User System: Cleanup and Bugs by ClaraLeigh · Pull Request #346 · LibreChat-AI/LibreChat

ClaraLeigh · 2023-05-21T11:31:27Z

Features:

Disable registration
Upgrade Script
One central .env file
Environment based files, ie .env.dev, .env.prod, .env.test

Chores:

Refactors User base code
JWT default key warning

Also being added soon:

Install script - node based

Closes #341 and Closes #273

- Simplifies the ENV vars too - Legacy fallbacks for older builds

- Moves both login and logout controllers to their own file

danorlando · 2023-05-22T16:07:38Z

@ClaraLeigh This is great! yes, the user system implementation was a bit hasty and there is still a LOT that can be done to improve it. Thank you for looking into it!

ie .env.prod, .env.dev, .env.test

danorlando · 2023-05-24T22:51:09Z

+      await new Promise((resolve) => setTimeout(resolve, 1000));
+
+      // TODO: We should change the process to always email and be generic is signup works or fails (user enum)
+      return { status: 500, message: 'Something went wrong' };


I have to disagree with the idea of making this generic because its a poor user experience to just tell them that "something went wrong", and will inevitably lead to them pinging us in the discord. If they know that the reg failed because the email is already in use, they can take take the appropriate action.

@danorlando user enumeration is the reason. Generic errors prevent enum. Some people don't care about this security facet but I worry the people deploying having weak pw.
An alternative idea is sending an email, "Hey did you forget your password?" sort of thing, then you avoid enum

I understand your point, but if someone knows how to deploy the app where it is exposed to the internet, the onus is on them to use strong passwords and implement rate limiting and whatever other security measures are necessary for production deployment. I don't think we should sacrifice user experience on our end because of the potential for some dummy to use a crappy password on a deployed instance of an open source app. I think it would be better if we instead strengthened the password requirements to perhaps 10 characters, upper and lower, special character, and number.

I'm not sure its that much of a sacrifice?

Flow would be:
Get an error about your login, try to reset pw, you get an email either saying the account doesn't exist or it does and how to reset.

Additional flow, if you attempt to login but use the wrong pw, an email gets sent saying "are you trying to login? click here to reset". or if they don't exist, "click here to register now".

When it comes to pw restrictions, they don't actually suggest that sort of complexity anymore as it doesn't make things more secure. all lower case with 4 words is recommended over 10 characters mixed. Even apples generated pw's don't have uppercase much anymore. There are definitely ways to avoid enum without loosing the notification side of things. Maybe a discussion for the discord at any rate. I'll leave till last if its contentious but honestly I don't see a problem with it, especially with the flows mentioned

Also, most people have pw managers these days anyway

ok, my only question would be along the lines of - how does the user in fact know that they forgot their password or are using the wrong password and that the error is not due to some other problem? 500 means Internal server error, and is almost always the result of an unhandled exception. This could be misleading, suggesting that there was in fact an internal server error rather than simply bad credentials.

If we don't want to tell the user that they are using invalid credentials from the UI, perhaps an alternative approach would be to send a 400 (Bad Request) back and modify the UI to no longer display the response message and instead add a line saying that there was a problem with their login attempt and to check their email for more information. If that were the approach, It would be probably also be good to have a ? icon, that when hovered, shows a tooltip stating that "For security, we do not display detailed error information here. Please check your email to find out more about why you are seeing this."

I guess another option would be to just add rate limiting on our end, such that after 3 attempts we lock them out and force a password reset from their email, although thats a little bit more work.

Nothing wrong with having it different from bad login vs 500. Didn't realised I changed the code

Lets the steer route ask the generating replica to seal its live model stream at the next provider-safe boundary instead of waiting for a tool step. The run is never aborted, job status never changes, the partial answer is kept, and generation resumes in the same assistant message after the injected steer. Consumes the SDK seam in @librechat/agents (LibreChat-AI/agents#335, #346). Transport: IEventTransport gains a fenced emitPreempt/onPreempt pair beside abort. RedisEventTransport fans PREEMPT out on the SAME events channel and subscription (no new connection, key, or subscribe call); onPreempt returns a registration-scoped unsubscribe with the same replacement-safe state-identity guard onAbort uses. InMemory implements neither — single-process preempt lives entirely in the runtime set. Runtime state: RuntimeJobState carries the per-generation request set, createdAt-fenced and capped at STEER_QUEUE_MAX_DEPTH, plus a bounded `cleared` tombstone so a late cross-replica arm cannot resurrect a request whose steer already drained. registerPreemptSubscription mirrors the abort registration's double fence (runtime identity + generation createdAt); releaseAbortSubscription retires BOTH listeners and the armed set, so every terminal path drops preempt state for free. Public surface: requestPreempt (arm + fenced publish, never a rejection surface, never touches job status), isPreemptRequested (O(1) level-triggered poll), noteSteersRemoved (drain/cancel bookkeeping + fenced clear), clearPreemptRequests (empty-boundary disarm). One drain body, two boundaries: createSteerDrainHook (PostToolBatch) and createSteerPreemptBoundaryHook (PreemptBoundary) share drainAndBuildInjections, so the two injection sites cannot drift — the SDK's provider-safety argument rests on identical HumanMessage shapes. The shared body builds injections incrementally under a swallow-all catch (a mid-loop throw still injects what was applied — those parts are already persisted), clears preempt requests in finally, and disarms the generation when a boundary drains nothing. Request path: POST /chat/steer accepts preempt: true. The guard ladder is unchanged in order and in every status code. A preempt request is NEVER a rejection reason — without the capability the steer still enqueues and the 202 echoes preempt: false. Armed strictly after a successful enqueue; cancel disarms. The capability is read from the OWNING replica's recorded `preemptCapable` rather than the route replica's own SDK probe, so a rolling deploy cannot label a steer "interrupting" that the older owner will only inject at a tool step. Durable label: SteerQueueItem.preempt → TPendingSteer.preempt, so a parked/claimed/replayed chip keeps its wording. Run wiring: createRun registers the PreemptBoundary hook and threads RunConfig.preemption, both gated on isSteerPreemptSupported() — a separate probe from isSteeringSupported(), so the client affordance can never arm against an SDK that only injects at tool boundaries. buildSteerWiring builds both hooks from one shared closures object, so preemption survives HITL pause/resume for free. Honest finalization: an empty preempt boundary persists and emits with unfinished: true — the same contract an abort gets — re-marked explicitly because BaseClient has already saved the row as unfinished: false by that point. Not changed: no new job status, store method, Lua, SSE event type, endpoint, or authorization surface. abortJob, completeJob, transitionStatus, closeAndDrainSteers, getResumeState, emitChunk, applySteerPart and the whole abort path are untouched. Tests: 120 packages/api steering specs (preempt lifecycle, tombstone, fences, caps, terminal release, both-boundary drain parity, level-triggered poll, request/cancel arming, owner-capability degradation) plus 5 in api for buildSteerWiring gating, and 2 Redis-gated cross-replica transport specs.

* 🛑 feat: Preemptive Steer — server half (Interrupt & Steer, PR 2 of 3) Lets the steer route ask the generating replica to seal its live model stream at the next provider-safe boundary instead of waiting for a tool step. The run is never aborted, job status never changes, the partial answer is kept, and generation resumes in the same assistant message after the injected steer. Consumes the SDK seam in @librechat/agents (LibreChat-AI/agents#335, #346). Transport: IEventTransport gains a fenced emitPreempt/onPreempt pair beside abort. RedisEventTransport fans PREEMPT out on the SAME events channel and subscription (no new connection, key, or subscribe call); onPreempt returns a registration-scoped unsubscribe with the same replacement-safe state-identity guard onAbort uses. InMemory implements neither — single-process preempt lives entirely in the runtime set. Runtime state: RuntimeJobState carries the per-generation request set, createdAt-fenced and capped at STEER_QUEUE_MAX_DEPTH, plus a bounded `cleared` tombstone so a late cross-replica arm cannot resurrect a request whose steer already drained. registerPreemptSubscription mirrors the abort registration's double fence (runtime identity + generation createdAt); releaseAbortSubscription retires BOTH listeners and the armed set, so every terminal path drops preempt state for free. Public surface: requestPreempt (arm + fenced publish, never a rejection surface, never touches job status), isPreemptRequested (O(1) level-triggered poll), noteSteersRemoved (drain/cancel bookkeeping + fenced clear), clearPreemptRequests (empty-boundary disarm). One drain body, two boundaries: createSteerDrainHook (PostToolBatch) and createSteerPreemptBoundaryHook (PreemptBoundary) share drainAndBuildInjections, so the two injection sites cannot drift — the SDK's provider-safety argument rests on identical HumanMessage shapes. The shared body builds injections incrementally under a swallow-all catch (a mid-loop throw still injects what was applied — those parts are already persisted), clears preempt requests in finally, and disarms the generation when a boundary drains nothing. Request path: POST /chat/steer accepts preempt: true. The guard ladder is unchanged in order and in every status code. A preempt request is NEVER a rejection reason — without the capability the steer still enqueues and the 202 echoes preempt: false. Armed strictly after a successful enqueue; cancel disarms. The capability is read from the OWNING replica's recorded `preemptCapable` rather than the route replica's own SDK probe, so a rolling deploy cannot label a steer "interrupting" that the older owner will only inject at a tool step. Durable label: SteerQueueItem.preempt → TPendingSteer.preempt, so a parked/claimed/replayed chip keeps its wording. Run wiring: createRun registers the PreemptBoundary hook and threads RunConfig.preemption, both gated on isSteerPreemptSupported() — a separate probe from isSteeringSupported(), so the client affordance can never arm against an SDK that only injects at tool boundaries. buildSteerWiring builds both hooks from one shared closures object, so preemption survives HITL pause/resume for free. Honest finalization: an empty preempt boundary persists and emits with unfinished: true — the same contract an abort gets — re-marked explicitly because BaseClient has already saved the row as unfinished: false by that point. Not changed: no new job status, store method, Lua, SSE event type, endpoint, or authorization surface. abortJob, completeJob, transitionStatus, closeAndDrainSteers, getResumeState, emitChunk, applySteerPart and the whole abort path are untouched. Tests: 120 packages/api steering specs (preempt lifecycle, tombstone, fences, caps, terminal release, both-boundary drain parity, level-triggered poll, request/cancel arming, owner-capability degradation) plus 5 in api for buildSteerWiring gating, and 2 Redis-gated cross-replica transport specs. * 🔒 fix: Codex round 2 — evict tombstones, scope the empty-boundary disarm, honest resumes All four server findings were fresh consequences of the round-1 fixes, which is the review doing exactly what it should. - Tombstone cap refused new entries instead of evicting. Every drained or cancelled steer is tombstoned, not just preempting ones, so a generation that processed 20 steers exhausted the set and the late-arm race resurfaced silently. Now evicts oldest-first (Set iteration is insertion-ordered), with the budget named PREEMPT_TOMBSTONE_MAX rather than an inline expression. - The empty-boundary disarm I added in round 1 wiped the generation's ENTIRE armed set. A second steer can enqueue and arm between the atomic drain returning empty and the disarm running — that arm is backed by a live, uninjected queue item and must survive. The drain now snapshots the armed ids BEFORE draining (getArmedPreemptIds) and clearPreemptRequests takes an explicit id list instead of clearing everything. - HITL resume finalized with a hardcoded unfinished: false. The boundary hook is re-registered on resume via buildSteerWiring, so a resumed segment can end on an empty preempt boundary exactly like a fresh one; finalizeResumedTurn now reads getPreemptStats() and the halt reason, matching the normal request path. - Ownership moves on resume, so the job's recorded preemptCapable must describe the replica that will actually generate. Refreshed before resumeCompletion; a job created on a capable replica that resumes on an older one during a rolling deploy no longer acknowledges steers as interrupting. Tests: +3 (scoped disarm sparing a post-snapshot arm, oldest-first tombstone eviction, id-list disarm). 122 packages/api steering specs green. * 🚨 fix: Codex round 3 — deserialize preemptCapable from Redis (feature was dead under Redis) The P1 here is the most consequential defect in the whole feature, and it was introduced by round 1's own capability fix. - `RedisJobStore.serializeJob` writes booleans generically, so `preemptCapable` reached Redis — but `deserializeJob` is an EXPLICIT field map and had no line for it. Every `getJob()` therefore dropped the flag, `job.metadata.preemptCapable` was always undefined, and `handleSteerRequest` computed `preemptArmed: false` unconditionally. Interrupt & steer would have silently degraded to ordinary tool-boundary steering in EVERY Redis deployment — i.e. the feature shipping as a no-op in production while passing every in-memory test. Now deserialized, with a round-trip assertion in the metadata spec that fails (`Received: undefined`) against the unfixed store. - The resume capability refresh moved from just-before `resumeCompletion` to immediately after `approvals.resolve` claims the run. That call already flips the job back to `running`, so the steer route accepts requests from that instant; leaving the refresh 135 lines later (across the whole client reconstruction) left a real window where a steer read the PREVIOUS owner's capability. Not the fully atomic transition Codex suggested — that reaches into the approvals Lua — but it shrinks the window from seconds to one await, which is proportionate for a label-accuracy issue. Refuted: "avoid triggering preemption inside subagents". The premise — that the run-wide poll can seal a subagent stream — does not hold against the shipped SDK. Child graphs are constructed with `subagentScope: true` (SubagentExecutor) and `preemption` is NOT propagated into child inputs, while `canClaimPreemptSeal()` requires `!subagentScope && preemption != null`. Both conditions fail independently, so a subagent can never claim a seal and the boundary cannot fire with `agentId` set. The `input.agentId != null` guard in the hook is defensive depth, not the thing standing between us and the described failure. 140 packages/api specs green. * 🔁 fix: Codex round 4 — re-arm durable interrupt steers when resume moves owners - An arm lives only in the owning replica's runtime plus a transient pub/sub message, while the steer's `preempt` flag is durable on the queue item. A HITL resume landing on a different replica therefore started with an empty armed set and a poll stuck false, so an interrupt the user had already been ACKed for silently waited for an ordinary tool boundary. New `GenerationJobManager.rearmQueuedPreempts` rebuilds the armed set by peeking the durable queue (fenced on the generation) and re-arming every item flagged `preempt`; resume calls it right after claiming. Safe by construction: every item peeked is still queued, so no drained steer can be resurrected. - Capability-refresh failure now logs at error rather than warn, but deliberately does NOT fail the resume — see the reply on that thread. Tests: +2 (rebuild from queue arms only the flagged item and reports the count; a stale generation arms nothing). 124 packages/api steering specs green. * 📡 fix: Codex round 5 — acknowledge only what was actually armed - A cross-replica arm was fire-and-forget: `emitPreempt` logged its own publish failure and `requestPreempt` returned void, so the route answered `preempt: true` even when the owner never armed a poll. The steer still injected at the next tool boundary, but the chip claimed an interrupt that could not happen — and unlike HITL resume, an ordinary running generation had no durable reconciliation to recover it. `emitPreempt` now resolves to the subscriber count and rejects on failure; `requestPreempt` is async and returns whether the arm truly landed (owned locally, or delivered to at least one subscriber). The 202 reports THAT rather than what was asked for, so the chip relabels to ordinary steering exactly as it does for a capability-degraded deployment. Errors are swallowed into `false` — an unarmed interrupt is a downgrade, never a failed steer. - The owner capability is re-read immediately before enqueue rather than reused from the top of the guard ladder. `checkAgentAccess` and file resolution are awaits, so a request can span an entire HITL pause/resume that moves ownership to a replica with different capability and rewrites that very flag. Only paid for by requests that actually asked to interrupt. Tests: +3 (not-armed when the publish reaches nobody; armed when this replica owns the generation; a throwing publish downgrades instead of propagating). 127 packages/api steering specs green. * 🎯 fix: Codex round 6 — real ownership, confirmed disarms, and a CI regression of my own Three review findings plus three CI failures the round-5 commit caused. Review: - Ownership came from `runtimeState`, which a cross-replica `getJob` populates with a FACADE runtime on any replica that merely read the job. Matching `createdAt` therefore proved only "we looked at this job", so a non-owner could arm nothing and report success. Ownership now comes from `ownedJobs`, the actual owner map. - `armPreemptIds` returns how many ids it accepted, and a local arm is only reported as armed when one was. A tombstoned id (its steer drained at an ordinary boundary mid-request) no longer answers `preempt: true` for an interrupt that cannot happen. - The cancel disarm is awaited. A dropped clear is worse than a dropped arm: the owner keeps a level-triggered request for a steer that no longer exists, seals its next chunk and truncates an unrelated answer. The boundary drain's own call stays non-blocking — there the owner is local, so the disarm is already effective and awaiting the informational publish would only delay injection. - Subscriber count is NOT read as proof of owner receipt: the count includes this replica's own facade subscription. A successful publish reports armed, a rejected one does not. Documented rather than papered over — see the acknowledgement-semantics note on the PR. CI regressions from round 5, all mine: - `registerPreemptSubscription` was AWAITED at both runtime-init sites, so job creation blocked on a second Redis channel subscription and hung when that subscribe was slow. Abort is awaited because a missed abort strands a run; a missed preempt only degrades that steer to the next tool boundary, so it now registers without gating createJob. - Two api specs mocked `@librechat/api` without the newly imported `isSteerPreemptSupported`, so the call threw before createJob; and one exact-match assertion needed the new `preemptCapable` metadata field. - My own Redis integration spec asserted arm-before-clear ordering, which two publishes carry no guarantee of — the receiving tombstone exists precisely because of that. Now asserts delivery and payload fidelity, order-independent. 158 packages/api specs, 27 api specs green. * 🧭 fix: Codex round 7 — settle the acknowledgement semantics (Option A) Round 7's second finding is the incoherence I flagged on the PR: the route persisted `preempt: true` on the durable queue item while returning `preempt: false` when delivery could not be confirmed. Those two then disagreed, and `rearmQueuedPreempts` trusts the DURABLE one — so a resumed owner would honour an interrupt the client had explicitly been told degraded to ordinary steering. Rather than patch the disagreement, this settles the meaning: `preempt` in the 202 means "queued as an interrupt request", NOT "a seal is guaranteed". It mirrors `SteerQueueItem.preempt` exactly, so the response, the durable record, and the resume-time re-arm can never disagree. The gates that ARE knowable stay — the owner's recorded capability and a successful enqueue. Everything past that degrades to the documented fallback of injecting at the next tool boundary. A route cannot synchronously know whether another replica will seal: proving it needs a correlated request/response over pub-sub, and even that only proves the owner heard, not that it is still streaming when the arm lands. Four rounds of tightening this boolean each surfaced a narrower case; the sequence does not converge, so the invariant is now "the flag describes the durable decision" and an unconfirmed arm logs a warning instead of rewriting the answer. Also from this round: a failed disarm publish is retried once and its outcome reported. `handleSteerCancel` keeps `removed: true` — the steer really did leave the queue, and saying otherwise would make the client re-show a chip for a steer that can never arrive — and adds `disarmed: false` so the residual risk is visible rather than swallowed. Damage stays bounded regardless: the empty-boundary self-clear disarms the generation after a single seal. Tests: +1 pinning the response/durable-flag invariant. 159 packages/api specs green. * 🧹 fix: Codex round 8 — remove the unverifiable disarm signal Round 8 found the same over-promise on the disarm side that round 7 corrected on the arm side, so this applies the same answer rather than patching around it. The `disarmed: false` field added in round 7 was both unreliable and unused: a resolved publish is not proof the owner heard it (the delivery count includes this replica's own facade subscription), and it was never threaded into `CancelSteerResponse` or read by any client. A signal that claims a certainty the transport cannot provide is worse than no signal — it invites callers to trust it. Removed from the response. The retry stays, because it genuinely reduces the failure rate, and `noteSteersRemoved` still returns whether the publish succeeded FOR LOGGING, now documented explicitly as "published without error", not "the owner disarmed". Disarm is best effort with a bounded, self-healing failure: if the clear is lost the owner seals once, the empty-boundary self-clear disarms the generation, and the turn is persisted `unfinished: true` rather than silently truncated. Tightening that further needs a correlated request/response over pub-sub with a timeout — noted on the PR as the deliberate boundary of this design rather than an oversight. 130 packages/api steering specs green. * 🧽 fix: Codex round 9 — spend snapshot arms on nonempty drains too The round-6 scoping fix only cleared the pre-drain snapshot when the drain came back EMPTY. On a nonempty drain the `finally` cleared just the drained ids, so a stale arm — typically a cancel whose cross-replica clear was lost — survived the boundary. It would then immediately seal the continuation meant to answer the steer that had just been injected, and land on an empty boundary as `preempt_incomplete`: the interrupt appears to work, and the answer to it is truncated. A boundary that runs has spent its seal, so everything armed at snapshot time is spent whether or not it came back from the drain. The `finally` now clears the union of the snapshot and the drained ids. Arms that land AFTER the snapshot are still spared — their queue items are live and uninjected, which is the property round 6 added. Also fixes an api-workspace CI failure of mine: `resume.spec.js` stubs `GenerationJobManager` wholesale, and the round-3/4 resume work added two calls (`updateMetadata`, `rearmQueuedPreempts`) the stub did not define, so 34 specs threw. Stub extended. Tests: +2 (a nonempty drain clears a stale snapshot arm; a nonempty drain spares an arm that landed mid-drain). Counterfactually verified — the stale-arm spec fails against the unfixed drain. 132 packages/api specs, 60 resume specs green. * fix: never let a failed preempt subscription reject into the void registerPreemptSubscription is called detached at both sites, so a rejected Redis SUBSCRIBE became an unhandled rejection — process-fatal under Node's default --unhandled-rejections=throw. The comment already promised this path merely degrades steering; it now does. Swallowed and logged inside the registration rather than at each call site, so a future third caller cannot reintroduce the trap. Losing the channel costs this generation's cross-replica preempts, not the server: same-replica arming is runtime state and still works, and remote arms fall back to the next tool boundary. Verified counterfactually — the new spec surfaces SUBSCRIBE failed as an unhandled rejection against the unfixed registration. * docs: state the real blast radius of a failed preempt subscription LibreChat's own entrypoints install a global unhandledRejection handler that logs and keeps serving, so the escaping rejection this guards was never fatal to this server — only to another consumer of @librechat/api that installs no handler. The fix stands either way; the comment just should not overstate what it prevents. * test: cover the cross-replica preempt hop with two manager instances Every other preempt test runs against a single manager, so the hop that actually carries an interrupt in production had no coverage: the steer POST lands on whichever replica the balancer picks, which is usually not the one generating. Non-owner publishes, owner arms, owner's level-triggered poll flips — none of that was exercised end to end. Two GenerationJobManagerClass instances are a faithful replica pair here. runtimeState and ownedJobs are private instance fields, there is no module-level mutable state between them, and createStreamServices duplicates a dedicated subscriber connection per call, so separate OS processes would exercise the same objects over the same Redis. Both assertions verified counterfactually against real Redis: - Deleting the preemptCapable deserialization in RedisJobStore fails this with 'Expected: true, Received: undefined' — the exact P1 that shipped past every in-memory test and would have made the feature a silent no-op on every Redis deployment. - Dropping the non-owner arm publish fails it with 'Received: false'. * test: remove the fixed sleeps and vacuity from the cross-replica preempt test Codex round 11, both findings, both on the test I added last commit. P2 — the 300ms waits were load-bearing. Redis pub/sub never replays and the owner's SUBSCRIBE is detached, so on a slow CI worker the publish could land before anyone was listening and the test would fail against correct code. Now it republishes until the owner's state converges, which is safe because arms and clears are idempotent set writes keyed by steerId. Side effect: the tests got ~10x faster (85ms/57ms vs 929ms/606ms) since they finish on delivery rather than on a timer. P3 — afterEach destroyed only the transports, leaving each manager alive in its own cleanup-interval closure, still working against a dead transport. Now tracks the managers and awaits destroy(), which disposes the job store and its timer too. Matches how the rest of this file cleans up. Fixing the sleeps exposed a third problem codex did not flag: the stale-arm test could pass vacuously, because an undelivered arm and a fenced one look identical. It now brackets the stale publish between two control arms — the first proves the owner is listening before the stale one is sent, the second proves it has had its chance to arrive. Verified counterfactually against real Redis, and stable over 5 runs: - dropping the preemptCapable deserialization fails with 'Received: undefined' - dropping the non-owner arm publish times out both tests - removing the generation fence fails the stale test with ["control-before", "steer-stale", "control-after"] — which also confirms the bracketing orders as intended rather than by luck * fix: gate interrupt on the OWNER's capability alone, not the route's Codex round 12. The comment above this gate already said 'the OWNER's recorded capability, not this replica's probe' — and then the code ANDed in isSteerPreemptSupported(), which is exactly this replica's probe. The contradiction dates to the original commit; round 6 made the gate owner-scoped and wrote that comment without removing the local conjunct. The route never seals. It enqueues and publishes an arm, neither of which touches the SDK, so during a rolling deploy a steer landing on an un-upgraded replica silently lost its interrupt even though the owner could seal. When the route IS the owner the probe is redundant anyway: the flag it would consult is the one this process wrote at createJob. The real degradation path is unchanged and still tested — an owner that recorded no capability relabels to an ordinary steer. The test that pinned the local probe asserted an impossible same-replica state (capable metadata plus an incapable local SDK, when the metadata is written from that probe); it now pins the mixed-SDK direction instead, and fails with 'Expected: true, Received: false' if the probe is put back. * fix: reconcile arms at handover, and stop holding the 202 on a publish Codex round 13, two of three findings. P2 — rearmQueuedPreempts only ever ADDED. A replica that merely read the job still installs a facade runtime and subscribes, so it can accept an arm and then miss the best-effort clear that follows the drain. HITL resume promotes that facade to owner, the union keeps the orphan, and the first resumed stream seals on a steer no longer in the queue, drains nothing, and truncates the resumed answer as preempt_incomplete. acquireResumedJobOwnership only sets ownedJobs, so nothing else was clearing it. The durable queue is the sole authority at a handover: arms it does not back are now disarmed and tombstoned, so an in-flight publish cannot revive them either. Worth recording that my own independent review raised this and my verifier refuted it. Codex found it separately; two reviewers converging should have outweighed one refutation. P2 — the route awaited the arm publish before answering. The 202 reports capability, not delivery, so the await could not change the response; it only exposed the caller to Redis latency after the queue item was already durable. A client that times out and retries mints a second steer while the first stays queued, injecting the same instruction twice, whereas a lost publish merely takes the tool-boundary fallback. Detached, with both outcomes logged. All three tests verified counterfactually: union-only rearm fails the two new handover specs, and re-awaiting the publish hangs the stalled-publish spec until jest kills it. * fix: snapshot arms before reading the queue at handover Codex round 14 — a regression from my own round-13 fix, and a worse failure than the one it corrected. Round 13 read the durable queue first, then tombstoned any armed id the snapshot did not back. But approvals.resolve reopens steering before reconciliation runs, so another replica can commit a preempt steer and publish its arm while the peek is in flight. That arm is then present locally but absent from a snapshot taken before the steer existed, so a LIVE interrupt the route already acknowledged got dropped — and tombstoned, which blocks the re-arm, making it unrecoverable rather than merely late. Fixed by inverting the two reads rather than by locking or paying a second round trip. A steer is durably enqueued BEFORE its arm is published, so any id in an arms-first snapshot was already queued when it was armed, and the later peek must observe it unless it has since drained — which is exactly the orphan this reconciliation exists to drop. Arms landing after the snapshot are simply not candidates. Also re-checks runtime identity across the await, since the generation can be replaced while the queue read is in flight. New spec injects a steer + arm during the peek and verifies it survives; against the round-13 ordering it fails with Received array: []. * fix: bound the cancel disarm wait and fence enqueue to its generation Codex round 15. P2 — the cancel awaited its disarm publish unbounded. ioredis queues commands during an outage rather than rejecting, so that await could hang for the length of the outage with the steer ALREADY durably cancelled; a client that gives up then treats the cancel as failed and restores a chip for a steer that can never produce an applied event. Every successful cancel publishes, so ordinary steers were exposed too, not only preemptive ones. Now bounded at 1s, with the publish continuing behind it — its retry and logging are unchanged, it is just no longer in front of the response. This is the sibling of round 13's arm-publish finding; I fixed one path and left this one. P3 — enqueue was not fenced to the generation the capability decision was made against. The access checks, file resolution and owner re-read are all awaits, so the run can be replaced before the enqueue: the item then lands on the REPLACEMENT queue while the durable preempt flag and the arm still name the previous epoch, the arm is fenced out at the owner, and the 202 promises an interrupt that cannot happen. enqueueSteer now takes an expected generation, mirroring drain/peek, and the Redis path enforces it inside STEER_ENQUEUE_LUA so the check is atomic with the push rather than racing it. All three new specs verified counterfactually, including the Lua guard against real Redis (removing it returns 1 where -1 is required). * fix: fence the steer to its authorized generation, bound resume setup, keep preempt when Redis parks Codex round 16, all three findings. P2 — round 15 fenced the enqueue to owner.createdAt, the RE-READ job. Every guard above it (ownership, tenant, paused-state, agent ACL) ran against the job read at the top, so if the run was replaced during those awaits the fence happily accepted the steer into a generation the request was never authorized against, carrying the wrong agent's metadata. Now rejects on any mismatch between the validated job and the re-read. P2 — resume awaited its steering bookkeeping unbounded, after approvals.resolve had consumed the action and flipped the job to running, and outside the resume lifecycle's own try/finally. .catch does not fire on a promise that never settles, which is what ioredis produces during an outage, so the client times out, its retry gets a 409 for a spent action, and no cleanup runs. Bounded at 1s with the writes finishing in the background. P3 — Redis parks leftover steers inside its terminal-transition Lua, which projects item fields one by one, so preempt was silently dropped and a steer recovered from /chat/status lost its interrupting label. Added to both projections. All three verified counterfactually, two against real Redis. Worth recording that my first version of the generation-mismatch test was VACUOUS — it faked a createdAt matching no live job, so the round-15 enqueue fence rejected it for the wrong reason and the test passed with the guard removed. Rewritten to replace the run for real; it now fails with 'Expected 404, Received 202'.

