a restricted placeholder.
* fix: remove duplicate logger import from agent initialization tests
* fix: honor prompt management capabilities for linked agent instructions
* test: seed prompt manager capabilities in authorization fixtures
* fix: clear acknowledged linked prompt edits before version restore
* test: adapt merged chain coverage to prompt-aware payload helper
---------
Co-authored-by: Lia <lia@librechat.ai>
* fix: Sync Spinner Rotation Phase Across Instances
* test: Colocate Spinner Spec In __tests__
* fix: Address Review Findings on Spinner Phase Pinning
* fix: Fall Back To Legacy Media Query Listeners In Spinner
* fix: Repin Recreated Spinner Rotations And Defer Without queueMicrotask
* feat: add app-wide UI scale setting
* test: align dev's new header, drawer and combobox specs with the scaled shell
* fix: keep design-recorded arbitrary values off the same-count swap
* fix: scale composer and reasoning spacing on the spacing scale
* test: reach scaled rows and wait on the composer under the unpainted drawer
* test: cover the scaled composer at both ends of the range
* test: reload the composer scenario at each scale from the live page
* fix: leave the root style untouched at the default scale
* test: align dev's tool output and subagent inset specs with the rem classes
* fix: scale a caller's combobox height cap with its rows
* fix: size the conversation list's row estimate on the scaled drawer breakpoint
* test: align dev's conversation row status spec with the scaled drawer breakpoint
* style: sort the settings dialog imports after the dialog layer merge
* fix: scale the collapsed combobox's fallback popover width
* fix: size the upload glyph on both axes so it grows with the scale
* perf: Async Local File Saves and Compile Hot-Path Regexes Once
* perf: Check Closing Code Fences Without a Per-Length Regex Cache
* test: Build the Delivery Route Config Before Routing Instead of Mutating It
mergeFileConfig now memoizes by config identity, so the CSV fallback tests
must set textFallbackWithoutTools when they build the request config rather
than mutate the object after the agent's routing already merged it.
* fix: Write Local Buffers Atomically So Concurrent Saves Cannot Interleave
Async writeFile calls to one deterministic path (OpenID and SAML avatars)
could truncate and interleave each other. Write through a temp file and a
rename instead, and add scenario specs for local saves and longer closing
fences.
* refactor: Move Local Save Logic Into packages/api
saveLocalBuffer, the temp-upload move and the atomic directory writes now live
in packages/api/src/storage/local.ts and receive their paths from the caller;
the Local strategy in /api keeps its logging and calls through. The traversal
check now runs before any directory is created.
* fix: Copy Moved Uploads Through a Temp File and Rename Into Place
Concurrent async copies to one destination could interleave just like direct
writes, so moveLocalFile now fills a sibling temp file and renames it.
* fix: Contain Every Local Write and Move Inside Its Target Directory
The traversal check that guarded saveLocalBuffer now runs in writeLocalFile
and moveLocalFile too, before any directory is created, and the buffer
save accepts only the images and uploads base paths.
* test: Define the Success Assertion Inside the Local File Saves Scenario
* fix: Give Atomic Write Temp Files a Fixed-Length Name So Long Targets Still Fit
* test: Send Image Dimensions and Read the Message Array in the Local Saves Scenarios
* feat: Return a Navigable Progress Tree When Checking a Running Subagent
check_background_task on a running or finished-but-undelivered subagent now
returns an activity digest: the child's turns, tool calls, and replies folded
to one line each and addressed by stable positional paths. Callers expand one
node or range with expand, and page with since against a settled-node cursor.
The owning process folds the child's update stream into a bounded tree that
keeps only tool names, the model-authored intent label, the tool-authored
outcome label, statuses, timings, and sizes. Routed claims carry the tree to a
poll on any replica, lists carry counts and the node in flight, and a finished
child whose owner is gone is rebuilt from its durable public projection.
The background-task card renders the digest as a foldable tree and links the
existing subagent activity panel.
* fix: Tighten Progress Digest Verdicts, Overflow Bounds, and Rebuilt Trees
Failed reply steps and schema-rejected calls now record as errors, and an
overflowed call is counted once without retaining it in the lookup maps.
A tree rebuilt from the durable projection re-checks completed calls with the
shared failure verdict, carries the projection's truncation, and tells the
caller its paths are a reconstruction. The card localizes folded summaries
instead of rendering the server's English tally.
* fix: Measure Single-Part Message Deltas in the Progress Digest
* fix: Scope Nested Progress Ownership by Run and Count Overflow Once
Nested runs now attach to the call recorded by their own spawning run, since
provider call ids repeat across runs. Only the step that announces a call can
count it as overflow, so later sightings of an unretained or evicted call never
re-count it or reopen history. A zero-valued since cursor is rejected, and the
card labels MCP tools with the configured server names.
* 🫗 perf: Release Message Histories of Conversations the User Has Left
Long agent conversations hold 12-18 MB of message JSON each in the React Query cache, and the default 5-minute cacheTime kept every conversation visited in that window. Switching between a few of them on mobile Safari grew the JS heap past 300 MB and crashed the tab.
retainMessages tracks when each per-conversation history loses its last observer and releases it once it has no observers, is not fetching, is not pinned, and is either older than 60 s or beyond the single most recently left conversation. The routed conversation, conversations with a running job, and conversations awaiting an approval decision stay pinned; releases wait while any mutation is pending. Returning to a released conversation refetches its history.
* 🫗 fix: Own History Collection, Configure the Grace, and Pin by Server State
Per-conversation histories get an infinite cacheTime while retention runs, so React Query's own five-minute collection can no longer drop a history a running job still pins; newly seeded histories schedule a pass; the grace and recent count come from interface.historyCacheTtlMs and interface.historyCacheRecent; the approval pin is replaced by the active-jobs set, which already lists jobs paused for a live approval; and Assistants histories stay pinned because deletion reads their thread id from the cache.
* 🫗 fix: Re-arm Collection for Histories Already Counting Down and Recheck Pins Promptly
A history left before retention started still had React Query's five-minute GC timer; a momentary observer now re-arms it under the infinite cacheTime. Histories kept only because they are pinned or fetching are judged again after 5 s instead of a full TTL, so a run that ends releases its transcript promptly.
* 🫗 fix: Exempt Assistants Histories Without Polling and Leave Fetching Histories Alone
Assistants histories are now exempt rather than pinned, so they are kept without a recurring recheck timer. The startup re-arm skips histories with a fetch in flight, since detaching a momentary last observer would cancel the request; settling the fetch re-arms collection on its own.