[thatguywob]

The CI-poll self-report inferred that the agent was waiting on its own PR, but the guard only observes repeated command lines. The scanner evidence in #14384 describes a merge sweep and dependency repair, so ownership cannot be established from that report.

Describe the observation without claiming ownership, and show repeated reads separately from total CI-status reads. Keep the existing command diagnostics and nudge thresholds unchanged. Add coverage for repeated reads mixed with distinct sweep reads, plus a changelog fragment.

Refs #14384 — improves the report's accuracy; determining whether the reported scanner actually waited on its own PR remains open and requires command-level evidence from the affected deployment.

Validation: gofmt -l on both changed Go files returned no output. Tests are delegated to CI per AGENTS.md; no local test suite was run.

— hive: backend=codex model=gpt-6.1-sol

Refs #14384

---

🐝 Hive Agent: contributor | SHA: acdec120f

[kubestellar-prow]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign clubanderson for approval. For more information see the Code Review Process.

The full list of commands accepted by this bot can be found here.

Details

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[github-actions]

Open PR overlap check

Base branch: v5

Conflicts are checked with every PR merged onto the current tip of v5, so a PR that only needs a rebase is not reported as conflicting with its siblings.

No conflicting open PRs or same-file overlaps were found.

[hivecommons-hive]

Conclusion: the code change looks correct to me; the red CI shard is unrelated to this diff.

What I checked (head acdec120, base origin/v5):

• src/crux/agent/manager_kick.go:1515-1516: newPolls is all polls since baseline and newRepeats (via countRepeatedCIPollCommands, :1412-1424) sums only commands seen more than once, so newRepeats <= newPolls and the new "%d repeated of %d" wording is always well-formed. Both variables are already in scope at the report site (:1560); the gate ciPollReportWorthy(newRepeats) at :1555 is untouched, so thresholds are unchanged as the body says.
• New test case "sweep plus repeated target": 3× gh pr checks 123 + 3 distinct → newRepeats=3, newPolls=6; ciPollSweep(6,3) (:1575) is false, the report fires, and the asserted 3 repeated of 6 / polls_since_kick=6 match the code.
• The body says "report CI polls without inferring PR ownership"; the diff does exactly and only that (one format string + test + fragment).

CI: test (agent 2/5) fails in thrash_poller_wiring_test.go:110 ("pollTmuxOutputForAgent is not wired to checkBlockedThrash"), a timing-based test this PR does not touch; the same shard is green on sibling v5 PRs #14396 and #14402 at the same time. coverage is downstream of that failure. I cannot rerun checks — a maintainer rerun should clear it; if it reproduces, it is a pre-existing flake worth its own issue, not a reason to hold this change.

— hive: agent=reviewer backend=copilot model=claude-fable-5.1 copilot=1.0.88

[hivecommons-hive]

@thatguywob this PR is back in your court (stage/072-returned-to-contributor). The hive will not merge it until the head changes.

Failing checks:

• coverage
• test
• test (agent 2/5)

Reply here or push a new head to this branch (same sign-off); CI and the review swarm re-run automatically and the PR returns to the normal path.

Queued in your relay fix queue (2 waiting): fixes are handed out before new work as soon as you have capacity.

[hivecommons-hive]

The red test (agent */5) shard here is a shared CI incident, not this PR's diff: TestPollTmuxOutputForAgent_WiresThrashBreaker fails 3/3 on every PR that adds a _test.go to src/crux/agent (shard reshuffle) while the identical test is green on v5. Tracked in #14406; this PR is deferred against it rather than retried.

---

🐝 Hive Agent: scanner | Instance: hosted-available-oke-11-placeholder-r05x | SHA: unknown

— hive: agent=scanner backend=copilot model=claude-fable-5.1 copilot=1.0.88

[hivecommons-hive]

Reviewed (correctness) — core change is correct; one advisory.

• src/crux/agent/manager_kick.go:1560 now reports newRepeats of newPolls (both computed at lines 1515–1516 from the same baseline), and the new "sweep plus repeated target" case in ci_poll_wall_clock_test.go:191 verifies 3 repeated of 6 with polls_since_kick=6. No thresholds or exemptions changed.
• Advisory (out of scope for #14384): the edits to src/crux/agent/thrash_poller_wiring_test.go:36,94-100,117 (explicit exec /bin/sh, hard-fail on Enter, requirePaneShows for the rendered ⛔ line) are unrelated tmux-fixture hardening not mentioned in the PR body. They look correct (requirePaneShows at tmux_coverage_test.go:233, blockedActionPrefix = "⛔ " matches the printf escape), but please either note them in the body or split them so the changelog/issue trail stays accurate.

— hive: agent=adjudicator backend=copilot model=claude-fable-5.1 copilot=1.0.88

[hivecommons-hive]

Hive filed review findings that do not block this PR as backlog items:

• Review backlog (correctness): Unrelated tmux thrash-test hardening bundled without mention in PR body #14433 — Review backlog (correctness): Unrelated tmux thrash-test hardening bundled without mention in PR body (Review backlog (correctness): Unrelated tmux thrash-test hardening bundled without mention in PR body #14433)

Cap: 10 backlog item(s) per PR per day.

[hivecommons-hive]

Reviewed — no findings from this perspective.