* 🛑 feat: Preemptive Steer — server half (Interrupt & Steer, PR 2 of 3) Lets the steer route ask the generating replica to seal its live model stream at the next provider-safe boundary instead of waiting for a tool step. The run is never aborted, job status never changes, the partial answer is kept, and generation resumes in the same assistant message after the injected steer. Consumes the SDK seam in @librechat/agents (LibreChat-AI/agents#335, #346). Transport: IEventTransport gains a fenced emitPreempt/onPreempt pair beside abort. RedisEventTransport fans PREEMPT out on the SAME events channel and subscription (no new connection, key, or subscribe call); onPreempt returns a registration-scoped unsubscribe with the same replacement-safe state-identity guard onAbort uses. InMemory implements neither — single-process preempt lives entirely in the runtime set. Runtime state: RuntimeJobState carries the per-generation request set, createdAt-fenced and capped at STEER_QUEUE_MAX_DEPTH, plus a bounded `cleared` tombstone so a late cross-replica arm cannot resurrect a request whose steer already drained. registerPreemptSubscription mirrors the abort registration's double fence (runtime identity + generation createdAt); releaseAbortSubscription retires BOTH listeners and the armed set, so every terminal path drops preempt state for free. Public surface: requestPreempt (arm + fenced publish, never a rejection surface, never touches job status), isPreemptRequested (O(1) level-triggered poll), noteSteersRemoved (drain/cancel bookkeeping + fenced clear), clearPreemptRequests (empty-boundary disarm). One drain body, two boundaries: createSteerDrainHook (PostToolBatch) and createSteerPreemptBoundaryHook (PreemptBoundary) share drainAndBuildInjections, so the two injection sites cannot drift — the SDK's provider-safety argument rests on identical HumanMessage shapes. The shared body builds injections incrementally under a swallow-all catch (a mid-loop throw still injects what was applied — those parts are already persisted), clears preempt requests in finally, and disarms the generation when a boundary drains nothing. Request path: POST /chat/steer accepts preempt: true. The guard ladder is unchanged in order and in every status code. A preempt request is NEVER a rejection reason — without the capability the steer still enqueues and the 202 echoes preempt: false. Armed strictly after a successful enqueue; cancel disarms. The capability is read from the OWNING replica's recorded `preemptCapable` rather than the route replica's own SDK probe, so a rolling deploy cannot label a steer "interrupting" that the older owner will only inject at a tool step. Durable label: SteerQueueItem.preempt → TPendingSteer.preempt, so a parked/claimed/replayed chip keeps its wording. Run wiring: createRun registers the PreemptBoundary hook and threads RunConfig.preemption, both gated on isSteerPreemptSupported() — a separate probe from isSteeringSupported(), so the client affordance can never arm against an SDK that only injects at tool boundaries. buildSteerWiring builds both hooks from one shared closures object, so preemption survives HITL pause/resume for free. Honest finalization: an empty preempt boundary persists and emits with unfinished: true — the same contract an abort gets — re-marked explicitly because BaseClient has already saved the row as unfinished: false by that point. Not changed: no new job status, store method, Lua, SSE event type, endpoint, or authorization surface. abortJob, completeJob, transitionStatus, closeAndDrainSteers, getResumeState, emitChunk, applySteerPart and the whole abort path are untouched. Tests: 120 packages/api steering specs (preempt lifecycle, tombstone, fences, caps, terminal release, both-boundary drain parity, level-triggered poll, request/cancel arming, owner-capability degradation) plus 5 in api for buildSteerWiring gating, and 2 Redis-gated cross-replica transport specs. * 🔒 fix: Codex round 2 — evict tombstones, scope the empty-boundary disarm, honest resumes All four server findings were fresh consequences of the round-1 fixes, which is the review doing exactly what it should. - Tombstone cap refused new entries instead of evicting. Every drained or cancelled steer is tombstoned, not just preempting ones, so a generation that processed 20 steers exhausted the set and the late-arm race resurfaced silently. Now evicts oldest-first (Set iteration is insertion-ordered), with the budget named PREEMPT_TOMBSTONE_MAX rather than an inline expression. - The empty-boundary disarm I added in round 1 wiped the generation's ENTIRE armed set. A second steer can enqueue and arm between the atomic drain returning empty and the disarm running — that arm is backed by a live, uninjected queue item and must survive. The drain now snapshots the armed ids BEFORE draining (getArmedPreemptIds) and clearPreemptRequests takes an explicit id list instead of clearing everything. - HITL resume finalized with a hardcoded unfinished: false. The boundary hook is re-registered on resume via buildSteerWiring, so a resumed segment can end on an empty preempt boundary exactly like a fresh one; finalizeResumedTurn now reads getPreemptStats() and the halt reason, matching the normal request path. - Ownership moves on resume, so the job's recorded preemptCapable must describe the replica that will actually generate. Refreshed before resumeCompletion; a job created on a capable replica that resumes on an older one during a rolling deploy no longer acknowledges steers as interrupting. Tests: +3 (scoped disarm sparing a post-snapshot arm, oldest-first tombstone eviction, id-list disarm). 122 packages/api steering specs green. * 🚨 fix: Codex round 3 — deserialize preemptCapable from Redis (feature was dead under Redis) The P1 here is the most consequential defect in the whole feature, and it was introduced by round 1's own capability fix. - `RedisJobStore.serializeJob` writes booleans generically, so `preemptCapable` reached Redis — but `deserializeJob` is an EXPLICIT field map and had no line for it. Every `getJob()` therefore dropped the flag, `job.metadata.preemptCapable` was always undefined, and `handleSteerRequest` computed `preemptArmed: false` unconditionally. Interrupt & steer would have silently degraded to ordinary tool-boundary steering in EVERY Redis deployment — i.e. the feature shipping as a no-op in production while passing every in-memory test. Now deserialized, with a round-trip assertion in the metadata spec that fails (`Received: undefined`) against the unfixed store. - The resume capability refresh moved from just-before `resumeCompletion` to immediately after `approvals.resolve` claims the run. That call already flips the job back to `running`, so the steer route accepts requests from that instant; leaving the refresh 135 lines later (across the whole client reconstruction) left a real window where a steer read the PREVIOUS owner's capability. Not the fully atomic transition Codex suggested — that reaches into the approvals Lua — but it shrinks the window from seconds to one await, which is proportionate for a label-accuracy issue. Refuted: "avoid triggering preemption inside subagents". The premise — that the run-wide poll can seal a subagent stream — does not hold against the shipped SDK. Child graphs are constructed with `subagentScope: true` (SubagentExecutor) and `preemption` is NOT propagated into child inputs, while `canClaimPreemptSeal()` requires `!subagentScope && preemption != null`. Both conditions fail independently, so a subagent can never claim a seal and the boundary cannot fire with `agentId` set. The `input.agentId != null` guard in the hook is defensive depth, not the thing standing between us and the described failure. 140 packages/api specs green. * 🔁 fix: Codex round 4 — re-arm durable interrupt steers when resume moves owners - An arm lives only in the owning replica's runtime plus a transient pub/sub message, while the steer's `preempt` flag is durable on the queue item. A HITL resume landing on a different replica therefore started with an empty armed set and a poll stuck false, so an interrupt the user had already been ACKed for silently waited for an ordinary tool boundary. New `GenerationJobManager.rearmQueuedPreempts` rebuilds the armed set by peeking the durable queue (fenced on the generation) and re-arming every item flagged `preempt`; resume calls it right after claiming. Safe by construction: every item peeked is still queued, so no drained steer can be resurrected. - Capability-refresh failure now logs at error rather than warn, but deliberately does NOT fail the resume — see the reply on that thread. Tests: +2 (rebuild from queue arms only the flagged item and reports the count; a stale generation arms nothing). 124 packages/api steering specs green. * 📡 fix: Codex round 5 — acknowledge only what was actually armed - A cross-replica arm was fire-and-forget: `emitPreempt` logged its own publish failure and `requestPreempt` returned void, so the route answered `preempt: true` even when the owner never armed a poll. The steer still injected at the next tool boundary, but the chip claimed an interrupt that could not happen — and unlike HITL resume, an ordinary running generation had no durable reconciliation to recover it. `emitPreempt` now resolves to the subscriber count and rejects on failure; `requestPreempt` is async and returns whether the arm truly landed (owned locally, or delivered to at least one subscriber). The 202 reports THAT rather than what was asked for, so the chip relabels to ordinary steering exactly as it does for a capability-degraded deployment. Errors are swallowed into `false` — an unarmed interrupt is a downgrade, never a failed steer. - The owner capability is re-read immediately before enqueue rather than reused from the top of the guard ladder. `checkAgentAccess` and file resolution are awaits, so a request can span an entire HITL pause/resume that moves ownership to a replica with different capability and rewrites that very flag. Only paid for by requests that actually asked to interrupt. Tests: +3 (not-armed when the publish reaches nobody; armed when this replica owns the generation; a throwing publish downgrades instead of propagating). 127 packages/api steering specs green. * 🎯 fix: Codex round 6 — real ownership, confirmed disarms, and a CI regression of my own Three review findings plus three CI failures the round-5 commit caused. Review: - Ownership came from `runtimeState`, which a cross-replica `getJob` populates with a FACADE runtime on any replica that merely read the job. Matching `createdAt` therefore proved only "we looked at this job", so a non-owner could arm nothing and report success. Ownership now comes from `ownedJobs`, the actual owner map. - `armPreemptIds` returns how many ids it accepted, and a local arm is only reported as armed when one was. A tombstoned id (its steer drained at an ordinary boundary mid-request) no longer answers `preempt: true` for an interrupt that cannot happen. - The cancel disarm is awaited. A dropped clear is worse than a dropped arm: the owner keeps a level-triggered request for a steer that no longer exists, seals its next chunk and truncates an unrelated answer. The boundary drain's own call stays non-blocking — there the owner is local, so the disarm is already effective and awaiting the informational publish would only delay injection. - Subscriber count is NOT read as proof of owner receipt: the count includes this replica's own facade subscription. A successful publish reports armed, a rejected one does not. Documented rather than papered over — see the acknowledgement-semantics note on the PR. CI regressions from round 5, all mine: - `registerPreemptSubscription` was AWAITED at both runtime-init sites, so job creation blocked on a second Redis channel subscription and hung when that subscribe was slow. Abort is awaited because a missed abort strands a run; a missed preempt only degrades that steer to the next tool boundary, so it now registers without gating createJob. - Two api specs mocked `@librechat/api` without the newly imported `isSteerPreemptSupported`, so the call threw before createJob; and one exact-match assertion needed the new `preemptCapable` metadata field. - My own Redis integration spec asserted arm-before-clear ordering, which two publishes carry no guarantee of — the receiving tombstone exists precisely because of that. Now asserts delivery and payload fidelity, order-independent. 158 packages/api specs, 27 api specs green. * 🧭 fix: Codex round 7 — settle the acknowledgement semantics (Option A) Round 7's second finding is the incoherence I flagged on the PR: the route persisted `preempt: true` on the durable queue item while returning `preempt: false` when delivery could not be confirmed. Those two then disagreed, and `rearmQueuedPreempts` trusts the DURABLE one — so a resumed owner would honour an interrupt the client had explicitly been told degraded to ordinary steering. Rather than patch the disagreement, this settles the meaning: `preempt` in the 202 means "queued as an interrupt request", NOT "a seal is guaranteed". It mirrors `SteerQueueItem.preempt` exactly, so the response, the durable record, and the resume-time re-arm can never disagree. The gates that ARE knowable stay — the owner's recorded capability and a successful enqueue. Everything past that degrades to the documented fallback of injecting at the next tool boundary. A route cannot synchronously know whether another replica will seal: proving it needs a correlated request/response over pub-sub, and even that only proves the owner heard, not that it is still streaming when the arm lands. Four rounds of tightening this boolean each surfaced a narrower case; the sequence does not converge, so the invariant is now "the flag describes the durable decision" and an unconfirmed arm logs a warning instead of rewriting the answer. Also from this round: a failed disarm publish is retried once and its outcome reported. `handleSteerCancel` keeps `removed: true` — the steer really did leave the queue, and saying otherwise would make the client re-show a chip for a steer that can never arrive — and adds `disarmed: false` so the residual risk is visible rather than swallowed. Damage stays bounded regardless: the empty-boundary self-clear disarms the generation after a single seal. Tests: +1 pinning the response/durable-flag invariant. 159 packages/api specs green. * 🧹 fix: Codex round 8 — remove the unverifiable disarm signal Round 8 found the same over-promise on the disarm side that round 7 corrected on the arm side, so this applies the same answer rather than patching around it. The `disarmed: false` field added in round 7 was both unreliable and unused: a resolved publish is not proof the owner heard it (the delivery count includes this replica's own facade subscription), and it was never threaded into `CancelSteerResponse` or read by any client. A signal that claims a certainty the transport cannot provide is worse than no signal — it invites callers to trust it. Removed from the response. The retry stays, because it genuinely reduces the failure rate, and `noteSteersRemoved` still returns whether the publish succeeded FOR LOGGING, now documented explicitly as "published without error", not "the owner disarmed". Disarm is best effort with a bounded, self-healing failure: if the clear is lost the owner seals once, the empty-boundary self-clear disarms the generation, and the turn is persisted `unfinished: true` rather than silently truncated. Tightening that further needs a correlated request/response over pub-sub with a timeout — noted on the PR as the deliberate boundary of this design rather than an oversight. 130 packages/api steering specs green. * 🧽 fix: Codex round 9 — spend snapshot arms on nonempty drains too The round-6 scoping fix only cleared the pre-drain snapshot when the drain came back EMPTY. On a nonempty drain the `finally` cleared just the drained ids, so a stale arm — typically a cancel whose cross-replica clear was lost — survived the boundary. It would then immediately seal the continuation meant to answer the steer that had just been injected, and land on an empty boundary as `preempt_incomplete`: the interrupt appears to work, and the answer to it is truncated. A boundary that runs has spent its seal, so everything armed at snapshot time is spent whether or not it came back from the drain. The `finally` now clears the union of the snapshot and the drained ids. Arms that land AFTER the snapshot are still spared — their queue items are live and uninjected, which is the property round 6 added. Also fixes an api-workspace CI failure of mine: `resume.spec.js` stubs `GenerationJobManager` wholesale, and the round-3/4 resume work added two calls (`updateMetadata`, `rearmQueuedPreempts`) the stub did not define, so 34 specs threw. Stub extended. Tests: +2 (a nonempty drain clears a stale snapshot arm; a nonempty drain spares an arm that landed mid-drain). Counterfactually verified — the stale-arm spec fails against the unfixed drain. 132 packages/api specs, 60 resume specs green. * fix: never let a failed preempt subscription reject into the void registerPreemptSubscription is called detached at both sites, so a rejected Redis SUBSCRIBE became an unhandled rejection — process-fatal under Node's default --unhandled-rejections=throw. The comment already promised this path merely degrades steering; it now does. Swallowed and logged inside the registration rather than at each call site, so a future third caller cannot reintroduce the trap. Losing the channel costs this generation's cross-replica preempts, not the server: same-replica arming is runtime state and still works, and remote arms fall back to the next tool boundary. Verified counterfactually — the new spec surfaces SUBSCRIBE failed as an unhandled rejection against the unfixed registration. * docs: state the real blast radius of a failed preempt subscription LibreChat's own entrypoints install a global unhandledRejection handler that logs and keeps serving, so the escaping rejection this guards was never fatal to this server — only to another consumer of @librechat/api that installs no handler. The fix stands either way; the comment just should not overstate what it prevents. * test: cover the cross-replica preempt hop with two manager instances Every other preempt test runs against a single manager, so the hop that actually carries an interrupt in production had no coverage: the steer POST lands on whichever replica the balancer picks, which is usually not the one generating. Non-owner publishes, owner arms, owner's level-triggered poll flips — none of that was exercised end to end. Two GenerationJobManagerClass instances are a faithful replica pair here. runtimeState and ownedJobs are private instance fields, there is no module-level mutable state between them, and createStreamServices duplicates a dedicated subscriber connection per call, so separate OS processes would exercise the same objects over the same Redis. Both assertions verified counterfactually against real Redis: - Deleting the preemptCapable deserialization in RedisJobStore fails this with 'Expected: true, Received: undefined' — the exact P1 that shipped past every in-memory test and would have made the feature a silent no-op on every Redis deployment. - Dropping the non-owner arm publish fails it with 'Received: false'. * test: remove the fixed sleeps and vacuity from the cross-replica preempt test Codex round 11, both findings, both on the test I added last commit. P2 — the 300ms waits were load-bearing. Redis pub/sub never replays and the owner's SUBSCRIBE is detached, so on a slow CI worker the publish could land before anyone was listening and the test would fail against correct code. Now it republishes until the owner's state converges, which is safe because arms and clears are idempotent set writes keyed by steerId. Side effect: the tests got ~10x faster (85ms/57ms vs 929ms/606ms) since they finish on delivery rather than on a timer. P3 — afterEach destroyed only the transports, leaving each manager alive in its own cleanup-interval closure, still working against a dead transport. Now tracks the managers and awaits destroy(), which disposes the job store and its timer too. Matches how the rest of this file cleans up. Fixing the sleeps exposed a third problem codex did not flag: the stale-arm test could pass vacuously, because an undelivered arm and a fenced one look identical. It now brackets the stale publish between two control arms — the first proves the owner is listening before the stale one is sent, the second proves it has had its chance to arrive. Verified counterfactually against real Redis, and stable over 5 runs: - dropping the preemptCapable deserialization fails with 'Received: undefined' - dropping the non-owner arm publish times out both tests - removing the generation fence fails the stale test with ["control-before", "steer-stale", "control-after"] — which also confirms the bracketing orders as intended rather than by luck * fix: gate interrupt on the OWNER's capability alone, not the route's Codex round 12. The comment above this gate already said 'the OWNER's recorded capability, not this replica's probe' — and then the code ANDed in isSteerPreemptSupported(), which is exactly this replica's probe. The contradiction dates to the original commit; round 6 made the gate owner-scoped and wrote that comment without removing the local conjunct. The route never seals. It enqueues and publishes an arm, neither of which touches the SDK, so during a rolling deploy a steer landing on an un-upgraded replica silently lost its interrupt even though the owner could seal. When the route IS the owner the probe is redundant anyway: the flag it would consult is the one this process wrote at createJob. The real degradation path is unchanged and still tested — an owner that recorded no capability relabels to an ordinary steer. The test that pinned the local probe asserted an impossible same-replica state (capable metadata plus an incapable local SDK, when the metadata is written from that probe); it now pins the mixed-SDK direction instead, and fails with 'Expected: true, Received: false' if the probe is put back. * fix: reconcile arms at handover, and stop holding the 202 on a publish Codex round 13, two of three findings. P2 — rearmQueuedPreempts only ever ADDED. A replica that merely read the job still installs a facade runtime and subscribes, so it can accept an arm and then miss the best-effort clear that follows the drain. HITL resume promotes that facade to owner, the union keeps the orphan, and the first resumed stream seals on a steer no longer in the queue, drains nothing, and truncates the resumed answer as preempt_incomplete. acquireResumedJobOwnership only sets ownedJobs, so nothing else was clearing it. The durable queue is the sole authority at a handover: arms it does not back are now disarmed and tombstoned, so an in-flight publish cannot revive them either. Worth recording that my own independent review raised this and my verifier refuted it. Codex found it separately; two reviewers converging should have outweighed one refutation. P2 — the route awaited the arm publish before answering. The 202 reports capability, not delivery, so the await could not change the response; it only exposed the caller to Redis latency after the queue item was already durable. A client that times out and retries mints a second steer while the first stays queued, injecting the same instruction twice, whereas a lost publish merely takes the tool-boundary fallback. Detached, with both outcomes logged. All three tests verified counterfactually: union-only rearm fails the two new handover specs, and re-awaiting the publish hangs the stalled-publish spec until jest kills it. * fix: snapshot arms before reading the queue at handover Codex round 14 — a regression from my own round-13 fix, and a worse failure than the one it corrected. Round 13 read the durable queue first, then tombstoned any armed id the snapshot did not back. But approvals.resolve reopens steering before reconciliation runs, so another replica can commit a preempt steer and publish its arm while the peek is in flight. That arm is then present locally but absent from a snapshot taken before the steer existed, so a LIVE interrupt the route already acknowledged got dropped — and tombstoned, which blocks the re-arm, making it unrecoverable rather than merely late. Fixed by inverting the two reads rather than by locking or paying a second round trip. A steer is durably enqueued BEFORE its arm is published, so any id in an arms-first snapshot was already queued when it was armed, and the later peek must observe it unless it has since drained — which is exactly the orphan this reconciliation exists to drop. Arms landing after the snapshot are simply not candidates. Also re-checks runtime identity across the await, since the generation can be replaced while the queue read is in flight. New spec injects a steer + arm during the peek and verifies it survives; against the round-13 ordering it fails with Received array: []. * fix: bound the cancel disarm wait and fence enqueue to its generation Codex round 15. P2 — the cancel awaited its disarm publish unbounded. ioredis queues commands during an outage rather than rejecting, so that await could hang for the length of the outage with the steer ALREADY durably cancelled; a client that gives up then treats the cancel as failed and restores a chip for a steer that can never produce an applied event. Every successful cancel publishes, so ordinary steers were exposed too, not only preemptive ones. Now bounded at 1s, with the publish continuing behind it — its retry and logging are unchanged, it is just no longer in front of the response. This is the sibling of round 13's arm-publish finding; I fixed one path and left this one. P3 — enqueue was not fenced to the generation the capability decision was made against. The access checks, file resolution and owner re-read are all awaits, so the run can be replaced before the enqueue: the item then lands on the REPLACEMENT queue while the durable preempt flag and the arm still name the previous epoch, the arm is fenced out at the owner, and the 202 promises an interrupt that cannot happen. enqueueSteer now takes an expected generation, mirroring drain/peek, and the Redis path enforces it inside STEER_ENQUEUE_LUA so the check is atomic with the push rather than racing it. All three new specs verified counterfactually, including the Lua guard against real Redis (removing it returns 1 where -1 is required). * ⚡ feat: Interrupt & Steer — client half (PR 3 of 3) Makes preemptive steering reachable. Consumes the server contract from PR 2 (POST /chat/steer `preempt`, echoed on the 202) and the SDK seam in @librechat/agents 3.3.5. Settings shape follows the agreed correction, NOT the earlier plan draft: `steerInterruptsByDefault` is a boolean ORTHOGONAL to `duringRunDefaultAction` — that enum still chooses steer-vs-queue, the new boolean chooses how soon a steer lands. This deliberately avoids widening the enum to three values, which would have silently broken two hard-coded binary TOGGLES (`DuringRunAction.tsx`'s setter and `SteerMenu`'s `useDefaultToggleEntry`, both `prev === 'steer' ? … : …`) where a third value collapses to the wrong branch and one click erases the setting. - useSteering: `submitSteer` takes an opts bag and threads `preempt` into the POST, the optimistic chip, and the failure chip. The ACK relabels from the SERVER's echo, so a deployment that cannot seal mid-stream downgrades the chip's wording instead of erroring — the entire UX surface of capability degradation. New `interruptSteer` reuses the whole chip lifecycle and degradation ladder, and falls back to `interruptAndSend` when `!canSteer`, because steering needs a server-side job and an always-visible button would otherwise be dead for the whole first turn. `steerFromComposer` honours the new preference. - Composer: always-visible `InterruptSteerButton` with one fixed meaning (stop now, keep what's written), disabled on a paused run to pre-empt the server's 409, `type="button"` so it never steals the form's Enter submit, RTL-correct margins. A fourth hovercard row on the during-run send button, and ⌘/Ctrl+Shift+Enter routed AHEAD of the bare ⌘/Ctrl+Enter branch that would otherwise swallow it. - Chips: an in-flight preempt chip reads "Interrupting" with a ZapOff glyph; `preempt` survives reconnect through `seedSteerChips`. - `RunEnd.interruptArmed`, `drainAfterAbortByIndex`, `useQueueDrain`, `stopGenerating` and `interruptAndSend` are untouched — the preempt path deliberately shares none of the abort machinery. Tests: 7 new specs (posts preempt, turn-1 fallback, empty-text refusal, default route with and without the preference, server-echo relabel, double-click). 66 useSteering specs green; tsc and lint clean. Round-1 review fixes folded in: - P1: interrupt & steer no longer hard-aborts a run paused on tool approval. `canSteer` is false there, so the fallback was routing the keyboard and hovercard paths into `interruptAndSend` — discarding the partial answer, the exact opposite of what the action promises. The fallback is now scoped to the missing-conversation case only, and a paused run refuses outright (the standalone button was already disabled; the guard now lives where all three paths reach it). - The preference no longer leaks into the explicit Steer action. `steerFromComposer` backs both the default Enter route AND the explicit hovercard row / Ctrl+Enter alternate; applying `steerInterruptsByDefault` inside it made ordinary Steer interrupt and the two rows indistinguishable. It now takes an explicit argument that only `submitDuringRun`'s default route sets. - Retry preserves preemption: a failed interrupt-steer chip keeps `preempt: true`, and `retrySteer` now forwards it rather than silently resubmitting as an ordinary tool-boundary steer. - ⌘/Ctrl+Shift+Enter defers to a rebound submit shortcut, mirroring the bare ⌘/Ctrl+Enter branch — a user who bound submit to that chord keeps getting submit. Round-2 fix: the preempt label now survives the page-reload resume path too. `seedSteerChips` (useResumableSSE) and `restoreSteerChips` (useResumeOnLoad) are two independent TPendingSteer→PendingSteer mappers with near-identical bodies; the first carried the flag and the second silently dropped it, so an armed interrupt reverted to plain "Steering" after a reload. Swept: those are the only two in production code. The reclaim/convert paths deliberately omit it — a queued follow-up starts its own turn, so there is nothing to interrupt. * fix: yield the interrupt-steer chord only to a submit shortcut bound to it The previous guard skipped the Ctrl/Cmd+Shift+Enter branch whenever ANY submitMessage override existed. Rebinding submit to something unrelated (Ctrl+J) or unbinding it entirely then fell through to the override resolver, which returns 'none' for shifted Enter — silently removing the shortcut the hovercard still advertises. Compare the pressed chord against the configured one instead. The adjacent bare Ctrl/Cmd+Enter branch keeps its any-override guard on purpose: that chord IS the default submit chord, so once submit moves the resolver should own it. The predicate already existed inside resolveSubmitOverrideAction; pulled it out as bindingsMatch so both sites compare chords the same way. That call is behavior-preserving — eventBinding.key is 'Enter' by the early return, and equal hashes imply equal keys, so the dropped explicit key check was redundant. * fix: disable the Interrupt & steer menu row while paused on approval interruptSteer hard-refuses when the run is paused for tool approval, but the hovercard row was never gated, so it rendered enabled and clicking it did nothing at all — no chip, no queue entry, no toast — at exactly the moment a user is trying to say "stop, don't run that command". The standalone button already gates on pausedOnApproval; the row contradicted it. Gated on pausedOnApproval rather than !canSteer like the steer row above, because canSteer is also false before a conversation exists, where interruptSteer deliberately falls back to interruptAndSend and the row must stay live for the whole first turn. Tests pin both directions and were verified counterfactually: removing the gate fails the paused case, and using !canSteer fails the first-turn case. * test: render the during-run hovercard eagerly instead of driving Ariakit The new spec passed locally and failed all four cases on CI's Ubuntu and Windows shards: Ariakit's show path keys off pointer geometry, which jsdom reports as zeros, so whether a synthetic mouseEnter opens the hovercard is environment-dependent. Driving it was testing Ariakit's hover behavior, not which rows this component disables. Mocking the three Ariakit primitives renders the rows unconditionally and drops the fake timers. Both counterfactuals still fail as they should: removing the gate fails the paused case, !canSteer fails the first-turn case. * test(e2e): cover interrupt & steer sealing mid-stream The mock Playwright suite covered every sibling during-run action — steer at a tool boundary, steer degrading to a queued follow-up, queue, and interrupt & send — but not interrupt & steer, the one this stack adds. Uses E2E_SLOW_REPLY, which streams pure text with no tools, so the scenario is the same one where an ordinary steer provably degrades to a queued follow-up turn. Injecting in-thread there is something only a mid-stream seal can do, which makes the assertion discriminating rather than incidental: the steer part lands in the response, the final chunk never arrives, the text written before the seal survives, and no follow-up turn pair is created. * test(e2e): assert the run resumes after the seal, not just that it sealed The other four assertions are all satisfied by a seal that killed the run: the steer part is persisted by applySteer during the drain, before the continuation starts, so 'sealed and resumed' and 'sealed and died' were indistinguishable — and resuming is the whole difference from interrupt & send. The continuation answers the injected steer, whose text carries no fake-model marker, so getLatestUserText falls through to the default reply. That string ('E2E mock reply') is distinct from the setup turn's ('E2E reply <label>'), so seeing it proves generation restarted rather than matching text that was already on screen. The test itself is confirmed working: it ran as 104/121 in the Playwright job on 0ece357 and the run concluded success. * test(e2e): drop the resume assertion pending an unresolved question The assertion that the run visibly resumes after the seal fails deterministically in CI across all three retries. Every other assertion in the test passes, so the chord, the server preempt, the seal, and the in-thread injection all work; what fails is only the continuation's reply becoming visible. I could not determine from CI logs whether that is the mock harness not surfacing a continuation in a no-tool scenario or generation genuinely stopping after the seal, and I am not willing to weaken it into something that passes either way — that would convert a real question into false assurance. Reverted to the four assertions that hold, with the open question recorded in the test and raised on the PR. * fix: yield the interrupt chord to any bound shortcut, and move the Enter hint Codex round 1 on this PR (its first — #14519's rounds predate these files). P2 — the chord yielded only to a rebound submitMessage. This composer handler runs before the document-level one in useKeyboardShortcuts, and that one does not check defaultPrevented, so binding any composer-allowed action (focusChat, focusSearch, showShortcuts) to Ctrl/Cmd+Shift+Enter fired BOTH: the run was interrupted and the bound action ran. Now yields to any chord the user has bound. No default binding uses this chord, so it only ever yields to a deliberate rebinding. P2 — with steerInterruptsByDefault on, plain Enter routes through submitDuringRun and preempts, but the hovercard still put the ⏎ hint on the ordinary Steer row, whose click deliberately does NOT preempt. The same row advertised a key that did something else. The hint now follows the preference: ⏎ moves to Interrupt & steer, and the Steer row shows none because no key reaches it in that mode. I had declined this twice on the grounds that the behaviour split is deliberate — it is, and it is unchanged; the finding was about the label, which was a different claim and a correct one. Lint also caught a real bug in the first fix: boundShortcutChords was missing from the handler's dependency array, so a rebinding would not have taken effect until the callback was recreated for another reason. Both verified counterfactually; 179 client specs green. * fix: resolve every composer Enter chord through one decision table Codex round 2: yielding the preempt branch to a bound chord dropped execution into the bare Ctrl/Cmd branch below it, which ignores Shift, and past that into the submit tail, where isCtrlEnter is true for the chord. So rebinding focusChat, focusSearch, or showShortcuts to Ctrl/Cmd+Shift+Enter fired the alternate action or a submit AND the document-level shortcut, consuming the draft. Two rounds in a row landed in this handler because the guard chain decided "whose chord is this?" piecemeal inside individual branches, with fall-through between them. This extracts the entire pipeline into resolveComposerKeyDown (utils/shortcuts, beside resolveSubmitOverrideAction, which it absorbs as a step): one pure decision table where every verdict is terminal, so the fall-through class is gone rather than patched around. The yield rule is now the first gate before all branches: an Enter chord bound to any shortcut the document handler runs while typing (EDITING_ALLOWED_SHORTCUTS, hoisted out of the handler and shared by both dispatchers) is left entirely to that handler. This also closes the identical latent holes in the branches codex did not flag: bound Alt+Enter and Ctrl/Cmd+Enter chords double-fired the same way during a run, and the idle submit tail consumed bound chords too. A chord bound to a shortcut the document handler does NOT run while typing keeps its composer meaning; there is nothing to collide with, and yielding would just make the chord dead. Also merged the spec file round 0 added at utils/__tests__/ into the pre-existing utils/shortcuts.spec.ts (duplicate coverage at a second path) and dropped the isNonShiftEnter+filesLoading preventDefault, which was subsumed by the unconditional one below it. The table has a spec locking every verdict. Counterfactually verified: removing the yield gate fails exactly the three yield tests. Full client suite green apart from six suites that fail identically without this change (local data-provider dist drift in unrelated areas). * fix: yield Alt+Enter to a rebound submit, derive hovercard hints from the decision table Codex round 3, both findings. P2, Alt+Enter submit rebinding. The interrupt branch reserved Alt+Enter unconditionally, so rebinding submitMessage to Alt+Enter meant the user's own submit chord aborted the run and consumed the draft as interrupt & send. Now guarded with the same bindingsMatch yield the preempt branch already had; the chord falls through to the submit override resolution and submits the default action. P2, hovercard hints. The rows advertised hardcoded chords, so a chord rebound to an editing-allowed global shortcut (yielded by the decision table) or claimed by a rebound submit still appeared as Steer, Queue, Interrupt & steer, or Interrupt & send. The hovercard now asks the same decision table the composer executes what each canonical chord does and only labels a row with a chord that still triggers it. That also fixes two dishonest hints codex did not flag: with Enter-to-send off, plain Enter inserts a newline during a run (the primary row advertised it anyway) and Ctrl/Cmd+Enter submits the default action (the alternate row claimed it). The default-action hint now moves to Ctrl/Cmd+Enter in that mode. The effective bindings (submitOverride plus yielded chords) moved from useTextarea into a shared useComposerBindings hook so the handler and the hints read the same source. resolveComposerKeyDown now takes a KeyChordSource pick of the event fields it reads, letting the hovercard pass synthetic chords. Counterfactually verified: reverting the Alt guard fails the new resolver test and the hint suppression test. 353 tests across the affected suites green; the two failing hook suites (useVisibleTools, useResumableSSE) fail identically without these changes (local data-provider dist drift). * fix: never advertise a chord on a disabled hovercard row Codex round 4, one P2. The Interrupt & steer row kept showing its chord while disabled for tool approval, but pressing it reaches interruptSteer's pausedOnApproval guard and no-ops. Fixed as a render rule rather than a per-row patch: a disabled row never shows its kbd, since its action refuses the chord by the same guard that disabled it. That also covers the disabled Steer row, whose alternate-action hint had the identical hole through steerFromComposer's canSteer refusal. Counterfactually verified: reverting the render guard fails the new test.

