fix(agent): log lifecycle-end completions at info/warn instead of error by chenxiaoyu209 · Pull Request #101704 · openclaw/openclaw

chenxiaoyu209 · 2026-07-07T13:59:41Z

What Problem This Solves

Successful agent run completions (stopReason=stop) were logged at console.error level, polluting error-level monitoring/alerting with false positives on every normal successful run. This also buries genuine failures in the noise.

Fixes #101678

Why This Change Was Made

The emitLifecycleEnd function unconditionally used console.error for every non-end_turn stop reason, even though stopReason=stop is a normal successful completion. The canonical terminal outcome helper (buildAgentRunTerminalOutcome) already classifies status=ok + stopReason=stop as reason=completed — but the logging path was not using this classification.

Per the AGENTS.md policy: "Agent run terminal state: normalize/merge via src/agents/agent-run-terminal-outcome.ts; do not rederive timeout/cancel precedence in projections."

User Impact

Before: Every successful run produces an error-level log line, triggering false alerts
After: Completed/cancelled runs log at info level; unusual non-error stop reasons log at warn level

Evidence

ClawSweeper review confirmed the bug with source-level reproduction
Existing agent-run-terminal-outcome.test.ts tests (14 passed) confirm the terminal outcome helper correctly classifies stopReason=stop as completed
agent-command.ingress-diagnostics.test.ts tests (13 passed) continue to pass with the change
The emitLifecycleEnd call site (agent-command.ts:2462) is only reached when no error is detected — the error path uses emitLifecycleResultError — so console.error was never the correct level for this path

Co-Authored-By: Claude <[email protected]>

clawsweeper · 2026-07-07T14:05:09Z

Thanks for the context here. I swept through the related work, and this is now duplicate or superseded.

Close: this PR overlaps the same agent lifecycle log-severity bug as the proof-positive sibling PR, while this branch still lacks after-fix real behavior proof and has an aborted-stop status regression that makes it a weaker landing path.

Root-cause cluster
Relationship: superseded
Canonical: #101703
Summary: This PR and the sibling PR both target the same successful stopReason=stop error-severity logging bug; the sibling PR is the stronger open landing candidate because it has real behavior proof and safer abort-status handling.

Members:

canonical: fix: lower successful agent stop completion logs #101703 - Open sibling PR for the same linked issue, same runtime/test files, sufficient real behavior proof, and no concrete code finding from this review.
fixed_by_candidate: Successful run completions (stopReason=stop) logged at error severity, sometimes twice per run #101678 - The shared open bug report is the issue that the canonical sibling PR is intended to fix.
duplicate: [Bug]: subagent normal completion logged as error level — "run ended with stopReason=stop" #101695 - Closed issue reports the same successful subagent stopReason=stop error-severity symptom and points back to the same root cause.

Proposal only: this assessment does not dispatch repair, suppress jobs, mutate sibling items, close, or merge anything.

Canonical path: Use #101703 as the canonical landing path for #101678, then close overlapping attempts once the chosen fix lands.

So I’m closing this here and keeping the remaining discussion on #101703 and #101678.

Review details

Best possible solution:

Use #101703 as the canonical landing path for #101678, then close overlapping attempts once the chosen fix lands.

Do we have a high-confidence way to reproduce the issue?

Yes from source inspection. Current main still routes a successful meta.stopReason === "stop" lifecycle end through console.error, and the terminal outcome helper already classifies successful stop metadata as completed.

Is this the best way to solve the issue?

No. Reusing the terminal outcome helper is the right direction, but this branch is not the best landing path because it lacks live proof, overlaps the proof-positive sibling PR, and misclassifies abort-path stop metadata when deriving the helper input status.

Security review:

Security review cleared: The diff only changes in-process agent logging severity and focused tests; it does not touch credentials, dependencies, workflows, package resolution, or other supply-chain surfaces.

AGENTS.md: found and applied where relevant.