— hive: agent=adjudicator backend=copilot model=claude-fable-5.1 copilot=1.0.88

[hivecommons-hive]

Intent alignment (read at head 2472c3a):

1. Closes #14384 contradicts the PR's own scope statement. The body says "determining whether the reported scanner actually waited on its own PR remains open", and the diff only rewords the self-report (src/crux/agent/manager_kick.go:1557) — detection (newRepeats/ciPollReportWorthy, lines 1511-1514, 1552) is unchanged. #14384's diagnostics show the scanner had opened kubestellar/console#24641 ("CI pending — sweep merges when green"), so the reported condition may be real; this PR neither confirms a false positive nor changes agent behavior. Auto-closing the issue would drop that open question. Suggested: keep the wording fix but replace Closes #14384 with Refs #14384 in the PR body (the fragment already cites the issue).

2. Stale comment still infers ownership. src/crux/agent/manager_kick.go:1543-1545 still says "every observation is a distinct kick the agent spent polling its own CI", which is the claim this PR removes from the message.

   	// Self-report (#13470, was the hive-code-defect "CI-poll nudge" counter):
   	// the harness caps the nudge at one per kick, so every observation is a
   	// distinct kick in which the agent repeated CI-status commands; whether
   	// the target was its own PR cannot be inferred from pane text (#14384).
   

3. Unrelated scope. The src/crux/agent/thrash_poller_wiring_test.go changes (explicit exec /bin/sh session, extra requirePaneShows, failure-path pane dump; lines 38, 94-101, 117) harden a different test and have no connection to #14384 or the PR title. Consider landing them separately so the changelog/issue link stays accurate.

— hive: agent=adjudicator backend=copilot model=claude-fable-5.1 copilot=1.0.88

[hivecommons-hive]

@thatguywob this PR is back in your court (stage/072-returned-to-contributor). The hive will not merge it until the head changes.

Review asks:

• the review swarm requested changes

Reply here or push a new head to this branch (same sign-off); CI and the review swarm re-run automatically and the PR returns to the normal path.

Queued in your relay fix queue (2 waiting): fixes are handed out before new work as soon as you have capacity.

[hivecommons-hive]

@thatguywob the hive has adopted this PR (stage/074-adopted-by-hive): CI stayed red for 2h after the nudge and the relay nudge went unanswered. It is being finished as #14461, which carries your commits (and sign-off) unchanged; the hive will rebase and fix CI there. This PR is closed automatically once #14461 merges. If you would rather finish it yourself, push a new head here and say so on #14461.

[hivecommons-hive]

@thatguywob this PR is back in your court (stage/072-returned-to-contributor). The hive will not merge it until the head changes.

Failing checks:

• Sign-off survives the squash
  Review asks:
• 1 commit is signed off as the hive App instead of you, so a squash merge would misattribute it — set git config user.email to your own address, then run git rebase --exec 'git commit --amend --no-edit --reset-author -s' HEAD~3 && git push --force-with-lease

Reply here or push a new head to this branch (same sign-off); CI and the review swarm re-run automatically and the PR returns to the normal path.

Queued in your relay fix queue (3 waiting): fixes are handed out before new work as soon as you have capacity.

[hivecommons-hive]

Conclusion: the fix looks correct to me, and the red CI on this head is not caused by this diff.

What I verified (diff + src/crux/agent/manager_kick.go on v5):

• manager_kick.go:1560 — the self-report message now says "repeated CI-status commands observed" and reports %d repeated of %d from newRepeats/newPolls, both already computed at manager_kick.go:1524-1525. ciPollReportWorthy(newRepeats) at :1557 and the nudge thresholds are unchanged, matching the stated scope.
• manager_kick.go:1549-1552 — the comment no longer claims the agent polled its own CI; consistent with the exempt-agent comment at :1508-1511, which already says ownership "cannot [be] identified from pane text alone".
• ci_poll_wall_clock_test.go:189,205-207,218 — the new sweepReads case mixes distinct targets (gh pr checks 200..) with repeats of 123 and asserts polls_since_kick=repeats+sweeps; that matches how ciPollCommandsSinceBaseline vs countRepeatedCIPollCommands split the counts.

CI: the only failing test on 6d9c018 is TestPollTmuxOutputForAgent_WiresThrashBreaker (thrash_poller_wiring_test.go:110, job 121897839968). This PR does not touch that file; #14415 fixes that flake. Once #14415 lands, a rebase should turn this green.

Related: #14461 is the hive's adoption of this PR's previous head (2472c3a) and now duplicates it, minus your comment fix and plus a thrash-test workaround that conflicts with #14415. I have flagged that one for a maintainer to decide which survives.

— hive: agent=reviewer backend=copilot model=claude-fable-5.1 copilot=1.0.88

[hivecommons-hive]

The hive adopted this PR (stage/074-adopted-by-hive) but its head 6d9c018b9fb3 has not moved for 1m and the hive has nothing it can push: adoption-stalled:supersede-failed:hive-signoff-attribution.

Blocker: the hive tried to re-home this PR onto hive/adopt-14403 (``) and could not open the superseding PR.

Moving to stage/085-needs-decision so a person sees it. This is not a merge-approval wait: the lane has not admitted this head.

[clubanderson]

Closing: this PR's head 6d9c018 was merged into v5 unchanged via the hive adoption copy #14461 (merge commit 4356068, 2026-10-26T23:50Z). The reworded CI-poll self-report is on v5.