) * 🛑 feat: Preemptive Steer — server half (Interrupt & Steer, PR 2 of 3) Lets the steer route ask the generating replica to seal its live model stream at the next provider-safe boundary instead of waiting for a tool step. The run is never aborted, job status never changes, the partial answer is kept, and generation resumes in the same assistant message after the injected steer. Consumes the SDK seam in @librechat/agents (LibreChat-AI/agents#335, LibreChat-AI#346). Transport: IEventTransport gains a fenced emitPreempt/onPreempt pair beside abort. RedisEventTransport fans PREEMPT out on the SAME events channel and subscription (no new connection, key, or subscribe call); onPreempt returns a registration-scoped unsubscribe with the same replacement-safe state-identity guard onAbort uses. InMemory implements neither — single-process preempt lives entirely in the runtime set. Runtime state: RuntimeJobState carries the per-generation request set, createdAt-fenced and capped at STEER_QUEUE_MAX_DEPTH, plus a bounded `cleared` tombstone so a late cross-replica arm cannot resurrect a request whose steer already drained. registerPreemptSubscription mirrors the abort registration's double fence (runtime identity + generation createdAt); releaseAbortSubscription retires BOTH listeners and the armed set, so every terminal path drops preempt state for free. Public surface: requestPreempt (arm + fenced publish, never a rejection surface, never touches job status), isPreemptRequested (O(1) level-triggered poll), noteSteersRemoved (drain/cancel bookkeeping + fenced clear), clearPreemptRequests (empty-boundary disarm). One drain body, two boundaries: createSteerDrainHook (PostToolBatch) and createSteerPreemptBoundaryHook (PreemptBoundary) share drainAndBuildInjections, so the two injection sites cannot drift — the SDK's provider-safety argument rests on identical HumanMessage shapes. The shared body builds injections incrementally under a swallow-all catch (a mid-loop throw still injects what was applied — those parts are already persisted), clears preempt requests in finally, and disarms the generation when a boundary drains nothing. Request path: POST /chat/steer accepts preempt: true. The guard ladder is unchanged in order and in every status code. A preempt request is NEVER a rejection reason — without the capability the steer still enqueues and the 202 echoes preempt: false. Armed strictly after a successful enqueue; cancel disarms. The capability is read from the OWNING replica's recorded `preemptCapable` rather than the route replica's own SDK probe, so a rolling deploy cannot label a steer "interrupting" that the older owner will only inject at a tool step. Durable label: SteerQueueItem.preempt → TPendingSteer.preempt, so a parked/claimed/replayed chip keeps its wording. Run wiring: createRun registers the PreemptBoundary hook and threads RunConfig.preemption, both gated on isSteerPreemptSupported() — a separate probe from isSteeringSupported(), so the client affordance can never arm against an SDK that only injects at tool boundaries. buildSteerWiring builds both hooks from one shared closures object, so preemption survives HITL pause/resume for free. Honest finalization: an empty preempt boundary persists and emits with unfinished: true — the same contract an abort gets — re-marked explicitly because BaseClient has already saved the row as unfinished: false by that point. Not changed: no new job status, store method, Lua, SSE event type, endpoint, or authorization surface. abortJob, completeJob, transitionStatus, closeAndDrainSteers, getResumeState, emitChunk, applySteerPart and the whole abort path are untouched. Tests: 120 packages/api steering specs (preempt lifecycle, tombstone, fences, caps, terminal release, both-boundary drain parity, level-triggered poll, request/cancel arming, owner-capability degradation) plus 5 in api for buildSteerWiring gating, and 2 Redis-gated cross-replica transport specs. * 🔒 fix: Codex round 2 — evict tombstones, scope the empty-boundary disarm, honest resumes All four server findings were fresh consequences of the round-1 fixes, which is the review doing exactly what it should. - Tombstone cap refused new entries instead of evicting. Every drained or cancelled steer is tombstoned, not just preempting ones, so a generation that processed 20 steers exhausted the set and the late-arm race resurfaced silently. Now evicts oldest-first (Set iteration is insertion-ordered), with the budget named PREEMPT_TOMBSTONE_MAX rather than an inline expression. - The empty-boundary disarm I added in round 1 wiped the generation's ENTIRE armed set. A second steer can enqueue and arm between the atomic drain returning empty and the disarm running — that arm is backed by a live, uninjected queue item and must survive. The drain now snapshots the armed ids BEFORE draining (getArmedPreemptIds) and clearPreemptRequests takes an explicit id list instead of clearing everything. - HITL resume finalized with a hardcoded unfinished: false. The boundary hook is re-registered on resume via buildSteerWiring, so a resumed segment can end on an empty preempt boundary exactly like a fresh one; finalizeResumedTurn now reads getPreemptStats() and the halt reason, matching the normal request path. - Ownership moves on resume, so the job's recorded preemptCapable must describe the replica that will actually generate. Refreshed before resumeCompletion; a job created on a capable replica that resumes on an older one during a rolling deploy no longer acknowledges steers as interrupting. Tests: +3 (scoped disarm sparing a post-snapshot arm, oldest-first tombstone eviction, id-list disarm). 122 packages/api steering specs green. * 🚨 fix: Codex round 3 — deserialize preemptCapable from Redis (feature was dead under Redis) The P1 here is the most consequential defect in the whole feature, and it was introduced by round 1's own capability fix. - `RedisJobStore.serializeJob` writes booleans generically, so `preemptCapable` reached Redis — but `deserializeJob` is an EXPLICIT field map and had no line for it. Every `getJob()` therefore dropped the flag, `job.metadata.preemptCapable` was always undefined, and `handleSteerRequest` computed `preemptArmed: false` unconditionally. Interrupt & steer would have silently degraded to ordinary tool-boundary steering in EVERY Redis deployment — i.e. the feature shipping as a no-op in production while passing every in-memory test. Now deserialized, with a round-trip assertion in the metadata spec that fails (`Received: undefined`) against the unfixed store. - The resume capability refresh moved from just-before `resumeCompletion` to immediately after `approvals.resolve` claims the run. That call already flips the job back to `running`, so the steer route accepts requests from that instant; leaving the refresh 135 lines later (across the whole client reconstruction) left a real window where a steer read the PREVIOUS owner's capability. Not the fully atomic transition Codex suggested — that reaches into the approvals Lua — but it shrinks the window from seconds to one await, which is proportionate for a label-accuracy issue. Refuted: "avoid triggering preemption inside subagents". The premise — that the run-wide poll can seal a subagent stream — does not hold against the shipped SDK. Child graphs are constructed with `subagentScope: true` (SubagentExecutor) and `preemption` is NOT propagated into child inputs, while `canClaimPreemptSeal()` requires `!subagentScope && preemption != null`. Both conditions fail independently, so a subagent can never claim a seal and the boundary cannot fire with `agentId` set. The `input.agentId != null` guard in the hook is defensive depth, not the thing standing between us and the described failure. 140 packages/api specs green. * 🔁 fix: Codex round 4 — re-arm durable interrupt steers when resume moves owners - An arm lives only in the owning replica's runtime plus a transient pub/sub message, while the steer's `preempt` flag is durable on the queue item. A HITL resume landing on a different replica therefore started with an empty armed set and a poll stuck false, so an interrupt the user had already been ACKed for silently waited for an ordinary tool boundary. New `GenerationJobManager.rearmQueuedPreempts` rebuilds the armed set by peeking the durable queue (fenced on the generation) and re-arming every item flagged `preempt`; resume calls it right after claiming. Safe by construction: every item peeked is still queued, so no drained steer can be resurrected. - Capability-refresh failure now logs at error rather than warn, but deliberately does NOT fail the resume — see the reply on that thread. Tests: +2 (rebuild from queue arms only the flagged item and reports the count; a stale generation arms nothing). 124 packages/api steering specs green. * 📡 fix: Codex round 5 — acknowledge only what was actually armed - A cross-replica arm was fire-and-forget: `emitPreempt` logged its own publish failure and `requestPreempt` returned void, so the route answered `preempt: true` even when the owner never armed a poll. The steer still injected at the next tool boundary, but the chip claimed an interrupt that could not happen — and unlike HITL resume, an ordinary running generation had no durable reconciliation to recover it. `emitPreempt` now resolves to the subscriber count and rejects on failure; `requestPreempt` is async and returns whether the arm truly landed (owned locally, or delivered to at least one subscriber). The 202 reports THAT rather than what was asked for, so the chip relabels to ordinary steering exactly as it does for a capability-degraded deployment. Errors are swallowed into `false` — an unarmed interrupt is a downgrade, never a failed steer. - The owner capability is re-read immediately before enqueue rather than reused from the top of the guard ladder. `checkAgentAccess` and file resolution are awaits, so a request can span an entire HITL pause/resume that moves ownership to a replica with different capability and rewrites that very flag. Only paid for by requests that actually asked to interrupt. Tests: +3 (not-armed when the publish reaches nobody; armed when this replica owns the generation; a throwing publish downgrades instead of propagating). 127 packages/api steering specs green. * 🎯 fix: Codex round 6 — real ownership, confirmed disarms, and a CI regression of my own Three review findings plus three CI failures the round-5 commit caused. Review: - Ownership came from `runtimeState`, which a cross-replica `getJob` populates with a FACADE runtime on any replica that merely read the job. Matching `createdAt` therefore proved only "we looked at this job", so a non-owner could arm nothing and report success. Ownership now comes from `ownedJobs`, the actual owner map. - `armPreemptIds` returns how many ids it accepted, and a local arm is only reported as armed when one was. A tombstoned id (its steer drained at an ordinary boundary mid-request) no longer answers `preempt: true` for an interrupt that cannot happen. - The cancel disarm is awaited. A dropped clear is worse than a dropped arm: the owner keeps a level-triggered request for a steer that no longer exists, seals its next chunk and truncates an unrelated answer. The boundary drain's own call stays non-blocking — there the owner is local, so the disarm is already effective and awaiting the informational publish would only delay injection. - Subscriber count is NOT read as proof of owner receipt: the count includes this replica's own facade subscription. A successful publish reports armed, a rejected one does not. Documented rather than papered over — see the acknowledgement-semantics note on the PR. CI regressions from round 5, all mine: - `registerPreemptSubscription` was AWAITED at both runtime-init sites, so job creation blocked on a second Redis channel subscription and hung when that subscribe was slow. Abort is awaited because a missed abort strands a run; a missed preempt only degrades that steer to the next tool boundary, so it now registers without gating createJob. - Two api specs mocked `@librechat/api` without the newly imported `isSteerPreemptSupported`, so the call threw before createJob; and one exact-match assertion needed the new `preemptCapable` metadata field. - My own Redis integration spec asserted arm-before-clear ordering, which two publishes carry no guarantee of — the receiving tombstone exists precisely because of that. Now asserts delivery and payload fidelity, order-independent. 158 packages/api specs, 27 api specs green. * 🧭 fix: Codex round 7 — settle the acknowledgement semantics (Option A) Round 7's second finding is the incoherence I flagged on the PR: the route persisted `preempt: true` on the durable queue item while returning `preempt: false` when delivery could not be confirmed. Those two then disagreed, and `rearmQueuedPreempts` trusts the DURABLE one — so a resumed owner would honour an interrupt the client had explicitly been told degraded to ordinary steering. Rather than patch the disagreement, this settles the meaning: `preempt` in the 202 means "queued as an interrupt request", NOT "a seal is guaranteed". It mirrors `SteerQueueItem.preempt` exactly, so the response, the durable record, and the resume-time re-arm can never disagree. The gates that ARE knowable stay — the owner's recorded capability and a successful enqueue. Everything past that degrades to the documented fallback of injecting at the next tool boundary. A route cannot synchronously know whether another replica will seal: proving it needs a correlated request/response over pub-sub, and even that only proves the owner heard, not that it is still streaming when the arm lands. Four rounds of tightening this boolean each surfaced a narrower case; the sequence does not converge, so the invariant is now "the flag describes the durable decision" and an unconfirmed arm logs a warning instead of rewriting the answer. Also from this round: a failed disarm publish is retried once and its outcome reported. `handleSteerCancel` keeps `removed: true` — the steer really did leave the queue, and saying otherwise would make the client re-show a chip for a steer that can never arrive — and adds `disarmed: false` so the residual risk is visible rather than swallowed. Damage stays bounded regardless: the empty-boundary self-clear disarms the generation after a single seal. Tests: +1 pinning the response/durable-flag invariant. 159 packages/api specs green. * 🧹 fix: Codex round 8 — remove the unverifiable disarm signal Round 8 found the same over-promise on the disarm side that round 7 corrected on the arm side, so this applies the same answer rather than patching around it. The `disarmed: false` field added in round 7 was both unreliable and unused: a resolved publish is not proof the owner heard it (the delivery count includes this replica's own facade subscription), and it was never threaded into `CancelSteerResponse` or read by any client. A signal that claims a certainty the transport cannot provide is worse than no signal — it invites callers to trust it. Removed from the response. The retry stays, because it genuinely reduces the failure rate, and `noteSteersRemoved` still returns whether the publish succeeded FOR LOGGING, now documented explicitly as "published without error", not "the owner disarmed". Disarm is best effort with a bounded, self-healing failure: if the clear is lost the owner seals once, the empty-boundary self-clear disarms the generation, and the turn is persisted `unfinished: true` rather than silently truncated. Tightening that further needs a correlated request/response over pub-sub with a timeout — noted on the PR as the deliberate boundary of this design rather than an oversight. 130 packages/api steering specs green. * 🧽 fix: Codex round 9 — spend snapshot arms on nonempty drains too The round-6 scoping fix only cleared the pre-drain snapshot when the drain came back EMPTY. On a nonempty drain the `finally` cleared just the drained ids, so a stale arm — typically a cancel whose cross-replica clear was lost — survived the boundary. It would then immediately seal the continuation meant to answer the steer that had just been injected, and land on an empty boundary as `preempt_incomplete`: the interrupt appears to work, and the answer to it is truncated. A boundary that runs has spent its seal, so everything armed at snapshot time is spent whether or not it came back from the drain. The `finally` now clears the union of the snapshot and the drained ids. Arms that land AFTER the snapshot are still spared — their queue items are live and uninjected, which is the property round 6 added. Also fixes an api-workspace CI failure of mine: `resume.spec.js` stubs `GenerationJobManager` wholesale, and the round-3/4 resume work added two calls (`updateMetadata`, `rearmQueuedPreempts`) the stub did not define, so 34 specs threw. Stub extended. Tests: +2 (a nonempty drain clears a stale snapshot arm; a nonempty drain spares an arm that landed mid-drain). Counterfactually verified — the stale-arm spec fails against the unfixed drain. 132 packages/api specs, 60 resume specs green. * fix: never let a failed preempt subscription reject into the void registerPreemptSubscription is called detached at both sites, so a rejected Redis SUBSCRIBE became an unhandled rejection — process-fatal under Node's default --unhandled-rejections=throw. The comment already promised this path merely degrades steering; it now does. Swallowed and logged inside the registration rather than at each call site, so a future third caller cannot reintroduce the trap. Losing the channel costs this generation's cross-replica preempts, not the server: same-replica arming is runtime state and still works, and remote arms fall back to the next tool boundary. Verified counterfactually — the new spec surfaces SUBSCRIBE failed as an unhandled rejection against the unfixed registration. * docs: state the real blast radius of a failed preempt subscription LibreChat's own entrypoints install a global unhandledRejection handler that logs and keeps serving, so the escaping rejection this guards was never fatal to this server — only to another consumer of @librechat/api that installs no handler. The fix stands either way; the comment just should not overstate what it prevents. * test: cover the cross-replica preempt hop with two manager instances Every other preempt test runs against a single manager, so the hop that actually carries an interrupt in production had no coverage: the steer POST lands on whichever replica the balancer picks, which is usually not the one generating. Non-owner publishes, owner arms, owner's level-triggered poll flips — none of that was exercised end to end. Two GenerationJobManagerClass instances are a faithful replica pair here. runtimeState and ownedJobs are private instance fields, there is no module-level mutable state between them, and createStreamServices duplicates a dedicated subscriber connection per call, so separate OS processes would exercise the same objects over the same Redis. Both assertions verified counterfactually against real Redis: - Deleting the preemptCapable deserialization in RedisJobStore fails this with 'Expected: true, Received: undefined' — the exact P1 that shipped past every in-memory test and would have made the feature a silent no-op on every Redis deployment. - Dropping the non-owner arm publish fails it with 'Received: false'. * test: remove the fixed sleeps and vacuity from the cross-replica preempt test Codex round 11, both findings, both on the test I added last commit. P2 — the 300ms waits were load-bearing. Redis pub/sub never replays and the owner's SUBSCRIBE is detached, so on a slow CI worker the publish could land before anyone was listening and the test would fail against correct code. Now it republishes until the owner's state converges, which is safe because arms and clears are idempotent set writes keyed by steerId. Side effect: the tests got ~10x faster (85ms/57ms vs 929ms/606ms) since they finish on delivery rather than on a timer. P3 — afterEach destroyed only the transports, leaving each manager alive in its own cleanup-interval closure, still working against a dead transport. Now tracks the managers and awaits destroy(), which disposes the job store and its timer too. Matches how the rest of this file cleans up. Fixing the sleeps exposed a third problem codex did not flag: the stale-arm test could pass vacuously, because an undelivered arm and a fenced one look identical. It now brackets the stale publish between two control arms — the first proves the owner is listening before the stale one is sent, the second proves it has had its chance to arrive. Verified counterfactually against real Redis, and stable over 5 runs: - dropping the preemptCapable deserialization fails with 'Received: undefined' - dropping the non-owner arm publish times out both tests - removing the generation fence fails the stale test with ["control-before", "steer-stale", "control-after"] — which also confirms the bracketing orders as intended rather than by luck * fix: gate interrupt on the OWNER's capability alone, not the route's Codex round 12. The comment above this gate already said 'the OWNER's recorded capability, not this replica's probe' — and then the code ANDed in isSteerPreemptSupported(), which is exactly this replica's probe. The contradiction dates to the original commit; round 6 made the gate owner-scoped and wrote that comment without removing the local conjunct. The route never seals. It enqueues and publishes an arm, neither of which touches the SDK, so during a rolling deploy a steer landing on an un-upgraded replica silently lost its interrupt even though the owner could seal. When the route IS the owner the probe is redundant anyway: the flag it would consult is the one this process wrote at createJob. The real degradation path is unchanged and still tested — an owner that recorded no capability relabels to an ordinary steer. The test that pinned the local probe asserted an impossible same-replica state (capable metadata plus an incapable local SDK, when the metadata is written from that probe); it now pins the mixed-SDK direction instead, and fails with 'Expected: true, Received: false' if the probe is put back. * fix: reconcile arms at handover, and stop holding the 202 on a publish Codex round 13, two of three findings. P2 — rearmQueuedPreempts only ever ADDED. A replica that merely read the job still installs a facade runtime and subscribes, so it can accept an arm and then miss the best-effort clear that follows the drain. HITL resume promotes that facade to owner, the union keeps the orphan, and the first resumed stream seals on a steer no longer in the queue, drains nothing, and truncates the resumed answer as preempt_incomplete. acquireResumedJobOwnership only sets ownedJobs, so nothing else was clearing it. The durable queue is the sole authority at a handover: arms it does not back are now disarmed and tombstoned, so an in-flight publish cannot revive them either. Worth recording that my own independent review raised this and my verifier refuted it. Codex found it separately; two reviewers converging should have outweighed one refutation. P2 — the route awaited the arm publish before answering. The 202 reports capability, not delivery, so the await could not change the response; it only exposed the caller to Redis latency after the queue item was already durable. A client that times out and retries mints a second steer while the first stays queued, injecting the same instruction twice, whereas a lost publish merely takes the tool-boundary fallback. Detached, with both outcomes logged. All three tests verified counterfactually: union-only rearm fails the two new handover specs, and re-awaiting the publish hangs the stalled-publish spec until jest kills it. * fix: snapshot arms before reading the queue at handover Codex round 14 — a regression from my own round-13 fix, and a worse failure than the one it corrected. Round 13 read the durable queue first, then tombstoned any armed id the snapshot did not back. But approvals.resolve reopens steering before reconciliation runs, so another replica can commit a preempt steer and publish its arm while the peek is in flight. That arm is then present locally but absent from a snapshot taken before the steer existed, so a LIVE interrupt the route already acknowledged got dropped — and tombstoned, which blocks the re-arm, making it unrecoverable rather than merely late. Fixed by inverting the two reads rather than by locking or paying a second round trip. A steer is durably enqueued BEFORE its arm is published, so any id in an arms-first snapshot was already queued when it was armed, and the later peek must observe it unless it has since drained — which is exactly the orphan this reconciliation exists to drop. Arms landing after the snapshot are simply not candidates. Also re-checks runtime identity across the await, since the generation can be replaced while the queue read is in flight. New spec injects a steer + arm during the peek and verifies it survives; against the round-13 ordering it fails with Received array: []. * fix: bound the cancel disarm wait and fence enqueue to its generation Codex round 15. P2 — the cancel awaited its disarm publish unbounded. ioredis queues commands during an outage rather than rejecting, so that await could hang for the length of the outage with the steer ALREADY durably cancelled; a client that gives up then treats the cancel as failed and restores a chip for a steer that can never produce an applied event. Every successful cancel publishes, so ordinary steers were exposed too, not only preemptive ones. Now bounded at 1s, with the publish continuing behind it — its retry and logging are unchanged, it is just no longer in front of the response. This is the sibling of round 13's arm-publish finding; I fixed one path and left this one. P3 — enqueue was not fenced to the generation the capability decision was made against. The access checks, file resolution and owner re-read are all awaits, so the run can be replaced before the enqueue: the item then lands on the REPLACEMENT queue while the durable preempt flag and the arm still name the previous epoch, the arm is fenced out at the owner, and the 202 promises an interrupt that cannot happen. enqueueSteer now takes an expected generation, mirroring drain/peek, and the Redis path enforces it inside STEER_ENQUEUE_LUA so the check is atomic with the push rather than racing it. All three new specs verified counterfactually, including the Lua guard against real Redis (removing it returns 1 where -1 is required). * fix: fence the steer to its authorized generation, bound resume setup, keep preempt when Redis parks Codex round 16, all three findings. P2 — round 15 fenced the enqueue to owner.createdAt, the RE-READ job. Every guard above it (ownership, tenant, paused-state, agent ACL) ran against the job read at the top, so if the run was replaced during those awaits the fence happily accepted the steer into a generation the request was never authorized against, carrying the wrong agent's metadata. Now rejects on any mismatch between the validated job and the re-read. P2 — resume awaited its steering bookkeeping unbounded, after approvals.resolve had consumed the action and flipped the job to running, and outside the resume lifecycle's own try/finally. .catch does not fire on a promise that never settles, which is what ioredis produces during an outage, so the client times out, its retry gets a 409 for a spent action, and no cleanup runs. Bounded at 1s with the writes finishing in the background. P3 — Redis parks leftover steers inside its terminal-transition Lua, which projects item fields one by one, so preempt was silently dropped and a steer recovered from /chat/status lost its interrupting label. Added to both projections. All three verified counterfactually, two against real Redis. Worth recording that my first version of the generation-mismatch test was VACUOUS — it faked a createdAt matching no live job, so the round-15 enqueue fence rejected it for the wrong reason and the test passed with the guard removed. Rewritten to replace the run for real; it now fails with 'Expected 404, Received 202'.