What I checked:

Root policy read and applied: Root AGENTS.md was read fully; its terminal-state guidance says agent run terminal state should normalize through src/agents/agent-run-terminal-outcome.ts, which is the key review contract for this PR. (AGENTS.md:98, 453f5968bbca)
Scoped agent policy read: src/agents/AGENTS.md was read fully; its focused-test/runtime-import guidance was applied while checking the new agent-command import and diagnostics test shape. (src/agents/AGENTS.md:1, 453f5968bbca)
Current main still has the reported bug: Current main logs every lifecycle-end stop reason except end_turn through console.error, so the linked bug is real and still needs one canonical fix path. (src/agents/agent-command.ts:1877, 453f5968bbca)
Terminal outcome contract distinguishes successful stop from cancellation: The helper treats stopReason=stop as cancellation only when the terminal input is non-success, and returns error status for cancelled outcomes. (src/agents/agent-run-terminal-outcome.ts:122, 453f5968bbca)
This PR misderives aborted stop metadata: The current head constructs the terminal outcome with status ok whenever runResult.meta.error is absent, so an abort-path meta.aborted=true plus stopReason=stop can be normalized as completed. (src/agents/agent-command.ts:1904, 0b1238bbd15c)
Abort-path metadata source: buildAcpResult can emit cancelled ACP results as aborted=true with stopReason=stop, which is the metadata shape this PR needs to preserve when building the normalized outcome. (src/agents/command/attempt-execution.ts:899, 453f5968bbca)

Likely related people:

steipete: GitHub path history shows repeated recent work on terminal outcome precedence and agent-command-adjacent runtime behavior, including central terminal-outcome commits and audit/agent runtime changes. (role: recent area contributor; confidence: high; commits: f53346944db0, acb0e9c1552f, b1e5c9d7fa16; files: src/agents/agent-command.ts, src/agents/agent-run-terminal-outcome.ts, src/agents/agent-run-terminal-outcome.test.ts)
vincentkoc: GitHub path history shows recent terminal outcome helper refactors, and local blame/release-tag provenance also points through recent Vincent Koc commits around the affected files. (role: recent terminal-outcome contributor; confidence: high; commits: 68ef80116f42, dc384393fc8b, 098b471143c1; files: src/agents/agent-command.ts, src/agents/agent-run-terminal-outcome.ts)

Codex review notes: model internal, reasoning high; reviewed against 453f5968bbca.

…m logger, add test coverage

chenxiaoyu209 · 2026-07-07T16:29:55Z

@clawsweeper re-review

clawsweeper · 2026-07-07T16:29:58Z

🦞🧹
ClawSweeper re-review requested.

I asked ClawSweeper to review this item again.
Action: item re-review queued (workflow sweep.yml, event repository_dispatch).
Result: the existing ClawSweeper review comment will be edited in place when the review finishes.

fix(agent): use terminal outcome to set lifecycle-end log severity

602579d

openclaw-barnacle Bot added agents Agent runtime and tooling size: S labels Jul 7, 2026

chore: resolve merge conflict with upstream/main

56aea04

Co-Authored-By: Claude <[email protected]>

openclaw-barnacle Bot added size: XS and removed size: S labels Jul 7, 2026

clawsweeper Bot mentioned this pull request Jul 7, 2026

fix: lower successful agent stop completion logs #101703

Merged

fix(agent): extract resolveAgentRunLifecycleEndLogLevel, use subsyste…

0b1238b

…m logger, add test coverage

openclaw-barnacle Bot added size: S and removed size: XS labels Jul 7, 2026

clawsweeper Bot added the merge-risk: 🚨 other 🚨 Merging this PR has meaningful risk outside the owned taxonomy. label Jul 7, 2026

clawsweeper Bot mentioned this pull request Jul 7, 2026

fix(agents): downgrade stop/abort stop reasons from error to correct level #101897

Closed

chenxiaoyu209 closed this Jul 8, 2026

chenxiaoyu209 deleted the fix/issue-101678-stop-reason-log-level branch July 8, 2026 07:56

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

fix(agent): log lifecycle-end completions at info/warn instead of error#101704

fix(agent): log lifecycle-end completions at info/warn instead of error#101704
chenxiaoyu209 wants to merge 3 commits into
openclaw:mainfrom
chenxiaoyu209:fix/issue-101678-stop-reason-log-level

chenxiaoyu209 commented Jul 7, 2026

Uh oh!

clawsweeper Bot commented Jul 7, 2026 •

edited

Loading

Uh oh!

chenxiaoyu209 commented Jul 7, 2026

Uh oh!

clawsweeper Bot commented Jul 7, 2026 •

edited

Loading

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Uh oh!

Uh oh!

Conversation

chenxiaoyu209 commented Jul 7, 2026

What Problem This Solves

Why This Change Was Made

User Impact

Evidence

Uh oh!

clawsweeper Bot commented Jul 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

chenxiaoyu209 commented Jul 7, 2026

Uh oh!

clawsweeper Bot commented Jul 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

clawsweeper Bot commented Jul 7, 2026 •

edited

Loading

clawsweeper Bot commented Jul 7, 2026 •

edited

Loading