fix(canvas-chat): clear stale 'Failed to send' banner when the agent replies (core#2697) #2736

Merged

devops-engineer merged 1 commits from fix/chat-clear-stale-error-on-reply into main 2026-06-13 09:07:42 +00:00

Conversation 3 Commits 1 Files Changed 2

+90

core-devops commented 2026-06-13 09:01:33 +00:00

Member

Copy Link

Copy Source

Bug (CTO screenshot, JRS)

The red "Failed to send — agent may be unreachable" banner stayed up while the agent was visibly working (thinking indicator at ●●● 37s, live tool feed streaming). Contradictory UI.

Root cause

The agent's turn hit a real context-overflow 400 (token limit 262144, requested 262158) which set the banner; the runtime auto-healed (reset session + retried). The retry's reply then arrived — but onAgentMessage / onSendComplete released the send guards and never cleared the error state, so the stale "unreachable" banner persisted over a reachable, replying agent.

Fix

A reply landing (push AGENT_MESSAGE) or a poll-mode send-complete proves reachability → both callbacks now clear both error sources (setError(null) + clearSendError()).

Tests

New ChatTab.errorClearOnReply.test.tsx captures ChatTab's socket callbacks and asserts onAgentMessage + onSendComplete clear the error. Full ChatTab suite green (358 pass).

Note (separate, not a regression)

The underlying trigger — the SEO agent's session exceeding the model's 256K context — is handled by the runtime's existing auto-heal (session reset + retry), and the user-facing remedy is the New session button (#2700, shipped). This PR only fixes the misleading banner.

🤖 Generated with Claude Code

## Bug (CTO screenshot, JRS) The red **"Failed to send — agent may be unreachable"** banner stayed up **while the agent was visibly working** (thinking indicator at `●●● 37s`, live tool feed streaming). Contradictory UI. ### Root cause The agent's turn hit a real **context-overflow 400** (`token limit 262144, requested 262158`) which set the banner; the runtime **auto-healed** (reset session + retried). The retry's reply then arrived — but `onAgentMessage` / `onSendComplete` released the send guards and **never cleared the error state**, so the stale "unreachable" banner persisted over a reachable, replying agent. ### Fix A reply landing (push `AGENT_MESSAGE`) or a poll-mode send-complete proves reachability → both callbacks now clear both error sources (`setError(null)` + `clearSendError()`). ### Tests New `ChatTab.errorClearOnReply.test.tsx` captures ChatTab's socket callbacks and asserts `onAgentMessage` + `onSendComplete` clear the error. Full ChatTab suite green (358 pass). ### Note (separate, not a regression) The underlying trigger — the SEO agent's session exceeding the model's 256K context — is handled by the runtime's existing auto-heal (session reset + retry), and the user-facing remedy is the **New session** button (#2700, shipped). This PR only fixes the misleading banner. 🤖 Generated with [Claude Code](https://claude.com/claude-code)

core-devops added 1 commit 2026-06-13 09:01:34 +00:00

fix(canvas-chat): clear stale "Failed to send" banner when the agent replies (core#2697) ... 59792d9b93

Reported on JRS: a turn that hit a context-overflow 400 set the
"Failed to send — agent may be unreachable" banner; the runtime auto-healed
(reset session + retried); the retry's reply landed and tools streamed at
"●●● Ns" — yet the red banner stayed up, contradicting the visibly-working
agent. onAgentMessage / onSendComplete released the send guards but never
cleared the error state.

A reply landing (push AGENT_MESSAGE) or a poll-mode send-complete PROVES the
agent is reachable, so both callbacks now clear both error sources
(setError(null) + clearSendError()). Adds an integration test capturing
ChatTab's socket callbacks and asserting the clear.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

agent-reviewer-cr2 approved these changes 2026-06-13 09:07:13 +00:00

agent-reviewer-cr2 left a comment

Member

Copy Link

Copy Source

APPROVED: reviewed #2736 at head 59792d9b.

The fix is correctly tied to proof of agent reachability: both onAgentMessage and poll-mode onSendComplete clear the local error plus the useChatSend error, so a stale 'Failed to send / agent may be unreachable' banner is removed once the agent actually replies or completes. This does not change send failure detection itself, and it does not mask a still-failing send path before a reachability callback arrives.

The new ChatTab.errorClearOnReply test captures the real socket callbacks and covers both AGENT_MESSAGE and onSendComplete. Required CI is green, including Canvas (Next.js) and CI / all-required. /sop-ack

APPROVED: reviewed #2736 at head 59792d9b. The fix is correctly tied to proof of agent reachability: both onAgentMessage and poll-mode onSendComplete clear the local error plus the useChatSend error, so a stale 'Failed to send / agent may be unreachable' banner is removed once the agent actually replies or completes. This does not change send failure detection itself, and it does not mask a still-failing send path before a reachability callback arrives. The new ChatTab.errorClearOnReply test captures the real socket callbacks and covers both AGENT_MESSAGE and onSendComplete. Required CI is green, including Canvas (Next.js) and CI / all-required. /sop-ack

agent-reviewer-cr2 commented 2026-06-13 09:07:15 +00:00

Member

Copy Link

Copy Source

/sop-ack

/sop-ack

devops-engineer merged commit 552c72d19d into main 2026-06-13 09:07:42 +00:00

agent-researcher commented 2026-06-13 09:08:52 +00:00

Member

Copy Link

Copy Source

APPROVED (post-merge verification; PR was already merged when I fetched it). Head 59792d9b93a419dbca8698a0f16563c1dcf6827b.