* 🛑 feat: Preemptive Steer — server half (Interrupt & Steer, PR 2 of 3) Lets the steer route ask the generating replica to seal its live model stream at the next provider-safe boundary instead of waiting for a tool step. The run is never aborted, job status never changes, the partial answer is kept, and generation resumes in the same assistant message after the injected steer. Consumes the SDK seam in @librechat/agents (LibreChat-AI/agents#335, LibreChat-AI#346). Transport: IEventTransport gains a fenced emitPreempt/onPreempt pair beside abort. RedisEventTransport fans PREEMPT out on the SAME events channel and subscription (no new connection, key, or subscribe call); onPreempt returns a registration-scoped unsubscribe with the same replacement-safe state-identity guard onAbort uses. InMemory implements neither — single-process preempt lives entirely in the runtime set. Runtime state: RuntimeJobState carries the per-generation request set, createdAt-fenced and capped at STEER_QUEUE_MAX_DEPTH, plus a bounded `cleared` tombstone so a late cross-replica arm cannot resurrect a request whose steer already drained. registerPreemptSubscription mirrors the abort registration's double fence (runtime identity + generation createdAt); releaseAbortSubscription retires BOTH listeners and the armed set, so every terminal path drops preempt state for free. Public surface: requestPreempt (arm + fenced publish, never a rejection surface, never touches job status), isPreemptRequested (O(1) level-triggered poll), noteSteersRemoved (drain/cancel bookkeeping + fenced clear), clearPreemptRequests (empty-boundary disarm). One drain body, two boundaries: createSteerDrainHook (PostToolBatch) and createSteerPreemptBoundaryHook (PreemptBoundary) share drainAndBuildInjections, so the two injection sites cannot drift — the SDK's provider-safety argument rests on identical HumanMessage shapes. The shared body builds injections incrementally under a swallow-all catch (a mid-loop throw still injects what was applied — those parts are already persisted), clears preempt requests in finally, and disarms the generation when a boundary drains nothing. Request path: POST /chat/steer accepts preempt: true. The guard ladder is unchanged in order and in every status code. A preempt request is NEVER a rejection reason — without the capability the steer still enqueues and the 202 echoes preempt: false. Armed strictly after a successful enqueue; cancel disarms. The capability is read from the OWNING replica's recorded `preemptCapable` rather than the route replica's own SDK probe, so a rolling deploy cannot label a steer "interrupting" that the older owner will only inject at a tool step. Durable label: SteerQueueItem.preempt → TPendingSteer.preempt, so a parked/claimed/replayed chip keeps its wording. Run wiring: createRun registers the PreemptBoundary hook and threads RunConfig.preemption, both gated on isSteerPreemptSupported() — a separate probe from isSteeringSupported(), so the client affordance can never arm against an SDK that only injects at tool boundaries. buildSteerWiring builds both hooks from one shared closures object, so preemption survives HITL pause/resume for free. Honest finalization: an empty preempt boundary persists and emits with unfinished: true — the same contract an abort gets — re-marked explicitly because BaseClient has already saved the row as unfinished: false by that point. Not changed: no new job status, store method, Lua, SSE event type, endpoint, or authorization surface. abortJob, completeJob, transitionStatus, closeAndDrainSteers, getResumeState, emitChunk, applySteerPart and the whole abort path are untouched. Tests: 120 packages/api steering specs (preempt lifecycle, tombstone, fences, caps, terminal release, both-boundary drain parity, level-triggered poll, request/cancel arming, owner-capability degradation) plus 5 in api for buildSteerWiring gating, and 2 Redis-gated cross-replica transport specs. * 🔒 fix: Codex round 2 — evict tombstones, scope the empty-boundary disarm, honest resumes All four server findings were fresh consequences of the round-1 fixes, which is the review doing exactly what it should. - Tombstone cap refused new entries instead of evicting. Every drained or cancelled steer is tombstoned, not just preempting ones, so a generation that processed 20 steers exhausted the set and the late-arm race resurfaced silently. Now evicts oldest-first (Set iteration is insertion-ordered), with the budget named PREEMPT_TOMBSTONE_MAX rather than an inline expression. - The empty-boundary disarm I added in round 1 wiped the generation's ENTIRE armed set. A second steer can enqueue and arm between the atomic drain returning empty and the disarm running — that arm is backed by a live, uninjected queue item and must survive. The drain now snapshots the armed ids BEFORE draining (getArmedPreemptIds) and clearPreemptRequests takes an explicit id list instead of clearing everything. - HITL resume finalized with a hardcoded unfinished: false. The boundary hook is re-registered on resume via buildSteerWiring, so a resumed segment can end on an empty preempt boundary exactly like a fresh one; finalizeResumedTurn now reads getPreemptStats() and the halt reason, matching the normal request path. - Ownership moves on resume, so the job's recorded preemptCapable must describe the replica that will actually generate. Refreshed before resumeCompletion; a job created on a capable replica that resumes on an older one during a rolling deploy no longer acknowledges steers as interrupting. Tests: +3 (scoped disarm sparing a post-snapshot arm, oldest-first tombstone eviction, id-list disarm). 122 packages/api steering specs green. * 🚨 fix: Codex round 3 — deserialize preemptCapable from Redis (feature was dead under Redis) The P1 here is the most consequential defect in the whole feature, and it was introduced by round 1's own capability fix. - `RedisJobStore.serializeJob` writes booleans generically, so `preemptCapable` reached Redis — but `deserializeJob` is an EXPLICIT field map and had no line for it. Every `getJob()` therefore dropped the flag, `job.metadata.preemptCapable` was always undefined, and `handleSteerRequest` computed `preemptArmed: false` unconditionally. Interrupt & steer would have silently degraded to ordinary tool-boundary steering in EVERY Redis deployment — i.e. the feature shipping as a no-op in production while passing every in-memory test. Now deserialized, with a round-trip assertion in the metadata spec that fails (`Received: undefined`) against the unfixed store. - The resume capability refresh moved from just-before `resumeCompletion` to immediately after `approvals.resolve` claims the run. That call already flips the job back to `running`, so the steer route accepts requests from that instant; leaving the refresh 135 lines later (across the whole client reconstruction) left a real window where a steer read the PREVIOUS owner's capability. Not the fully atomic transition Codex suggested — that reaches into the approvals Lua — but it shrinks the window from seconds to one await, which is proportionate for a label-accuracy issue. Refuted: "avoid triggering preemption inside subagents". The premise — that the run-wide poll can seal a subagent stream — does not hold against the shipped SDK. Child graphs are constructed with `subagentScope: true` (SubagentExecutor) and `preemption` is NOT propagated into child inputs, while `canClaimPreemptSeal()` requires `!subagentScope && preemption != null`. Both conditions fail independently, so a subagent can never claim a seal and the boundary cannot fire with `agentId` set. The `input.agentId != null` guard in the hook is defensive depth, not the thing standing between us and the described failure. 140 packages/api specs green. * 🔁 fix: Codex round 4 — re-arm durable interrupt steers when resume moves owners - An arm lives only in the owning replica's runtime plus a transient pub/sub message, while the steer's `preempt` flag is durable on the queue item. A HITL resume landing on a different replica therefore started with an empty armed set and a poll stuck false, so an interrupt the user had already been ACKed for silently waited for an ordinary tool boundary. New `GenerationJobManager.rearmQueuedPreempts` rebuilds the armed set by peeking the durable queue (fenced on the generation) and re-arming every item flagged `preempt`; resume calls it right after claiming. Safe by construction: every item peeked is still queued, so no drained steer can be resurrected. - Capability-refresh failure now logs at error rather than warn, but deliberately does NOT fail the resume — see the reply on that thread. Tests: +2 (rebuild from queue arms only the flagged item and reports the count; a stale generation arms nothing). 124 packages/api steering specs green. * 📡 fix: Codex round 5 — acknowledge only what was actually armed - A cross-replica arm was fire-and-forget: `emitPreempt` logged its own publish failure and `requestPreempt` returned void, so the route answered `preempt: true` even when the owner never armed a poll. The steer still injected at the next tool boundary, but the chip claimed an interrupt that could not happen — and unlike HITL resume, an ordinary running generation had no durable reconciliation to recover it. `emitPreempt` now resolves to the subscriber count and rejects on failure; `requestPreempt` is async and returns whether the arm truly landed (owned locally, or delivered to at least one subscriber). The 202 reports THAT rather than what was asked for, so the chip relabels to ordinary steering exactly as it does for a capability-degraded deployment. Errors are swallowed into `false` — an unarmed interrupt is a downgrade, never a failed steer. - The owner capability is re-read immediately before enqueue rather than reused from the top of the guard ladder. `checkAgentAccess` and file resolution are awaits, so a request can span an entire HITL pause/resume that moves ownership to a replica with different capability and rewrites that very flag. Only paid for by requests that actually asked to interrupt. Tests: +3 (not-armed when the publish reaches nobody; armed when this replica owns the generation; a throwing publish downgrades instead of propagating). 127 packages/api steering specs green. * 🎯 fix: Codex round 6 — real ownership, confirmed disarms, and a CI regression of my own Three review findings plus three CI failures the round-5 commit caused. Review: - Ownership came from `runtimeState`, which a cross-replica `getJob` populates with a FACADE runtime on any replica that merely read the job. Matching `createdAt` therefore proved only "we looked at this job", so a non-owner could arm nothing and report success. Ownership now comes from `ownedJobs`, the actual owner map. - `armPreemptIds` returns how many ids it accepted, and a local arm is only reported as armed when one was. A tombstoned id (its steer drained at an ordinary boundary mid-request) no longer answers `preempt: true` for an interrupt that cannot happen. - The cancel disarm is awaited. A dropped clear is worse than a dropped arm: the owner keeps a level-triggered request for a steer that no longer exists, seals its next chunk and truncates an unrelated answer. The boundary drain's own call stays non-blocking — there the owner is local, so the disarm is already effective and awaiting the informational publish would only delay injection. - Subscriber count is NOT read as proof of owner receipt: the count includes this replica's own facade subscription. A successful publish reports armed, a rejected one does not. Documented rather than papered over — see the acknowledgement-semantics note on the PR. CI regressions from round 5, all mine: - `registerPreemptSubscription` was AWAITED at both runtime-init sites, so job creation blocked on a second Redis channel subscription and hung when that subscribe was slow. Abort is awaited because a missed abort strands a run; a missed preempt only degrades that steer to the next tool boundary, so it now registers without gating createJob. - Two api specs mocked `@librechat/api` without the newly imported `isSteerPreemptSupported`, so the call threw before createJob; and one exact-match assertion needed the new `preemptCapable` metadata field. - My own Redis integration spec asserted arm-before-clear ordering, which two publishes carry no guarantee of — the receiving tombstone exists precisely because of that. Now asserts delivery and payload fidelity, order-independent. 158 packages/api specs, 27 api specs green. * 🧭 fix: Codex round 7 — settle the acknowledgement semantics (Option A) Round 7's second finding is the incoherence I flagged on the PR: the route persisted `preempt: true` on the durable queue item while returning `preempt: false` when delivery could not be confirmed. Those two then disagreed, and `rearmQueuedPreempts` trusts the DURABLE one — so a resumed owner would honour an interrupt the client had explicitly been told degraded to ordinary steering. Rather than patch the disagreement, this settles the meaning: `preempt` in the 202 means "queued as an interrupt request", NOT "a seal is guaranteed". It mirrors `SteerQueueItem.preempt` exactly, so the response, the durable record, and the resume-time re-arm can never disagree. The gates that ARE knowable stay — the owner's recorded capability and a successful enqueue. Everything past that degrades to the documented fallback of injecting at the next tool boundary. A route cannot synchronously know whether another replica will seal: proving it needs a correlated request/response over pub-sub, and even that only proves the owner heard, not that it is still streaming when the arm lands. Four rounds of tightening this boolean each surfaced a narrower case; the sequence does not converge, so the invariant is now "the flag describes the durable decision" and an unconfirmed arm logs a warning instead of rewriting the answer. Also from this round: a failed disarm publish is retried once and its outcome reported. `handleSteerCancel` keeps `removed: true` — the steer really did leave the queue, and saying otherwise would make the client re-show a chip for a steer that can never arrive — and adds `disarmed: false` so the residual risk is visible rather than swallowed. Damage stays bounded regardless: the empty-boundary self-clear disarms the generation after a single seal. Tests: +1 pinning the response/durable-flag invariant. 159 packages/api specs green. * 🧹 fix: Codex round 8 — remove the unverifiable disarm signal Round 8 found the same over-promise on the disarm side that round 7 corrected on the arm side, so this applies the same answer rather than patching around it. The `disarmed: false` field added in round 7 was both unreliable and unused: a resolved publish is not proof the owner heard it (the delivery count includes this replica's own facade subscription), and it was never threaded into `CancelSteerResponse` or read by any client. A signal that claims a certainty the transport cannot provide is worse than no signal — it invites callers to trust it. Removed from the response. The retry stays, because it genuinely reduces the failure rate, and `noteSteersRemoved` still returns whether the publish succeeded FOR LOGGING, now documented explicitly as "published without error", not "the owner disarmed". Disarm is best effort with a bounded, self-healing failure: if the clear is lost the owner seals once, the empty-boundary self-clear disarms the generation, and the turn is persisted `unfinished: true` rather than silently truncated. Tightening that further needs a correlated request/response over pub-sub with a timeout — noted on the PR as the deliberate boundary of this design rather than an oversight. 130 packages/api steering specs green. * 🧽 fix: Codex round 9 — spend snapshot arms on nonempty drains too The round-6 scoping fix only cleared the pre-drain snapshot when the drain came back EMPTY. On a nonempty drain the `finally` cleared just the drained ids, so a stale arm — typically a cancel whose cross-replica clear was lost — survived the boundary. It would then immediately seal the continuation meant to answer the steer that had just been injected, and land on an empty boundary as `preempt_incomplete`: the interrupt appears to work, and the answer to it is truncated. A boundary that runs has spent its seal, so everything armed at snapshot time is spent whether or not it came back from the drain. The `finally` now clears the union of the snapshot and the drained ids. Arms that land AFTER the snapshot are still spared — their queue items are live and uninjected, which is the property round 6 added. Also fixes an api-workspace CI failure of mine: `resume.spec.js` stubs `GenerationJobManager` wholesale, and the round-3/4 resume work added two calls (`updateMetadata`, `rearmQueuedPreempts`) the stub did not define, so 34 specs threw. Stub extended. Tests: +2 (a nonempty drain clears a stale snapshot arm; a nonempty drain spares an arm that landed mid-drain). Counterfactually verified — the stale-arm spec fails against the unfixed drain. 132 packages/api specs, 60 resume specs green. * fix: never let a failed preempt subscription reject into the void registerPreemptSubscription is called detached at both sites, so a rejected Redis SUBSCRIBE became an unhandled rejection — process-fatal under Node's default --unhandled-rejections=throw. The comment already promised this path merely degrades steering; it now does. Swallowed and logged inside the registration rather than at each call site, so a future third caller cannot reintroduce the trap. Losing the channel costs this generation's cross-replica preempts, not the server: same-replica arming is runtime state and still works, and remote arms fall back to the next tool boundary. Verified counterfactually — the new spec surfaces SUBSCRIBE failed as an unhandled rejection against the unfixed registration. * docs: state the real blast radius of a failed preempt subscription LibreChat's own entrypoints install a global unhandledRejection handler that logs and keeps serving, so the escaping rejection this guards was never fatal to this server — only to another consumer of @librechat/api that installs no handler. The fix stands either way; the comment just should not overstate what it prevents. * test: cover the cross-replica preempt hop with two manager instances Every other preempt test runs against a single manager, so the hop that actually carries an interrupt in production had no coverage: the steer POST lands on whichever replica the balancer picks, which is usually not the one generating. Non-owner publishes, owner arms, owner's level-triggered poll flips — none of that was exercised end to end. Two GenerationJobManagerClass instances are a faithful replica pair here. runtimeState and ownedJobs are private instance fields, there is no module-level mutable state between them, and createStreamServices duplicates a dedicated subscriber connection per call, so separate OS processes would exercise the same objects over the same Redis. Both assertions verified counterfactually against real Redis: - Deleting the preemptCapable deserialization in RedisJobStore fails this with 'Expected: true, Received: undefined' — the exact P1 that shipped past every in-memory test and would have made the feature a silent no-op on every Redis deployment. - Dropping the non-owner arm publish fails it with 'Received: false'. * test: remove the fixed sleeps and vacuity from the cross-replica preempt test Codex round 11, both findings, both on the test I added last commit. P2 — the 300ms waits were load-bearing. Redis pub/sub never replays and the owner's SUBSCRIBE is detached, so on a slow CI worker the publish could land before anyone was listening and the test would fail against correct code. Now it republishes until the owner's state converges, which is safe because arms and clears are idempotent set writes keyed by steerId. Side effect: the tests got ~10x faster (85ms/57ms vs 929ms/606ms) since they finish on delivery rather than on a timer. P3 — afterEach destroyed only the transports, leaving each manager alive in its own cleanup-interval closure, still working against a dead transport. Now tracks the managers and awaits destroy(), which disposes the job store and its timer too. Matches how the rest of this file cleans up. Fixing the sleeps exposed a third problem codex did not flag: the stale-arm test could pass vacuously, because an undelivered arm and a fenced one look identical. It now brackets the stale publish between two control arms — the first proves the owner is listening before the stale one is sent, the second proves it has had its chance to arrive. Verified counterfactually against real Redis, and stable over 5 runs: - dropping the preemptCapable deserialization fails with 'Received: undefined' - dropping the non-owner arm publish times out both tests - removing the generation fence fails the stale test with ["control-before", "steer-stale", "control-after"] — which also confirms the bracketing orders as intended rather than by luck * fix: gate interrupt on the OWNER's capability alone, not the route's Codex round 12. The comment above this gate already said 'the OWNER's recorded capability, not this replica's probe' — and then the code ANDed in isSteerPreemptSupported(), which is exactly this replica's probe. The contradiction dates to the original commit; round 6 made the gate owner-scoped and wrote that comment without removing the local conjunct. The route never seals. It enqueues and publishes an arm, neither of which touches the SDK, so during a rolling deploy a steer landing on an un-upgraded replica silently lost its interrupt even though the owner could seal. When the route IS the owner the probe is redundant anyway: the flag it would consult is the one this process wrote at createJob. The real degradation path is unchanged and still tested — an owner that recorded no capability relabels to an ordinary steer. The test that pinned the local probe asserted an impossible same-replica state (capable metadata plus an incapable local SDK, when the metadata is written from that probe); it now pins the mixed-SDK direction instead, and fails with 'Expected: true, Received: false' if the probe is put back. * fix: reconcile arms at handover, and stop holding the 202 on a publish Codex round 13, two of three findings. P2 — rearmQueuedPreempts only ever ADDED. A replica that merely read the job still installs a facade runtime and subscribes, so it can accept an arm and then miss the best-effort clear that follows the drain. HITL resume promotes that facade to owner, the union keeps the orphan, and the first resumed stream seals on a steer no longer in the queue, drains nothing, and truncates the resumed answer as preempt_incomplete. acquireResumedJobOwnership only sets ownedJobs, so nothing else was clearing it. The durable queue is the sole authority at a handover: arms it does not back are now disarmed and tombstoned, so an in-flight publish cannot revive them either. Worth recording that my own independent review raised this and my verifier refuted it. Codex found it separately; two reviewers converging should have outweighed one refutation. P2 — the route awaited the arm publish before answering. The 202 reports capability, not delivery, so the await could not change the response; it only exposed the caller to Redis latency after the queue item was already durable. A client that times out and retries mints a second steer while the first stays queued, injecting the same instruction twice, whereas a lost publish merely takes the tool-boundary fallback. Detached, with both outcomes logged. All three tests verified counterfactually: union-only rearm fails the two new handover specs, and re-awaiting the publish hangs the stalled-publish spec until jest kills it. * fix: snapshot arms before reading the queue at handover Codex round 14 — a regression from my own round-13 fix, and a worse failure than the one it corrected. Round 13 read the durable queue first, then tombstoned any armed id the snapshot did not back. But approvals.resolve reopens steering before reconciliation runs, so another replica can commit a preempt steer and publish its arm while the peek is in flight. That arm is then present locally but absent from a snapshot taken before the steer existed, so a LIVE interrupt the route already acknowledged got dropped — and tombstoned, which blocks the re-arm, making it unrecoverable rather than merely late. Fixed by inverting the two reads rather than by locking or paying a second round trip. A steer is durably enqueued BEFORE its arm is published, so any id in an arms-first snapshot was already queued when it was armed, and the later peek must observe it unless it has since drained — which is exactly the orphan this reconciliation exists to drop. Arms landing after the snapshot are simply not candidates. Also re-checks runtime identity across the await, since the generation can be replaced while the queue read is in flight. New spec injects a steer + arm during the peek and verifies it survives; against the round-13 ordering it fails with Received array: []. * fix: bound the cancel disarm wait and fence enqueue to its generation Codex round 15. P2 — the cancel awaited its disarm publish unbounded. ioredis queues commands during an outage rather than rejecting, so that await could hang for the length of the outage with the steer ALREADY durably cancelled; a client that gives up then treats the cancel as failed and restores a chip for a steer that can never produce an applied event. Every successful cancel publishes, so ordinary steers were exposed too, not only preemptive ones. Now bounded at 1s, with the publish continuing behind it — its retry and logging are unchanged, it is just no longer in front of the response. This is the sibling of round 13's arm-publish finding; I fixed one path and left this one. P3 — enqueue was not fenced to the generation the capability decision was made against. The access checks, file resolution and owner re-read are all awaits, so the run can be replaced before the enqueue: the item then lands on the REPLACEMENT queue while the durable preempt flag and the arm still name the previous epoch, the arm is fenced out at the owner, and the 202 promises an interrupt that cannot happen. enqueueSteer now takes an expected generation, mirroring drain/peek, and the Redis path enforces it inside STEER_ENQUEUE_LUA so the check is atomic with the push rather than racing it. All three new specs verified counterfactually, including the Lua guard against real Redis (removing it returns 1 where -1 is required). * ⚡ feat: Interrupt & Steer — client half (PR 3 of 3) Makes preemptive steering reachable. Consumes the server contract from PR 2 (POST /chat/steer `preempt`, echoed on the 202) and the SDK seam in @librechat/agents 3.3.5. Settings shape follows the agreed correction, NOT the earlier plan draft: `steerInterruptsByDefault` is a boolean ORTHOGONAL to `duringRunDefaultAction` — that enum still chooses steer-vs-queue, the new boolean chooses how soon a steer lands. This deliberately avoids widening the enum to three values, which would have silently broken two hard-coded binary TOGGLES (`DuringRunAction.tsx`'s setter and `SteerMenu`'s `useDefaultToggleEntry`, both `prev === 'steer' ? … : …`) where a third value collapses to the wrong branch and one click erases the setting. - useSteering: `submitSteer` takes an opts bag and threads `preempt` into the POST, the optimistic chip, and the failure chip. The ACK relabels from the SERVER's echo, so a deployment that cannot seal mid-stream downgrades the chip's wording instead of erroring — the entire UX surface of capability degradation. New `interruptSteer` reuses the whole chip lifecycle and degradation ladder, and falls back to `interruptAndSend` when `!canSteer`, because steering needs a server-side job and an always-visible button would otherwise be dead for the whole first turn. `steerFromComposer` honours the new preference. - Composer: always-visible `InterruptSteerButton` with one fixed meaning (stop now, keep what's written), disabled on a paused run to pre-empt the server's 409, `type="button"` so it never steals the form's Enter submit, RTL-correct margins. A fourth hovercard row on the during-run send button, and ⌘/Ctrl+Shift+Enter routed AHEAD of the bare ⌘/Ctrl+Enter branch that would otherwise swallow it. - Chips: an in-flight preempt chip reads "Interrupting" with a ZapOff glyph; `preempt` survives reconnect through `seedSteerChips`. - `RunEnd.interruptArmed`, `drainAfterAbortByIndex`, `useQueueDrain`, `stopGenerating` and `interruptAndSend` are untouched — the preempt path deliberately shares none of the abort machinery. Tests: 7 new specs (posts preempt, turn-1 fallback, empty-text refusal, default route with and without the preference, server-echo relabel, double-click). 66 useSteering specs green; tsc and lint clean. Round-1 review fixes folded in: - P1: interrupt & steer no longer hard-aborts a run paused on tool approval. `canSteer` is false there, so the fallback was routing the keyboard and hovercard paths into `interruptAndSend` — discarding the partial answer, the exact opposite of what the action promises. The fallback is now scoped to the missing-conversation case only, and a paused run refuses outright (the standalone button was already disabled; the guard now lives where all three paths reach it). - The preference no longer leaks into the explicit Steer action. `steerFromComposer` backs both the default Enter route AND the explicit hovercard row / Ctrl+Enter alternate; applying `steerInterruptsByDefault` inside it made ordinary Steer interrupt and the two rows indistinguishable. It now takes an explicit argument that only `submitDuringRun`'s default route sets. - Retry preserves preemption: a failed interrupt-steer chip keeps `preempt: true`, and `retrySteer` now forwards it rather than silently resubmitting as an ordinary tool-boundary steer. - ⌘/Ctrl+Shift+Enter defers to a rebound submit shortcut, mirroring the bare ⌘/Ctrl+Enter branch — a user who bound submit to that chord keeps getting submit. Round-2 fix: the preempt label now survives the page-reload resume path too. `seedSteerChips` (useResumableSSE) and `restoreSteerChips` (useResumeOnLoad) are two independent TPendingSteer→PendingSteer mappers with near-identical bodies; the first carried the flag and the second silently dropped it, so an armed interrupt reverted to plain "Steering" after a reload. Swept: those are the only two in production code. The reclaim/convert paths deliberately omit it — a queued follow-up starts its own turn, so there is nothing to interrupt. * fix: yield the interrupt-steer chord only to a submit shortcut bound to it The previous guard skipped the Ctrl/Cmd+Shift+Enter branch whenever ANY submitMessage override existed. Rebinding submit to something unrelated (Ctrl+J) or unbinding it entirely then fell through to the override resolver, which returns 'none' for shifted Enter — silently removing the shortcut the hovercard still advertises. Compare the pressed chord against the configured one instead. The adjacent bare Ctrl/Cmd+Enter branch keeps its any-override guard on purpose: that chord IS the default submit chord, so once submit moves the resolver should own it. The predicate already existed inside resolveSubmitOverrideAction; pulled it out as bindingsMatch so both sites compare chords the same way. That call is behavior-preserving — eventBinding.key is 'Enter' by the early return, and equal hashes imply equal keys, so the dropped explicit key check was redundant. * fix: disable the Interrupt & steer menu row while paused on approval interruptSteer hard-refuses when the run is paused for tool approval, but the hovercard row was never gated, so it rendered enabled and clicking it did nothing at all — no chip, no queue entry, no toast — at exactly the moment a user is trying to say "stop, don't run that command". The standalone button already gates on pausedOnApproval; the row contradicted it. Gated on pausedOnApproval rather than !canSteer like the steer row above, because canSteer is also false before a conversation exists, where interruptSteer deliberately falls back to interruptAndSend and the row must stay live for the whole first turn. Tests pin both directions and were verified counterfactually: removing the gate fails the paused case, and using !canSteer fails the first-turn case. * test: render the during-run hovercard eagerly instead of driving Ariakit The new spec passed locally and failed all four cases on CI's Ubuntu and Windows shards: Ariakit's show path keys off pointer geometry, which jsdom reports as zeros, so whether a synthetic mouseEnter opens the hovercard is environment-dependent. Driving it was testing Ariakit's hover behavior, not which rows this component disables. Mocking the three Ariakit primitives renders the rows unconditionally and drops the fake timers. Both counterfactuals still fail as they should: removing the gate fails the paused case, !canSteer fails the first-turn case. * test(e2e): cover interrupt & steer sealing mid-stream The mock Playwright suite covered every sibling during-run action — steer at a tool boundary, steer degrading to a queued follow-up, queue, and interrupt & send — but not interrupt & steer, the one this stack adds. Uses E2E_SLOW_REPLY, which streams pure text with no tools, so the scenario is the same one where an ordinary steer provably degrades to a queued follow-up turn. Injecting in-thread there is something only a mid-stream seal can do, which makes the assertion discriminating rather than incidental: the steer part lands in the response, the final chunk never arrives, the text written before the seal survives, and no follow-up turn pair is created. * test(e2e): assert the run resumes after the seal, not just that it sealed The other four assertions are all satisfied by a seal that killed the run: the steer part is persisted by applySteer during the drain, before the continuation starts, so 'sealed and resumed' and 'sealed and died' were indistinguishable — and resuming is the whole difference from interrupt & send. The continuation answers the injected steer, whose text carries no fake-model marker, so getLatestUserText falls through to the default reply. That string ('E2E mock reply') is distinct from the setup turn's ('E2E reply <label>'), so seeing it proves generation restarted rather than matching text that was already on screen. The test itself is confirmed working: it ran as 104/121 in the Playwright job on 0ece357 and the run concluded success. * test(e2e): drop the resume assertion pending an unresolved question The assertion that the run visibly resumes after the seal fails deterministically in CI across all three retries. Every other assertion in the test passes, so the chord, the server preempt, the seal, and the in-thread injection all work; what fails is only the continuation's reply becoming visible. I could not determine from CI logs whether that is the mock harness not surfacing a continuation in a no-tool scenario or generation genuinely stopping after the seal, and I am not willing to weaken it into something that passes either way — that would convert a real question into false assurance. Reverted to the four assertions that hold, with the open question recorded in the test and raised on the PR. * fix: yield the interrupt chord to any bound shortcut, and move the Enter hint Codex round 1 on this PR (its first — LibreChat-AI#14519's rounds predate these files). P2 — the chord yielded only to a rebound submitMessage. This composer handler runs before the document-level one in useKeyboardShortcuts, and that one does not check defaultPrevented, so binding any composer-allowed action (focusChat, focusSearch, showShortcuts) to Ctrl/Cmd+Shift+Enter fired BOTH: the run was interrupted and the bound action ran. Now yields to any chord the user has bound. No default binding uses this chord, so it only ever yields to a deliberate rebinding. P2 — with steerInterruptsByDefault on, plain Enter routes through submitDuringRun and preempts, but the hovercard still put the ⏎ hint on the ordinary Steer row, whose click deliberately does NOT preempt. The same row advertised a key that did something else. The hint now follows the preference: ⏎ moves to Interrupt & steer, and the Steer row shows none because no key reaches it in that mode. I had declined this twice on the grounds that the behaviour split is deliberate — it is, and it is unchanged; the finding was about the label, which was a different claim and a correct one. Lint also caught a real bug in the first fix: boundShortcutChords was missing from the handler's dependency array, so a rebinding would not have taken effect until the callback was recreated for another reason. Both verified counterfactually; 179 client specs green. * fix: resolve every composer Enter chord through one decision table Codex round 2: yielding the preempt branch to a bound chord dropped execution into the bare Ctrl/Cmd branch below it, which ignores Shift, and past that into the submit tail, where isCtrlEnter is true for the chord. So rebinding focusChat, focusSearch, or showShortcuts to Ctrl/Cmd+Shift+Enter fired the alternate action or a submit AND the document-level shortcut, consuming the draft. Two rounds in a row landed in this handler because the guard chain decided "whose chord is this?" piecemeal inside individual branches, with fall-through between them. This extracts the entire pipeline into resolveComposerKeyDown (utils/shortcuts, beside resolveSubmitOverrideAction, which it absorbs as a step): one pure decision table where every verdict is terminal, so the fall-through class is gone rather than patched around. The yield rule is now the first gate before all branches: an Enter chord bound to any shortcut the document handler runs while typing (EDITING_ALLOWED_SHORTCUTS, hoisted out of the handler and shared by both dispatchers) is left entirely to that handler. This also closes the identical latent holes in the branches codex did not flag: bound Alt+Enter and Ctrl/Cmd+Enter chords double-fired the same way during a run, and the idle submit tail consumed bound chords too. A chord bound to a shortcut the document handler does NOT run while typing keeps its composer meaning; there is nothing to collide with, and yielding would just make the chord dead. Also merged the spec file round 0 added at utils/__tests__/ into the pre-existing utils/shortcuts.spec.ts (duplicate coverage at a second path) and dropped the isNonShiftEnter+filesLoading preventDefault, which was subsumed by the unconditional one below it. The table has a spec locking every verdict. Counterfactually verified: removing the yield gate fails exactly the three yield tests. Full client suite green apart from six suites that fail identically without this change (local data-provider dist drift in unrelated areas). * fix: yield Alt+Enter to a rebound submit, derive hovercard hints from the decision table Codex round 3, both findings. P2, Alt+Enter submit rebinding. The interrupt branch reserved Alt+Enter unconditionally, so rebinding submitMessage to Alt+Enter meant the user's own submit chord aborted the run and consumed the draft as interrupt & send. Now guarded with the same bindingsMatch yield the preempt branch already had; the chord falls through to the submit override resolution and submits the default action. P2, hovercard hints. The rows advertised hardcoded chords, so a chord rebound to an editing-allowed global shortcut (yielded by the decision table) or claimed by a rebound submit still appeared as Steer, Queue, Interrupt & steer, or Interrupt & send. The hovercard now asks the same decision table the composer executes what each canonical chord does and only labels a row with a chord that still triggers it. That also fixes two dishonest hints codex did not flag: with Enter-to-send off, plain Enter inserts a newline during a run (the primary row advertised it anyway) and Ctrl/Cmd+Enter submits the default action (the alternate row claimed it). The default-action hint now moves to Ctrl/Cmd+Enter in that mode. The effective bindings (submitOverride plus yielded chords) moved from useTextarea into a shared useComposerBindings hook so the handler and the hints read the same source. resolveComposerKeyDown now takes a KeyChordSource pick of the event fields it reads, letting the hovercard pass synthetic chords. Counterfactually verified: reverting the Alt guard fails the new resolver test and the hint suppression test. 353 tests across the affected suites green; the two failing hook suites (useVisibleTools, useResumableSSE) fail identically without these changes (local data-provider dist drift). * fix: never advertise a chord on a disabled hovercard row Codex round 4, one P2. The Interrupt & steer row kept showing its chord while disabled for tool approval, but pressing it reaches interruptSteer's pausedOnApproval guard and no-ops. Fixed as a render rule rather than a per-row patch: a disabled row never shows its kbd, since its action refuses the chord by the same guard that disabled it. That also covers the disabled Steer row, whose alternate-action hint had the identical hole through steerFromComposer's canSteer refusal. Counterfactually verified: reverting the render guard fails the new test.

) * 🛑 feat: Preemptive Steer — server half (Interrupt & Steer, PR 2 of 3) Lets the steer route ask the generating replica to seal its live model stream at the next provider-safe boundary instead of waiting for a tool step. The run is never aborted, job status never changes, the partial answer is kept, and generation resumes in the same assistant message after the injected steer. Consumes the SDK seam in @librechat/agents (LibreChat-AI/agents#335, LibreChat-AI#346). Transport: IEventTransport gains a fenced emitPreempt/onPreempt pair beside abort. RedisEventTransport fans PREEMPT out on the SAME events channel and subscription (no new connection, key, or subscribe call); onPreempt returns a registration-scoped unsubscribe with the same replacement-safe state-identity guard onAbort uses. InMemory implements neither — single-process preempt lives entirely in the runtime set. Runtime state: RuntimeJobState carries the per-generation request set, createdAt-fenced and capped at STEER_QUEUE_MAX_DEPTH, plus a bounded `cleared` tombstone so a late cross-replica arm cannot resurrect a request whose steer already drained. registerPreemptSubscription mirrors the abort registration's double fence (runtime identity + generation createdAt); releaseAbortSubscription retires BOTH listeners and the armed set, so every terminal path drops preempt state for free. Public surface: requestPreempt (arm + fenced publish, never a rejection surface, never touches job status), isPreemptRequested (O(1) level-triggered poll), noteSteersRemoved (drain/cancel bookkeeping + fenced clear), clearPreemptRequests (empty-boundary disarm). One drain body, two boundaries: createSteerDrainHook (PostToolBatch) and createSteerPreemptBoundaryHook (PreemptBoundary) share drainAndBuildInjections, so the two injection sites cannot drift — the SDK's provider-safety argument rests on identical HumanMessage shapes. The shared body builds injections incrementally under a swallow-all catch (a mid-loop throw still injects what was applied — those parts are already persisted), clears preempt requests in finally, and disarms the generation when a boundary drains nothing. Request path: POST /chat/steer accepts preempt: true. The guard ladder is unchanged in order and in every status code. A preempt request is NEVER a rejection reason — without the capability the steer still enqueues and the 202 echoes preempt: false. Armed strictly after a successful enqueue; cancel disarms. The capability is read from the OWNING replica's recorded `preemptCapable` rather than the route replica's own SDK probe, so a rolling deploy cannot label a steer "interrupting" that the older owner will only inject at a tool step. Durable label: SteerQueueItem.preempt → TPendingSteer.preempt, so a parked/claimed/replayed chip keeps its wording. Run wiring: createRun registers the PreemptBoundary hook and threads RunConfig.preemption, both gated on isSteerPreemptSupported() — a separate probe from isSteeringSupported(), so the client affordance can never arm against an SDK that only injects at tool boundaries. buildSteerWiring builds both hooks from one shared closures object, so preemption survives HITL pause/resume for free. Honest finalization: an empty preempt boundary persists and emits with unfinished: true — the same contract an abort gets — re-marked explicitly because BaseClient has already saved the row as unfinished: false by that point. Not changed: no new job status, store method, Lua, SSE event type, endpoint, or authorization surface. abortJob, completeJob, transitionStatus, closeAndDrainSteers, getResumeState, emitChunk, applySteerPart and the whole abort path are untouched. Tests: 120 packages/api steering specs (preempt lifecycle, tombstone, fences, caps, terminal release, both-boundary drain parity, level-triggered poll, request/cancel arming, owner-capability degradation) plus 5 in api for buildSteerWiring gating, and 2 Redis-gated cross-replica transport specs. * 🔒 fix: Codex round 2 — evict tombstones, scope the empty-boundary disarm, honest resumes All four server findings were fresh consequences of the round-1 fixes, which is the review doing exactly what it should. - Tombstone cap refused new entries instead of evicting. Every drained or cancelled steer is tombstoned, not just preempting ones, so a generation that processed 20 steers exhausted the set and the late-arm race resurfaced silently. Now evicts oldest-first (Set iteration is insertion-ordered), with the budget named PREEMPT_TOMBSTONE_MAX rather than an inline expression. - The empty-boundary disarm I added in round 1 wiped the generation's ENTIRE armed set. A second steer can enqueue and arm between the atomic drain returning empty and the disarm running — that arm is backed by a live, uninjected queue item and must survive. The drain now snapshots the armed ids BEFORE draining (getArmedPreemptIds) and clearPreemptRequests takes an explicit id list instead of clearing everything. - HITL resume finalized with a hardcoded unfinished: false. The boundary hook is re-registered on resume via buildSteerWiring, so a resumed segment can end on an empty preempt boundary exactly like a fresh one; finalizeResumedTurn now reads getPreemptStats() and the halt reason, matching the normal request path. - Ownership moves on resume, so the job's recorded preemptCapable must describe the replica that will actually generate. Refreshed before resumeCompletion; a job created on a capable replica that resumes on an older one during a rolling deploy no longer acknowledges steers as interrupting. Tests: +3 (scoped disarm sparing a post-snapshot arm, oldest-first tombstone eviction, id-list disarm). 122 packages/api steering specs green. * 🚨 fix: Codex round 3 — deserialize preemptCapable from Redis (feature was dead under Redis) The P1 here is the most consequential defect in the whole feature, and it was introduced by round 1's own capability fix. - `RedisJobStore.serializeJob` writes booleans generically, so `preemptCapable` reached Redis — but `deserializeJob` is an EXPLICIT field map and had no line for it. Every `getJob()` therefore dropped the flag, `job.metadata.preemptCapable` was always undefined, and `handleSteerRequest` computed `preemptArmed: false` unconditionally. Interrupt & steer would have silently degraded to ordinary tool-boundary steering in EVERY Redis deployment — i.e. the feature shipping as a no-op in production while passing every in-memory test. Now deserialized, with a round-trip assertion in the metadata spec that fails (`Received: undefined`) against the unfixed store. - The resume capability refresh moved from just-before `resumeCompletion` to immediately after `approvals.resolve` claims the run. That call already flips the job back to `running`, so the steer route accepts requests from that instant; leaving the refresh 135 lines later (across the whole client reconstruction) left a real window where a steer read the PREVIOUS owner's capability. Not the fully atomic transition Codex suggested — that reaches into the approvals Lua — but it shrinks the window from seconds to one await, which is proportionate for a label-accuracy issue. Refuted: "avoid triggering preemption inside subagents". The premise — that the run-wide poll can seal a subagent stream — does not hold against the shipped SDK. Child graphs are constructed with `subagentScope: true` (SubagentExecutor) and `preemption` is NOT propagated into child inputs, while `canClaimPreemptSeal()` requires `!subagentScope && preemption != null`. Both conditions fail independently, so a subagent can never claim a seal and the boundary cannot fire with `agentId` set. The `input.agentId != null` guard in the hook is defensive depth, not the thing standing between us and the described failure. 140 packages/api specs green. * 🔁 fix: Codex round 4 — re-arm durable interrupt steers when resume moves owners - An arm lives only in the owning replica's runtime plus a transient pub/sub message, while the steer's `preempt` flag is durable on the queue item. A HITL resume landing on a different replica therefore started with an empty armed set and a poll stuck false, so an interrupt the user had already been ACKed for silently waited for an ordinary tool boundary. New `GenerationJobManager.rearmQueuedPreempts` rebuilds the armed set by peeking the durable queue (fenced on the generation) and re-arming every item flagged `preempt`; resume calls it right after claiming. Safe by construction: every item peeked is still queued, so no drained steer can be resurrected. - Capability-refresh failure now logs at error rather than warn, but deliberately does NOT fail the resume — see the reply on that thread. Tests: +2 (rebuild from queue arms only the flagged item and reports the count; a stale generation arms nothing). 124 packages/api steering specs green. * 📡 fix: Codex round 5 — acknowledge only what was actually armed - A cross-replica arm was fire-and-forget: `emitPreempt` logged its own publish failure and `requestPreempt` returned void, so the route answered `preempt: true` even when the owner never armed a poll. The steer still injected at the next tool boundary, but the chip claimed an interrupt that could not happen — and unlike HITL resume, an ordinary running generation had no durable reconciliation to recover it. `emitPreempt` now resolves to the subscriber count and rejects on failure; `requestPreempt` is async and returns whether the arm truly landed (owned locally, or delivered to at least one subscriber). The 202 reports THAT rather than what was asked for, so the chip relabels to ordinary steering exactly as it does for a capability-degraded deployment. Errors are swallowed into `false` — an unarmed interrupt is a downgrade, never a failed steer. - The owner capability is re-read immediately before enqueue rather than reused from the top of the guard ladder. `checkAgentAccess` and file resolution are awaits, so a request can span an entire HITL pause/resume that moves ownership to a replica with different capability and rewrites that very flag. Only paid for by requests that actually asked to interrupt. Tests: +3 (not-armed when the publish reaches nobody; armed when this replica owns the generation; a throwing publish downgrades instead of propagating). 127 packages/api steering specs green. * 🎯 fix: Codex round 6 — real ownership, confirmed disarms, and a CI regression of my own Three review findings plus three CI failures the round-5 commit caused. Review: - Ownership came from `runtimeState`, which a cross-replica `getJob` populates with a FACADE runtime on any replica that merely read the job. Matching `createdAt` therefore proved only "we looked at this job", so a non-owner could arm nothing and report success. Ownership now comes from `ownedJobs`, the actual owner map. - `armPreemptIds` returns how many ids it accepted, and a local arm is only reported as armed when one was. A tombstoned id (its steer drained at an ordinary boundary mid-request) no longer answers `preempt: true` for an interrupt that cannot happen. - The cancel disarm is awaited. A dropped clear is worse than a dropped arm: the owner keeps a level-triggered request for a steer that no longer exists, seals its next chunk and truncates an unrelated answer. The boundary drain's own call stays non-blocking — there the owner is local, so the disarm is already effective and awaiting the informational publish would only delay injection. - Subscriber count is NOT read as proof of owner receipt: the count includes this replica's own facade subscription. A successful publish reports armed, a rejected one does not. Documented rather than papered over — see the acknowledgement-semantics note on the PR. CI regressions from round 5, all mine: - `registerPreemptSubscription` was AWAITED at both runtime-init sites, so job creation blocked on a second Redis channel subscription and hung when that subscribe was slow. Abort is awaited because a missed abort strands a run; a missed preempt only degrades that steer to the next tool boundary, so it now registers without gating createJob. - Two api specs mocked `@librechat/api` without the newly imported `isSteerPreemptSupported`, so the call threw before createJob; and one exact-match assertion needed the new `preemptCapable` metadata field. - My own Redis integration spec asserted arm-before-clear ordering, which two publishes carry no guarantee of — the receiving tombstone exists precisely because of that. Now asserts delivery and payload fidelity, order-independent. 158 packages/api specs, 27 api specs green. * 🧭 fix: Codex round 7 — settle the acknowledgement semantics (Option A) Round 7's second finding is the incoherence I flagged on the PR: the route persisted `preempt: true` on the durable queue item while returning `preempt: false` when delivery could not be confirmed. Those two then disagreed, and `rearmQueuedPreempts` trusts the DURABLE one — so a resumed owner would honour an interrupt the client had explicitly been told degraded to ordinary steering. Rather than patch the disagreement, this settles the meaning: `preempt` in the 202 means "queued as an interrupt request", NOT "a seal is guaranteed". It mirrors `SteerQueueItem.preempt` exactly, so the response, the durable record, and the resume-time re-arm can never disagree. The gates that ARE knowable stay — the owner's recorded capability and a successful enqueue. Everything past that degrades to the documented fallback of injecting at the next tool boundary. A route cannot synchronously know whether another replica will seal: proving it needs a correlated request/response over pub-sub, and even that only proves the owner heard, not that it is still streaming when the arm lands. Four rounds of tightening this boolean each surfaced a narrower case; the sequence does not converge, so the invariant is now "the flag describes the durable decision" and an unconfirmed arm logs a warning instead of rewriting the answer. Also from this round: a failed disarm publish is retried once and its outcome reported. `handleSteerCancel` keeps `removed: true` — the steer really did leave the queue, and saying otherwise would make the client re-show a chip for a steer that can never arrive — and adds `disarmed: false` so the residual risk is visible rather than swallowed. Damage stays bounded regardless: the empty-boundary self-clear disarms the generation after a single seal. Tests: +1 pinning the response/durable-flag invariant. 159 packages/api specs green. * 🧹 fix: Codex round 8 — remove the unverifiable disarm signal Round 8 found the same over-promise on the disarm side that round 7 corrected on the arm side, so this applies the same answer rather than patching around it. The `disarmed: false` field added in round 7 was both unreliable and unused: a resolved publish is not proof the owner heard it (the delivery count includes this replica's own facade subscription), and it was never threaded into `CancelSteerResponse` or read by any client. A signal that claims a certainty the transport cannot provide is worse than no signal — it invites callers to trust it. Removed from the response. The retry stays, because it genuinely reduces the failure rate, and `noteSteersRemoved` still returns whether the publish succeeded FOR LOGGING, now documented explicitly as "published without error", not "the owner disarmed". Disarm is best effort with a bounded, self-healing failure: if the clear is lost the owner seals once, the empty-boundary self-clear disarms the generation, and the turn is persisted `unfinished: true` rather than silently truncated. Tightening that further needs a correlated request/response over pub-sub with a timeout — noted on the PR as the deliberate boundary of this design rather than an oversight. 130 packages/api steering specs green. * 🧽 fix: Codex round 9 — spend snapshot arms on nonempty drains too The round-6 scoping fix only cleared the pre-drain snapshot when the drain came back EMPTY. On a nonempty drain the `finally` cleared just the drained ids, so a stale arm — typically a cancel whose cross-replica clear was lost — survived the boundary. It would then immediately seal the continuation meant to answer the steer that had just been injected, and land on an empty boundary as `preempt_incomplete`: the interrupt appears to work, and the answer to it is truncated. A boundary that runs has spent its seal, so everything armed at snapshot time is spent whether or not it came back from the drain. The `finally` now clears the union of the snapshot and the drained ids. Arms that land AFTER the snapshot are still spared — their queue items are live and uninjected, which is the property round 6 added. Also fixes an api-workspace CI failure of mine: `resume.spec.js` stubs `GenerationJobManager` wholesale, and the round-3/4 resume work added two calls (`updateMetadata`, `rearmQueuedPreempts`) the stub did not define, so 34 specs threw. Stub extended. Tests: +2 (a nonempty drain clears a stale snapshot arm; a nonempty drain spares an arm that landed mid-drain). Counterfactually verified — the stale-arm spec fails against the unfixed drain. 132 packages/api specs, 60 resume specs green. * fix: never let a failed preempt subscription reject into the void registerPreemptSubscription is called detached at both sites, so a rejected Redis SUBSCRIBE became an unhandled rejection — process-fatal under Node's default --unhandled-rejections=throw. The comment already promised this path merely degrades steering; it now does. Swallowed and logged inside the registration rather than at each call site, so a future third caller cannot reintroduce the trap. Losing the channel costs this generation's cross-replica preempts, not the server: same-replica arming is runtime state and still works, and remote arms fall back to the next tool boundary. Verified counterfactually — the new spec surfaces SUBSCRIBE failed as an unhandled rejection against the unfixed registration. * docs: state the real blast radius of a failed preempt subscription LibreChat's own entrypoints install a global unhandledRejection handler that logs and keeps serving, so the escaping rejection this guards was never fatal to this server — only to another consumer of @librechat/api that installs no handler. The fix stands either way; the comment just should not overstate what it prevents. * test: cover the cross-replica preempt hop with two manager instances Every other preempt test runs against a single manager, so the hop that actually carries an interrupt in production had no coverage: the steer POST lands on whichever replica the balancer picks, which is usually not the one generating. Non-owner publishes, owner arms, owner's level-triggered poll flips — none of that was exercised end to end. Two GenerationJobManagerClass instances are a faithful replica pair here. runtimeState and ownedJobs are private instance fields, there is no module-level mutable state between them, and createStreamServices duplicates a dedicated subscriber connection per call, so separate OS processes would exercise the same objects over the same Redis. Both assertions verified counterfactually against real Redis: - Deleting the preemptCapable deserialization in RedisJobStore fails this with 'Expected: true, Received: undefined' — the exact P1 that shipped past every in-memory test and would have made the feature a silent no-op on every Redis deployment. - Dropping the non-owner arm publish fails it with 'Received: false'. * test: remove the fixed sleeps and vacuity from the cross-replica preempt test Codex round 11, both findings, both on the test I added last commit. P2 — the 300ms waits were load-bearing. Redis pub/sub never replays and the owner's SUBSCRIBE is detached, so on a slow CI worker the publish could land before anyone was listening and the test would fail against correct code. Now it republishes until the owner's state converges, which is safe because arms and clears are idempotent set writes keyed by steerId. Side effect: the tests got ~10x faster (85ms/57ms vs 929ms/606ms) since they finish on delivery rather than on a timer. P3 — afterEach destroyed only the transports, leaving each manager alive in its own cleanup-interval closure, still working against a dead transport. Now tracks the managers and awaits destroy(), which disposes the job store and its timer too. Matches how the rest of this file cleans up. Fixing the sleeps exposed a third problem codex did not flag: the stale-arm test could pass vacuously, because an undelivered arm and a fenced one look identical. It now brackets the stale publish between two control arms — the first proves the owner is listening before the stale one is sent, the second proves it has had its chance to arrive. Verified counterfactually against real Redis, and stable over 5 runs: - dropping the preemptCapable deserialization fails with 'Received: undefined' - dropping the non-owner arm publish times out both tests - removing the generation fence fails the stale test with ["control-before", "steer-stale", "control-after"] — which also confirms the bracketing orders as intended rather than by luck * fix: gate interrupt on the OWNER's capability alone, not the route's Codex round 12. The comment above this gate already said 'the OWNER's recorded capability, not this replica's probe' — and then the code ANDed in isSteerPreemptSupported(), which is exactly this replica's probe. The contradiction dates to the original commit; round 6 made the gate owner-scoped and wrote that comment without removing the local conjunct. The route never seals. It enqueues and publishes an arm, neither of which touches the SDK, so during a rolling deploy a steer landing on an un-upgraded replica silently lost its interrupt even though the owner could seal. When the route IS the owner the probe is redundant anyway: the flag it would consult is the one this process wrote at createJob. The real degradation path is unchanged and still tested — an owner that recorded no capability relabels to an ordinary steer. The test that pinned the local probe asserted an impossible same-replica state (capable metadata plus an incapable local SDK, when the metadata is written from that probe); it now pins the mixed-SDK direction instead, and fails with 'Expected: true, Received: false' if the probe is put back. * fix: reconcile arms at handover, and stop holding the 202 on a publish Codex round 13, two of three findings. P2 — rearmQueuedPreempts only ever ADDED. A replica that merely read the job still installs a facade runtime and subscribes, so it can accept an arm and then miss the best-effort clear that follows the drain. HITL resume promotes that facade to owner, the union keeps the orphan, and the first resumed stream seals on a steer no longer in the queue, drains nothing, and truncates the resumed answer as preempt_incomplete. acquireResumedJobOwnership only sets ownedJobs, so nothing else was clearing it. The durable queue is the sole authority at a handover: arms it does not back are now disarmed and tombstoned, so an in-flight publish cannot revive them either. Worth recording that my own independent review raised this and my verifier refuted it. Codex found it separately; two reviewers converging should have outweighed one refutation. P2 — the route awaited the arm publish before answering. The 202 reports capability, not delivery, so the await could not change the response; it only exposed the caller to Redis latency after the queue item was already durable. A client that times out and retries mints a second steer while the first stays queued, injecting the same instruction twice, whereas a lost publish merely takes the tool-boundary fallback. Detached, with both outcomes logged. All three tests verified counterfactually: union-only rearm fails the two new handover specs, and re-awaiting the publish hangs the stalled-publish spec until jest kills it. * fix: snapshot arms before reading the queue at handover Codex round 14 — a regression from my own round-13 fix, and a worse failure than the one it corrected. Round 13 read the durable queue first, then tombstoned any armed id the snapshot did not back. But approvals.resolve reopens steering before reconciliation runs, so another replica can commit a preempt steer and publish its arm while the peek is in flight. That arm is then present locally but absent from a snapshot taken before the steer existed, so a LIVE interrupt the route already acknowledged got dropped — and tombstoned, which blocks the re-arm, making it unrecoverable rather than merely late. Fixed by inverting the two reads rather than by locking or paying a second round trip. A steer is durably enqueued BEFORE its arm is published, so any id in an arms-first snapshot was already queued when it was armed, and the later peek must observe it unless it has since drained — which is exactly the orphan this reconciliation exists to drop. Arms landing after the snapshot are simply not candidates. Also re-checks runtime identity across the await, since the generation can be replaced while the queue read is in flight. New spec injects a steer + arm during the peek and verifies it survives; against the round-13 ordering it fails with Received array: []. * fix: bound the cancel disarm wait and fence enqueue to its generation Codex round 15. P2 — the cancel awaited its disarm publish unbounded. ioredis queues commands during an outage rather than rejecting, so that await could hang for the length of the outage with the steer ALREADY durably cancelled; a client that gives up then treats the cancel as failed and restores a chip for a steer that can never produce an applied event. Every successful cancel publishes, so ordinary steers were exposed too, not only preemptive ones. Now bounded at 1s, with the publish continuing behind it — its retry and logging are unchanged, it is just no longer in front of the response. This is the sibling of round 13's arm-publish finding; I fixed one path and left this one. P3 — enqueue was not fenced to the generation the capability decision was made against. The access checks, file resolution and owner re-read are all awaits, so the run can be replaced before the enqueue: the item then lands on the REPLACEMENT queue while the durable preempt flag and the arm still name the previous epoch, the arm is fenced out at the owner, and the 202 promises an interrupt that cannot happen. enqueueSteer now takes an expected generation, mirroring drain/peek, and the Redis path enforces it inside STEER_ENQUEUE_LUA so the check is atomic with the push rather than racing it. All three new specs verified counterfactually, including the Lua guard against real Redis (removing it returns 1 where -1 is required). * fix: fence the steer to its authorized generation, bound resume setup, keep preempt when Redis parks Codex round 16, all three findings. P2 — round 15 fenced the enqueue to owner.createdAt, the RE-READ job. Every guard above it (ownership, tenant, paused-state, agent ACL) ran against the job read at the top, so if the run was replaced during those awaits the fence happily accepted the steer into a generation the request was never authorized against, carrying the wrong agent's metadata. Now rejects on any mismatch between the validated job and the re-read. P2 — resume awaited its steering bookkeeping unbounded, after approvals.resolve had consumed the action and flipped the job to running, and outside the resume lifecycle's own try/finally. .catch does not fire on a promise that never settles, which is what ioredis produces during an outage, so the client times out, its retry gets a 409 for a spent action, and no cleanup runs. Bounded at 1s with the writes finishing in the background. P3 — Redis parks leftover steers inside its terminal-transition Lua, which projects item fields one by one, so preempt was silently dropped and a steer recovered from /chat/status lost its interrupting label. Added to both projections. All three verified counterfactually, two against real Redis. Worth recording that my first version of the generation-mismatch test was VACUOUS — it faked a createdAt matching no live job, so the round-15 enqueue fence rejected it for the wrong reason and the test passed with the guard removed. Rewritten to replace the run for real; it now fails with 'Expected 404, Received 202'.