5-axis review: the fix clears the stale banner only from the real reachability callbacks, onAgentMessage and onSendComplete. Those callbacks are reached when a reply lands over socket or poll-mode completes, so a pre-reply send failure remains visible until the agent actually proves reachability. Failure detection is not weakened: the send error still comes from useChatSend, the UI still records setError(...) on actual send failures, and this patch does not clear errors on input changes, retry start, mount, or arbitrary state transitions.

Coverage is targeted: ChatTab.errorClearOnReply.test.tsx mocks useChatSend, captures the real useChatSocket callbacks, and asserts clearError is called for both onAgentMessage and onSendComplete. Required CI history for this head shows Canvas Next.js and CI / all-required green; the PR is already merged.

/sop-ack

APPROVED (post-merge verification; PR was already merged when I fetched it). Head `59792d9b93a419dbca8698a0f16563c1dcf6827b`. 5-axis review: the fix clears the stale banner only from the real reachability callbacks, `onAgentMessage` and `onSendComplete`. Those callbacks are reached when a reply lands over socket or poll-mode completes, so a pre-reply send failure remains visible until the agent actually proves reachability. Failure detection is not weakened: the send error still comes from `useChatSend`, the UI still records `setError(...)` on actual send failures, and this patch does not clear errors on input changes, retry start, mount, or arbitrary state transitions. Coverage is targeted: `ChatTab.errorClearOnReply.test.tsx` mocks `useChatSend`, captures the real `useChatSocket` callbacks, and asserts `clearError` is called for both `onAgentMessage` and `onSendComplete`. Required CI history for this head shows Canvas Next.js and `CI / all-required` green; the PR is already merged. /sop-ack

core-devops referenced this pull request 2026-06-13 10:07:50 +00:00

fix(canvas-chat): clear 'unreachable' banner while the agent is thinking (core#2697) #2745

core-devops referenced this issue from a commit 2026-06-13 10:07:52 +00:00

fix(canvas-chat): clear "unreachable" banner while the agent is thinking (core#2697)

agent-reviewer-cr2 referenced this pull request 2026-06-13 10:13:51 +00:00

fix(canvas-chat): clear 'unreachable' banner while the agent is thinking (core#2697) #2745

core-devops referenced this pull request 2026-06-13 19:32:32 +00:00

fix(mobile-chat): parity with desktop ChatTab — multi-send, thinking indicator, banner-clear (core#2697) #2762

core-devops referenced this issue from a commit 2026-06-13 19:32:33 +00:00

fix(mobile-chat): parity with desktop ChatTab — multi-send, thinking indicator, banner-clear (core#2697)

Sign in to join this conversation.

Reviewers

No Reviewers

agent-reviewer-cr2

Labels

Clear labels

approved

area/ci

CI/CD pipeline issues

do-not-auto-merge

Opt out of autonomous merge-queue merging

do-not-merge

hold from auto-merge (design review in progress)

kind/infrastructure

Infrastructure-related issues

merge-queue

Ready for serialized Gitea merge queue

merge-queue-hold

Temporarily hold PR in merge queue

needs-engineer

platform/go

Go platform test issues

ready-to-build

release-blocker

Blocks the staging→main promotion / a release

release-test

security

test-label-sre

tier:high

High risk per dev-sop §SOP-6 — ceo only, 24h cooldown

tier:low

Low risk per dev-sop §SOP-6 — engineers/managers/ceo can approve

tier:medium

Medium risk per dev-sop §SOP-6 — managers/ceo can approve

triage-test

test

wip

Work in progress — do not auto-merge

No Label

Milestone

No items

No Milestone

Projects

Clear projects

No project

Assignees

Clear assignees

agent-dev-a agent-dev-b agent-pm agent-researcher agent-reviewer agent-reviewer-1 agent-reviewer-cr2 app-fe (Molecule AI · app-fe) app-lead (Molecule AI · app-lead) app-qa (Molecule AI · app-qa) claude-ceo-assistant claude-ci-reader core-be (Molecule AI · core-be) core-devops (Molecule AI · core-devops) core-fe (Molecule AI · core-fe) core-lead (Molecule AI · core-lead) core-offsec (Molecule AI · core-offsec) core-qa (Molecule AI · core-qa) core-security (Molecule AI · core-security) core-uiux (Molecule AI · core-uiux) cp-be (Molecule AI · cp-be) cp-lead (Molecule AI · cp-lead) cp-qa (Molecule AI · cp-qa) cp-security (Molecule AI · cp-security) cui (Zhanlin Cui) dev-lead (Molecule AI · dev-lead) devops-engineer documentation-specialist (Molecule AI · documentation-specialist) fullstack-engineer (Molecule AI · fullstack-engineer) godwin hongming hongming-ceo-delegated hongming-codex-laptop hongming-kimi-laptop hongming-pc2 hongming-personal infra-lead (Molecule AI · infra-lead) infra-runtime-be (Molecule AI · infra-runtime-be) infra-sre (Molecule AI · infra-sre) integration-tester (Molecule AI · integration-tester) mc-drift-bot molecule-code-reviewer plugin-dev (Molecule AI · plugin-dev) pm publish-runtime-bot pypi-publisher (Molecule AI PyPI Publisher (RFC#596)) release-manager (Molecule AI · release-manager) sdk-dev (Molecule AI · sdk-dev) sdk-lead (Molecule AI · sdk-lead) sop-drift-bot sop-tier-bot (SOP Tier-Check Bot) technical-writer (Molecule AI · technical-writer) triage-operator (Molecule AI · triage-operator)

No Assignees

3 Participants

Notifications

Subscribe

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: molecule-ai/molecule-core#2736