* 🛑 feat: Preemptive Steer — server half (Interrupt & Steer, PR 2 of 3) Lets the steer route ask the generating replica to seal its live model stream at the next provider-safe boundary instead of waiting for a tool step. The run is never aborted, job status never changes, the partial answer is kept, and generation resumes in the same assistant message after the injected steer. Consumes the SDK seam in @librechat/agents (LibreChat-AI/agents#335, LibreChat-AI#346). Transport: IEventTransport gains a fenced emitPreempt/onPreempt pair beside abort. RedisEventTransport fans PREEMPT out on the SAME events channel and subscription (no new connection, key, or subscribe call); onPreempt returns a registration-scoped unsubscribe with the same replacement-safe state-identity guard onAbort uses. InMemory implements neither — single-process preempt lives entirely in the runtime set. Runtime state: RuntimeJobState carries the per-generation request set, createdAt-fenced and capped at STEER_QUEUE_MAX_DEPTH, plus a bounded `cleared` tombstone so a late cross-replica arm cannot resurrect a request whose steer already drained. registerPreemptSubscription mirrors the abort registration's double fence (runtime identity + generation createdAt); releaseAbortSubscription retires BOTH listeners and the armed set, so every terminal path drops preempt state for free. Public surface: requestPreempt (arm + fenced publish, never a rejection surface, never touches job status), isPreemptRequested (O(1) level-triggered poll), noteSteersRemoved (drain/cancel bookkeeping + fenced clear), clearPreemptRequests (empty-boundary disarm). One drain body, two boundaries: createSteerDrainHook (PostToolBatch) and createSteerPreemptBoundaryHook (PreemptBoundary) share drainAndBuildInjections, so the two injection sites cannot drift — the SDK's provider-safety argument rests on identical HumanMessage shapes. The shared body builds injections incrementally under a swallow-all catch (a mid-loop throw still injects what was applied — those parts are already persisted), clears preempt requests in finally, and disarms the generation when a boundary drains nothing. Request path: POST /chat/steer accepts preempt: true. The guard ladder is unchanged in order and in every status code. A preempt request is NEVER a rejection reason — without the capability the steer still enqueues and the 202 echoes preempt: false. Armed strictly after a successful enqueue; cancel disarms. The capability is read from the OWNING replica's recorded `preemptCapable` rather than the route replica's own SDK probe, so a rolling deploy cannot label a steer "interrupting" that the older owner will only inject at a tool step. Durable label: SteerQueueItem.preempt → TPendingSteer.preempt, so a parked/claimed/replayed chip keeps its wording. Run wiring: createRun registers the PreemptBoundary hook and threads RunConfig.preemption, both gated on isSteerPreemptSupported() — a separate probe from isSteeringSupported(), so the client affordance can never arm against an SDK that only injects at tool boundaries. buildSteerWiring builds both hooks from one shared closures object, so preemption survives HITL pause/resume for free. Honest finalization: an empty preempt boundary persists and emits with unfinished: true — the same contract an abort gets — re-marked explicitly because BaseClient has already saved the row as unfinished: false by that point. Not changed: no new job status, store method, Lua, SSE event type, endpoint, or authorization surface. abortJob, completeJob, transitionStatus, closeAndDrainSteers, getResumeState, emitChunk, applySteerPart and the whole abort path are untouched. Tests: 120 packages/api steering specs (preempt lifecycle, tombstone, fences, caps, terminal release, both-boundary drain parity, level-triggered poll, request/cancel arming, owner-capability degradation) plus 5 in api for buildSteerWiring gating, and 2 Redis-gated cross-replica transport specs. * 🔒 fix: Codex round 2 — evict tombstones, scope the empty-boundary disarm, honest resumes All four server findings were fresh consequences of the round-1 fixes, which is the review doing exactly what it should. - Tombstone cap refused new entries instead of evicting. Every drained or cancelled steer is tombstoned, not just preempting ones, so a generation that processed 20 steers exhausted the set and the late-arm race resurfaced silently. Now evicts oldest-first (Set iteration is insertion-ordered), with the budget named PREEMPT_TOMBSTONE_MAX rather than an inline expression. - The empty-boundary disarm I added in round 1 wiped the generation's ENTIRE armed set. A second steer can enqueue and arm between the atomic drain returning empty and the disarm running — that arm is backed by a live, uninjected queue item and must survive. The drain now snapshots the armed ids BEFORE draining (getArmedPreemptIds) and clearPreemptRequests takes an explicit id list instead of clearing everything. - HITL resume finalized with a hardcoded unfinished: false. The boundary hook is re-registered on resume via buildSteerWiring, so a resumed segment can end on an empty preempt boundary exactly like a fresh one; finalizeResumedTurn now reads getPreemptStats() and the halt reason, matching the normal request path. - Ownership moves on resume, so the job's recorded preemptCapable must describe the replica that will actually generate. Refreshed before resumeCompletion; a job created on a capable replica that resumes on an older one during a rolling deploy no longer acknowledges steers as interrupting. Tests: +3 (scoped disarm sparing a post-snapshot arm, oldest-first tombstone eviction, id-list disarm). 122 packages/api steering specs green. * 🚨 fix: Codex round 3 — deserialize preemptCapable from Redis (feature was dead under Redis) The P1 here is the most consequential defect in the whole feature, and it was introduced by round 1's own capability fix. - `RedisJobStore.serializeJob` writes booleans generically, so `preemptCapable` reached Redis — but `deserializeJob` is an EXPLICIT field map and had no line for it. Every `getJob()` therefore dropped the flag, `job.metadata.preemptCapable` was always undefined, and `handleSteerRequest` computed `preemptArmed: false` unconditionally. Interrupt & steer would have silently degraded to ordinary tool-boundary steering in EVERY Redis deployment — i.e. the feature shipping as a no-op in production while passing every in-memory test. Now deserialized, with a round-trip assertion in the metadata spec that fails (`Received: undefined`) against the unfixed store. - The resume capability refresh moved from just-before `resumeCompletion` to immediately after `approvals.resolve` claims the run. That call already flips the job back to `running`, so the steer route accepts requests from that instant; leaving the refresh 135 lines later (across the whole client reconstruction) left a real window where a steer read the PREVIOUS owner's capability. Not the fully atomic transition Codex suggested — that reaches into the approvals Lua — but it shrinks the window from seconds to one await, which is proportionate for a label-accuracy issue. Refuted: "avoid triggering preemption inside subagents". The premise — that the run-wide poll can seal a subagent stream — does not hold against the shipped SDK. Child graphs are constructed with `subagentScope: true` (SubagentExecutor) and `preemption` is NOT propagated into child inputs, while `canClaimPreemptSeal()` requires `!subagentScope && preemption != null`. Both conditions fail independently, so a subagent can never claim a seal and the boundary cannot fire with `agentId` set. The `input.agentId != null` guard in the hook is defensive depth, not the thing standing between us and the described failure. 140 packages/api specs green. * 🔁 fix: Codex round 4 — re-arm durable interrupt steers when resume moves owners - An arm lives only in the owning replica's runtime plus a transient pub/sub message, while the steer's `preempt` flag is durable on the queue item. A HITL resume landing on a different replica therefore started with an empty armed set and a poll stuck false, so an interrupt the user had already been ACKed for silently waited for an ordinary tool boundary. New `GenerationJobManager.rearmQueuedPreempts` rebuilds the armed set by peeking the durable queue (fenced on the generation) and re-arming every item flagged `preempt`; resume calls it right after claiming. Safe by construction: every item peeked is still queued, so no drained steer can be resurrected. - Capability-refresh failure now logs at error rather than warn, but deliberately does NOT fail the resume — see the reply on that thread. Tests: +2 (rebuild from queue arms only the flagged item and reports the count; a stale generation arms nothing). 124 packages/api steering specs green. * 📡 fix: Codex round 5 — acknowledge only what was actually armed - A cross-replica arm was fire-and-forget: `emitPreempt` logged its own publish failure and `requestPreempt` returned void, so the route answered `preempt: true` even when the owner never armed a poll. The steer still injected at the next tool boundary, but the chip claimed an interrupt that could not happen — and unlike HITL resume, an ordinary running generation had no durable reconciliation to recover it. `emitPreempt` now resolves to the subscriber count and rejects on failure; `requestPreempt` is async and returns whether the arm truly landed (owned locally, or delivered to at least one subscriber). The 202 reports THAT rather than what was asked for, so the chip relabels to ordinary steering exactly as it does for a capability-degraded deployment. Errors are swallowed into `false` — an unarmed interrupt is a downgrade, never a failed steer. - The owner capability is re-read immediately before enqueue rather than reused from the top of the guard ladder. `checkAgentAccess` and file resolution are awaits, so a request can span an entire HITL pause/resume that moves ownership to a replica with different capability and rewrites that very flag. Only paid for by requests that actually asked to interrupt. Tests: +3 (not-armed when the publish reaches nobody; armed when this replica owns the generation; a throwing publish downgrades instead of propagating). 127 packages/api steering specs green. * 🎯 fix: Codex round 6 — real ownership, confirmed disarms, and a CI regression of my own Three review findings plus three CI failures the round-5 commit caused. Review: - Ownership came from `runtimeState`, which a cross-replica `getJob` populates with a FACADE runtime on any replica that merely read the job. Matching `createdAt` therefore proved only "we looked at this job", so a non-owner could arm nothing and report success. Ownership now comes from `ownedJobs`, the actual owner map. - `armPreemptIds` returns how many ids it accepted, and a local arm is only reported as armed when one was. A tombstoned id (its steer drained at an ordinary boundary mid-request) no longer answers `preempt: true` for an interrupt that cannot happen. - The cancel disarm is awaited. A dropped clear is worse than a dropped arm: the owner keeps a level-triggered request for a steer that no longer exists, seals its next chunk and truncates an unrelated answer. The boundary drain's own call stays non-blocking — there the owner is local, so the disarm is already effective and awaiting the informational publish would only delay injection. - Subscriber count is NOT read as proof of owner receipt: the count includes this replica's own facade subscription. A successful publish reports armed, a rejected one does not. Documented rather than papered over — see the acknowledgement-semantics note on the PR. CI regressions from round 5, all mine: - `registerPreemptSubscription` was AWAITED at both runtime-init sites, so job creation blocked on a second Redis channel subscription and hung when that subscribe was slow. Abort is awaited because a missed abort strands a run; a missed preempt only degrades that steer to the next tool boundary, so it now registers without gating createJob. - Two api specs mocked `@librechat/api` without the newly imported `isSteerPreemptSupported`, so the call threw before createJob; and one exact-match assertion needed the new `preemptCapable` metadata field. - My own Redis integration spec asserted arm-before-clear ordering, which two publishes carry no guarantee of — the receiving tombstone exists precisely because of that. Now asserts delivery and payload fidelity, order-independent. 158 packages/api specs, 27 api specs green. * 🧭 fix: Codex round 7 — settle the acknowledgement semantics (Option A) Round 7's second finding is the incoherence I flagged on the PR: the route persisted `preempt: true` on the durable queue item while returning `preempt: false` when delivery could not be confirmed. Those two then disagreed, and `rearmQueuedPreempts` trusts the DURABLE one — so a resumed owner would honour an interrupt the client had explicitly been told degraded to ordinary steering. Rather than patch the disagreement, this settles the meaning: `preempt` in the 202 means "queued as an interrupt request", NOT "a seal is guaranteed". It mirrors `SteerQueueItem.preempt` exactly, so the response, the durable record, and the resume-time re-arm can never disagree. The gates that ARE knowable stay — the owner's recorded capability and a successful enqueue. Everything past that degrades to the documented fallback of injecting at the next tool boundary. A route cannot synchronously know whether another replica will seal: proving it needs a correlated request/response over pub-sub, and even that only proves the owner heard, not that it is still streaming when the arm lands. Four rounds of tightening this boolean each surfaced a narrower case; the sequence does not converge, so the invariant is now "the flag describes the durable decision" and an unconfirmed arm logs a warning instead of rewriting the answer. Also from this round: a failed disarm publish is retried once and its outcome reported. `handleSteerCancel` keeps `removed: true` — the steer really did leave the queue, and saying otherwise would make the client re-show a chip for a steer that can never arrive — and adds `disarmed: false` so the residual risk is visible rather than swallowed. Damage stays bounded regardless: the empty-boundary self-clear disarms the generation after a single seal. Tests: +1 pinning the response/durable-flag invariant. 159 packages/api specs green. * 🧹 fix: Codex round 8 — remove the unverifiable disarm signal Round 8 found the same over-promise on the disarm side that round 7 corrected on the arm side, so this applies the same answer rather than patching around it. The `disarmed: false` field added in round 7 was both unreliable and unused: a resolved publish is not proof the owner heard it (the delivery count includes this replica's own facade subscription), and it was never threaded into `CancelSteerResponse` or read by any client. A signal that claims a certainty the transport cannot provide is worse than no signal — it invites callers to trust it. Removed from the response. The retry stays, because it genuinely reduces the failure rate, and `noteSteersRemoved` still returns whether the publish succeeded FOR LOGGING, now documented explicitly as "published without error", not "the owner disarmed". Disarm is best effort with a bounded, self-healing failure: if the clear is lost the owner seals once, the empty-boundary self-clear disarms the generation, and the turn is persisted `unfinished: true` rather than silently truncated. Tightening that further needs a correlated request/response over pub-sub with a timeout — noted on the PR as the deliberate boundary of this design rather than an oversight. 130 packages/api steering specs green. * 🧽 fix: Codex round 9 — spend snapshot arms on nonempty drains too The round-6 scoping fix only cleared the pre-drain snapshot when the drain came back EMPTY. On a nonempty drain the `finally` cleared just the drained ids, so a stale arm — typically a cancel whose cross-replica clear was lost — survived the boundary. It would then immediately seal the continuation meant to answer the steer that had just been injected, and land on an empty boundary as `preempt_incomplete`: the interrupt appears to work, and the answer to it is truncated. A boundary that runs has spent its seal, so everything armed at snapshot time is spent whether or not it came back from the drain. The `finally` now clears the union of the snapshot and the drained ids. Arms that land AFTER the snapshot are still spared — their queue items are live and uninjected, which is the property round 6 added. Also fixes an api-workspace CI failure of mine: `resume.spec.js` stubs `GenerationJobManager` wholesale, and the round-3/4 resume work added two calls (`updateMetadata`, `rearmQueuedPreempts`) the stub did not define, so 34 specs threw. Stub extended. Tests: +2 (a nonempty drain clears a stale snapshot arm; a nonempty drain spares an arm that landed mid-drain). Counterfactually verified — the stale-arm spec fails against the unfixed drain. 132 packages/api specs, 60 resume specs green. * fix: never let a failed preempt subscription reject into the void registerPreemptSubscription is called detached at both sites, so a rejected Redis SUBSCRIBE became an unhandled rejection — process-fatal under Node's default --unhandled-rejections=throw. The comment already promised this path merely degrades steering; it now does. Swallowed and logged inside the registration rather than at each call site, so a future third caller cannot reintroduce the trap. Losing the channel costs this generation's cross-replica preempts, not the server: same-replica arming is runtime state and still works, and remote arms fall back to the next tool boundary. Verified counterfactually — the new spec surfaces SUBSCRIBE failed as an unhandled rejection against the unfixed registration. * docs: state the real blast radius of a failed preempt subscription LibreChat's own entrypoints install a global unhandledRejection handler that logs and keeps serving, so the escaping rejection this guards was never fatal to this server — only to another consumer of @librechat/api that installs no handler. The fix stands either way; the comment just should not overstate what it prevents. * test: cover the cross-replica preempt hop with two manager instances Every other preempt test runs against a single manager, so the hop that actually carries an interrupt in production had no coverage: the steer POST lands on whichever replica the balancer picks, which is usually not the one generating. Non-owner publishes, owner arms, owner's level-triggered poll flips — none of that was exercised end to end. Two GenerationJobManagerClass instances are a faithful replica pair here. runtimeState and ownedJobs are private instance fields, there is no module-level mutable state between them, and createStreamServices duplicates a dedicated subscriber connection per call, so separate OS processes would exercise the same objects over the same Redis. Both assertions verified counterfactually against real Redis: - Deleting the preemptCapable deserialization in RedisJobStore fails this with 'Expected: true, Received: undefined' — the exact P1 that shipped past every in-memory test and would have made the feature a silent no-op on every Redis deployment. - Dropping the non-owner arm publish fails it with 'Received: false'. * test: remove the fixed sleeps and vacuity from the cross-replica preempt test Codex round 11, both findings, both on the test I added last commit. P2 — the 300ms waits were load-bearing. Redis pub/sub never replays and the owner's SUBSCRIBE is detached, so on a slow CI worker the publish could land before anyone was listening and the test would fail against correct code. Now it republishes until the owner's state converges, which is safe because arms and clears are idempotent set writes keyed by steerId. Side effect: the tests got ~10x faster (85ms/57ms vs 929ms/606ms) since they finish on delivery rather than on a timer. P3 — afterEach destroyed only the transports, leaving each manager alive in its own cleanup-interval closure, still working against a dead transport. Now tracks the managers and awaits destroy(), which disposes the job store and its timer too. Matches how the rest of this file cleans up. Fixing the sleeps exposed a third problem codex did not flag: the stale-arm test could pass vacuously, because an undelivered arm and a fenced one look identical. It now brackets the stale publish between two control arms — the first proves the owner is listening before the stale one is sent, the second proves it has had its chance to arrive. Verified counterfactually against real Redis, and stable over 5 runs: - dropping the preemptCapable deserialization fails with 'Received: undefined' - dropping the non-owner arm publish times out both tests - removing the generation fence fails the stale test with ["control-before", "steer-stale", "control-after"] — which also confirms the bracketing orders as intended rather than by luck * fix: gate interrupt on the OWNER's capability alone, not the route's Codex round 12. The comment above this gate already said 'the OWNER's recorded capability, not this replica's probe' — and then the code ANDed in isSteerPreemptSupported(), which is exactly this replica's probe. The contradiction dates to the original commit; round 6 made the gate owner-scoped and wrote that comment without removing the local conjunct. The route never seals. It enqueues and publishes an arm, neither of which touches the SDK, so during a rolling deploy a steer landing on an un-upgraded replica silently lost its interrupt even though the owner could seal. When the route IS the owner the probe is redundant anyway: the flag it would consult is the one this process wrote at createJob. The real degradation path is unchanged and still tested — an owner that recorded no capability relabels to an ordinary steer. The test that pinned the local probe asserted an impossible same-replica state (capable metadata plus an incapable local SDK, when the metadata is written from that probe); it now pins the mixed-SDK direction instead, and fails with 'Expected: true, Received: false' if the probe is put back. * fix: reconcile arms at handover, and stop holding the 202 on a publish Codex round 13, two of three findings. P2 — rearmQueuedPreempts only ever ADDED. A replica that merely read the job still installs a facade runtime and subscribes, so it can accept an arm and then miss the best-effort clear that follows the drain. HITL resume promotes that facade to owner, the union keeps the orphan, and the first resumed stream seals on a steer no longer in the queue, drains nothing, and truncates the resumed answer as preempt_incomplete. acquireResumedJobOwnership only sets ownedJobs, so nothing else was clearing it. The durable queue is the sole authority at a handover: arms it does not back are now disarmed and tombstoned, so an in-flight publish cannot revive them either. Worth recording that my own independent review raised this and my verifier refuted it. Codex found it separately; two reviewers converging should have outweighed one refutation. P2 — the route awaited the arm publish before answering. The 202 reports capability, not delivery, so the await could not change the response; it only exposed the caller to Redis latency after the queue item was already durable. A client that times out and retries mints a second steer while the first stays queued, injecting the same instruction twice, whereas a lost publish merely takes the tool-boundary fallback. Detached, with both outcomes logged. All three tests verified counterfactually: union-only rearm fails the two new handover specs, and re-awaiting the publish hangs the stalled-publish spec until jest kills it. * fix: snapshot arms before reading the queue at handover Codex round 14 — a regression from my own round-13 fix, and a worse failure than the one it corrected. Round 13 read the durable queue first, then tombstoned any armed id the snapshot did not back. But approvals.resolve reopens steering before reconciliation runs, so another replica can commit a preempt steer and publish its arm while the peek is in flight. That arm is then present locally but absent from a snapshot taken before the steer existed, so a LIVE interrupt the route already acknowledged got dropped — and tombstoned, which blocks the re-arm, making it unrecoverable rather than merely late. Fixed by inverting the two reads rather than by locking or paying a second round trip. A steer is durably enqueued BEFORE its arm is published, so any id in an arms-first snapshot was already queued when it was armed, and the later peek must observe it unless it has since drained — which is exactly the orphan this reconciliation exists to drop. Arms landing after the snapshot are simply not candidates. Also re-checks runtime identity across the await, since the generation can be replaced while the queue read is in flight. New spec injects a steer + arm during the peek and verifies it survives; against the round-13 ordering it fails with Received array: []. * fix: bound the cancel disarm wait and fence enqueue to its generation Codex round 15. P2 — the cancel awaited its disarm publish unbounded. ioredis queues commands during an outage rather than rejecting, so that await could hang for the length of the outage with the steer ALREADY durably cancelled; a client that gives up then treats the cancel as failed and restores a chip for a steer that can never produce an applied event. Every successful cancel publishes, so ordinary steers were exposed too, not only preemptive ones. Now bounded at 1s, with the publish continuing behind it — its retry and logging are unchanged, it is just no longer in front of the response. This is the sibling of round 13's arm-publish finding; I fixed one path and left this one. P3 — enqueue was not fenced to the generation the capability decision was made against. The access checks, file resolution and owner re-read are all awaits, so the run can be replaced before the enqueue: the item then lands on the REPLACEMENT queue while the durable preempt flag and the arm still name the previous epoch, the arm is fenced out at the owner, and the 202 promises an interrupt that cannot happen. enqueueSteer now takes an expected generation, mirroring drain/peek, and the Redis path enforces it inside STEER_ENQUEUE_LUA so the check is atomic with the push rather than racing it. All three new specs verified counterfactually, including the Lua guard against real Redis (removing it returns 1 where -1 is required). * ⚡ feat: Interrupt & Steer — client half (PR 3 of 3) Makes preemptive steering reachable. Consumes the server contract from PR 2 (POST /chat/steer `preempt`, echoed on the 202) and the SDK seam in @librechat/agents 3.3.5. Settings shape follows the agreed correction, NOT the earlier plan draft: `steerInterruptsByDefault` is a boolean ORTHOGONAL to `duringRunDefaultAction` — that enum still chooses steer-vs-queue, the new boolean chooses how soon a steer lands. This deliberately avoids widening the enum to three values, which would have silently broken two hard-coded binary TOGGLES (`DuringRunAction.tsx`'s setter and `SteerMenu`'s `useDefaultToggleEntry`, both `prev === 'steer' ? … : …`) where a third value collapses to the wrong branch and one click erases the setting. - useSteering: `submitSteer` takes an opts bag and threads `preempt` into the POST, the optimistic chip, and the failure chip. The ACK relabels from the SERVER's echo, so a deployment that cannot seal mid-stream downgrades the chip's wording instead of erroring — the entire UX surface of capability degradation. New `interruptSteer` reuses the whole chip lifecycle and degradation ladder, and falls back to `interruptAndSend` when `!canSteer`, because steering needs a server-side job and an always-visible button would otherwise be dead for the whole first turn. `steerFromComposer` honours the new preference. - Composer: always-visible `InterruptSteerButton` with one fixed meaning (stop now, keep what's written), disabled on a paused run to pre-empt the server's 409, `type="button"` so it never steals the form's Enter submit, RTL-correct margins. A fourth hovercard row on the during-run send button, and ⌘/Ctrl+Shift+Enter routed AHEAD of the bare ⌘/Ctrl+Enter branch that would otherwise swallow it. - Chips: an in-flight preempt chip reads "Interrupting" with a ZapOff glyph; `preempt` survives reconnect through `seedSteerChips`. - `RunEnd.interruptArmed`, `drainAfterAbortByIndex`, `useQueueDrain`, `stopGenerating` and `interruptAndSend` are untouched — the preempt path deliberately shares none of the abort machinery. Tests: 7 new specs (posts preempt, turn-1 fallback, empty-text refusal, default route with and without the preference, server-echo relabel, double-click). 66 useSteering specs green; tsc and lint clean. Round-1 review fixes folded in: - P1: interrupt & steer no longer hard-aborts a run paused on tool approval. `canSteer` is false there, so the fallback was routing the keyboard and hovercard paths into `interruptAndSend` — discarding the partial answer, the exact opposite of what the action promises. The fallback is now scoped to the missing-conversation case only, and a paused run refuses outright (the standalone button was already disabled; the guard now lives where all three paths reach it). - The preference no longer leaks into the explicit Steer action. `steerFromComposer` backs both the default Enter route AND the explicit hovercard row / Ctrl+Enter alternate; applying `steerInterruptsByDefault` inside it made ordinary Steer interrupt and the two rows indistinguishable. It now takes an explicit argument that only `submitDuringRun`'s default route sets. - Retry preserves preemption: a failed interrupt-steer chip keeps `preempt: true`, and `retrySteer` now forwards it rather than silently resubmitting as an ordinary tool-boundary steer. - ⌘/Ctrl+Shift+Enter defers to a rebound submit shortcut, mirroring the bare ⌘/Ctrl+Enter branch — a user who bound submit to that chord keeps getting submit. Round-2 fix: the preempt label now survives the page-reload resume path too. `seedSteerChips` (useResumableSSE) and `restoreSteerChips` (useResumeOnLoad) are two independent TPendingSteer→PendingSteer mappers with near-identical bodies; the first carried the flag and the second silently dropped it, so an armed interrupt reverted to plain "Steering" after a reload. Swept: those are the only two in production code. The reclaim/convert paths deliberately omit it — a queued follow-up starts its own turn, so there is nothing to interrupt. * fix: yield the interrupt-steer chord only to a submit shortcut bound to it The previous guard skipped the Ctrl/Cmd+Shift+Enter branch whenever ANY submitMessage override existed. Rebinding submit to something unrelated (Ctrl+J) or unbinding it entirely then fell through to the override resolver, which returns 'none' for shifted Enter — silently removing the shortcut the hovercard still advertises. Compare the pressed chord against the configured one instead. The adjacent bare Ctrl/Cmd+Enter branch keeps its any-override guard on purpose: that chord IS the default submit chord, so once submit moves the resolver should own it. The predicate already existed inside resolveSubmitOverrideAction; pulled it out as bindingsMatch so both sites compare chords the same way. That call is behavior-preserving — eventBinding.key is 'Enter' by the early return, and equal hashes imply equal keys, so the dropped explicit key check was redundant. * fix: disable the Interrupt & steer menu row while paused on approval interruptSteer hard-refuses when the run is paused for tool approval, but the hovercard row was never gated, so it rendered enabled and clicking it did nothing at all — no chip, no queue entry, no toast — at exactly the moment a user is trying to say "stop, don't run that command". The standalone button already gates on pausedOnApproval; the row contradicted it. Gated on pausedOnApproval rather than !canSteer like the steer row above, because canSteer is also false before a conversation exists, where interruptSteer deliberately falls back to interruptAndSend and the row must stay live for the whole first turn. Tests pin both directions and were verified counterfactually: removing the gate fails the paused case, and using !canSteer fails the first-turn case. * test: render the during-run hovercard eagerly instead of driving Ariakit The new spec passed locally and failed all four cases on CI's Ubuntu and Windows shards: Ariakit's show path keys off pointer geometry, which jsdom reports as zeros, so whether a synthetic mouseEnter opens the hovercard is environment-dependent. Driving it was testing Ariakit's hover behavior, not which rows this component disables. Mocking the three Ariakit primitives renders the rows unconditionally and drops the fake timers. Both counterfactuals still fail as they should: removing the gate fails the paused case, !canSteer fails the first-turn case. * test(e2e): cover interrupt & steer sealing mid-stream The mock Playwright suite covered every sibling during-run action — steer at a tool boundary, steer degrading to a queued follow-up, queue, and interrupt & send — but not interrupt & steer, the one this stack adds. Uses E2E_SLOW_REPLY, which streams pure text with no tools, so the scenario is the same one where an ordinary steer provably degrades to a queued follow-up turn. Injecting in-thread there is something only a mid-stream seal can do, which makes the assertion discriminating rather than incidental: the steer part lands in the response, the final chunk never arrives, the text written before the seal survives, and no follow-up turn pair is created. * test(e2e): assert the run resumes after the seal, not just that it sealed The other four assertions are all satisfied by a seal that killed the run: the steer part is persisted by applySteer during the drain, before the continuation starts, so 'sealed and resumed' and 'sealed and died' were indistinguishable — and resuming is the whole difference from interrupt & send. The continuation answers the injected steer, whose text carries no fake-model marker, so getLatestUserText falls through to the default reply. That string ('E2E mock reply') is distinct from the setup turn's ('E2E reply <label>'), so seeing it proves generation restarted rather than matching text that was already on screen. The test itself is confirmed working: it ran as 104/121 in the Playwright job on 0ece357 and the run concluded success. * test(e2e): drop the resume assertion pending an unresolved question The assertion that the run visibly resumes after the seal fails deterministically in CI across all three retries. Every other assertion in the test passes, so the chord, the server preempt, the seal, and the in-thread injection all work; what fails is only the continuation's reply becoming visible. I could not determine from CI logs whether that is the mock harness not surfacing a continuation in a no-tool scenario or generation genuinely stopping after the seal, and I am not willing to weaken it into something that passes either way — that would convert a real question into false assurance. Reverted to the four assertions that hold, with the open question recorded in the test and raised on the PR. * fix: yield the interrupt chord to any bound shortcut, and move the Enter hint Codex round 1 on this PR (its first — LibreChat-AI#14519's rounds predate these files). P2 — the chord yielded only to a rebound submitMessage. This composer handler runs before the document-level one in useKeyboardShortcuts, and that one does not check defaultPrevented, so binding any composer-allowed action (focusChat, focusSearch, showShortcuts) to Ctrl/Cmd+Shift+Enter fired BOTH: the run was interrupted and the bound action ran. Now yields to any chord the user has bound. No default binding uses this chord, so it only ever yields to a deliberate rebinding. P2 — with steerInterruptsByDefault on, plain Enter routes through submitDuringRun and preempts, but the hovercard still put the ⏎ hint on the ordinary Steer row, whose click deliberately does NOT preempt. The same row advertised a key that did something else. The hint now follows the preference: ⏎ moves to Interrupt & steer, and the Steer row shows none because no key reaches it in that mode. I had declined this twice on the grounds that the behaviour split is deliberate — it is, and it is unchanged; the finding was about the label, which was a different claim and a correct one. Lint also caught a real bug in the first fix: boundShortcutChords was missing from the handler's dependency array, so a rebinding would not have taken effect until the callback was recreated for another reason. Both verified counterfactually; 179 client specs green. * fix: resolve every composer Enter chord through one decision table Codex round 2: yielding the preempt branch to a bound chord dropped execution into the bare Ctrl/Cmd branch below it, which ignores Shift, and past that into the submit tail, where isCtrlEnter is true for the chord. So rebinding focusChat, focusSearch, or showShortcuts to Ctrl/Cmd+Shift+Enter fired the alternate action or a submit AND the document-level shortcut, consuming the draft. Two rounds in a row landed in this handler because the guard chain decided "whose chord is this?" piecemeal inside individual branches, with fall-through between them. This extracts the entire pipeline into resolveComposerKeyDown (utils/shortcuts, beside resolveSubmitOverrideAction, which it absorbs as a step): one pure decision table where every verdict is terminal, so the fall-through class is gone rather than patched around. The yield rule is now the first gate before all branches: an Enter chord bound to any shortcut the document handler runs while typing (EDITING_ALLOWED_SHORTCUTS, hoisted out of the handler and shared by both dispatchers) is left entirely to that handler. This also closes the identical latent holes in the branches codex did not flag: bound Alt+Enter and Ctrl/Cmd+Enter chords double-fired the same way during a run, and the idle submit tail consumed bound chords too. A chord bound to a shortcut the document handler does NOT run while typing keeps its composer meaning; there is nothing to collide with, and yielding would just make the chord dead. Also merged the spec file round 0 added at utils/__tests__/ into the pre-existing utils/shortcuts.spec.ts (duplicate coverage at a second path) and dropped the isNonShiftEnter+filesLoading preventDefault, which was subsumed by the unconditional one below it. The table has a spec locking every verdict. Counterfactually verified: removing the yield gate fails exactly the three yield tests. Full client suite green apart from six suites that fail identically without this change (local data-provider dist drift in unrelated areas). * fix: yield Alt+Enter to a rebound submit, derive hovercard hints from the decision table Codex round 3, both findings. P2, Alt+Enter submit rebinding. The interrupt branch reserved Alt+Enter unconditionally, so rebinding submitMessage to Alt+Enter meant the user's own submit chord aborted the run and consumed the draft as interrupt & send. Now guarded with the same bindingsMatch yield the preempt branch already had; the chord falls through to the submit override resolution and submits the default action. P2, hovercard hints. The rows advertised hardcoded chords, so a chord rebound to an editing-allowed global shortcut (yielded by the decision table) or claimed by a rebound submit still appeared as Steer, Queue, Interrupt & steer, or Interrupt & send. The hovercard now asks the same decision table the composer executes what each canonical chord does and only labels a row with a chord that still triggers it. That also fixes two dishonest hints codex did not flag: with Enter-to-send off, plain Enter inserts a newline during a run (the primary row advertised it anyway) and Ctrl/Cmd+Enter submits the default action (the alternate row claimed it). The default-action hint now moves to Ctrl/Cmd+Enter in that mode. The effective bindings (submitOverride plus yielded chords) moved from useTextarea into a shared useComposerBindings hook so the handler and the hints read the same source. resolveComposerKeyDown now takes a KeyChordSource pick of the event fields it reads, letting the hovercard pass synthetic chords. Counterfactually verified: reverting the Alt guard fails the new resolver test and the hint suppression test. 353 tests across the affected suites green; the two failing hook suites (useVisibleTools, useResumableSSE) fail identically without these changes (local data-provider dist drift). * fix: never advertise a chord on a disabled hovercard row Codex round 4, one P2. The Interrupt & steer row kept showing its chord while disabled for tool approval, but pressing it reaches interruptSteer's pausedOnApproval guard and no-ops. Fixed as a render rule rather than a per-row patch: a disabled row never shows its kbd, since its action refuses the chord by the same guard that disabled it. That also covers the disabled Steer row, whose alternate-action hint had the identical hole through steerFromComposer's canSteer refusal. Counterfactually verified: reverting the render guard fails the new test.

ClaraLeigh added 3 commits May 21, 2023 21:15

feature: Init a app config file

e178b61

- Simplifies the ENV vars too - Legacy fallbacks for older builds

refactor(auth): Refactor log in/out controllers

911b8e2

- Moves both login and logout controllers to their own file

chore(jwt): Throw warning if secret is default

44ec9c7

danorlando reviewed May 22, 2023

View reviewed changes

Comment thread api/server/controllers/auth.controller.js

ClaraLeigh added 5 commits May 24, 2023 17:55

feature(frontend): Ability to disable registration

bb0fd45

feature(env): Env in the root + version support

567cb04

ie .env.prod, .env.dev, .env.test

feature: Upgrade .env script for users

d2b6060

chore(config): Refactor and remove legacy env refs

dce8067

feature(upgrade): Upgrade script for .env changes

e97a97a

This was referenced May 24, 2023

How to close the registration function？ #341

Closed

New user system breaks Apache Basic Auth #273

Closed

danorlando reviewed May 24, 2023

View reviewed changes

Comment thread api/server/services/auth.service.js

danorlando reviewed May 24, 2023

View reviewed changes

Comment thread api/server/services/auth.service.js

ClaraLeigh mentioned this pull request May 30, 2023

User System: Cleanup and Bugs (2) #408

Closed

6 of 7 tasks

ClaraLeigh closed this May 30, 2023

ClaraLeigh deleted the chore/user-cleanup branch May 30, 2023 20:23

This was referenced Jun 3, 2023

Feature: User/Reg cleanup + Install / Upgrade script #421

Closed

Feature: User/Reg cleanup + Install / Upgrade script for langchain #427

Merged

This was referenced Jul 29, 2026

🛑 feat: Preemptive Steer - Backend Interrupt & Steer #14518

Merged

⚡ feat: Interrupt & Steer, Client #14519

Merged

danny-avila mentioned this pull request Jul 30, 2026

⚡ feat: Interrupt & Steer, Client #14528

Merged

Navigation Menu

Repository navigation

User System: Cleanup and Bugs - #346

User System: Cleanup and Bugs#346
ClaraLeigh wants to merge 8 commits into
LibreChat-AI:mainfrom
ClaraLeigh:chore/user-cleanup

ClaraLeigh commented May 21, 2023 •

edited

Loading

Uh oh!

danorlando commented May 22, 2023

Uh oh!

Uh oh!

danorlando May 24, 2023

Uh oh!

ClaraLeigh May 25, 2023 •

edited

Loading

Uh oh!

danorlando May 25, 2023

Uh oh!

ClaraLeigh May 25, 2023

Uh oh!

ClaraLeigh May 25, 2023

Uh oh!

danorlando May 26, 2023 •

edited

Loading

Uh oh!

danorlando May 26, 2023

Uh oh!

ClaraLeigh May 26, 2023

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

ClaraLeigh commented May 21, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

danorlando commented May 22, 2023

Uh oh!

Uh oh!

danorlando May 24, 2023

Choose a reason for hiding this comment

Uh oh!

ClaraLeigh May 25, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

danorlando May 25, 2023

Choose a reason for hiding this comment

Uh oh!

ClaraLeigh May 25, 2023

Choose a reason for hiding this comment

Uh oh!

ClaraLeigh May 25, 2023

Choose a reason for hiding this comment

Uh oh!

danorlando May 26, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

danorlando May 26, 2023

Choose a reason for hiding this comment

Uh oh!

ClaraLeigh May 26, 2023

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

ClaraLeigh commented May 21, 2023 •

edited

Loading

ClaraLeigh May 25, 2023 •

edited

Loading

danorlando May 26, 2023 •

edited

Loading