[hivecommons-hive]

withholdFix re-recorded an already-withheld head's exclusion every eval cycle, producing the 6-in-10m repetition behind #14870 and contradicting auditWithheldReviewFixes' doc ("The planner reports each head once"). Now the plan entry is recorded once per head (when the State.Withheld entry is first created) and the doc comment describes that. Adds a regression test running the planner twice over the same head.

Supersedes #14886 (opened against v6 by mistake; v5 is the fix base and forward-merges into v6).

Closes #14882
Closes #14883

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

[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 danathar 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

[hivecommons-hive]

hive/awaiting-ci: CI pending — sweep will merge when green.

---

🐝 Hive Agent: scanner | Instance: hosted-available-oke-11-placeholder-r05x | SHA: 73c30624a

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

[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.

PR	Title	Relation	Files
#14524	✨ feat(hub,agent): Hub-coordinated live CLI session pool for multi-tenant runner capacity	same files	src/cmd/hive/main.go
#14821	🐛 fix: preserve supersession decisions across stage reconciliation	same files	src/cmd/hive/main.go
#14862	🐛 fix: preserve supersession decisions across stage reconciliation	same files	src/cmd/hive/main.go

[hivecommons-hive]

Reviewed — no findings from this perspective.

— hive: agent=adjudicator-2 backend=copilot model=claude-fable-5.1 copilot=1.0.88

[hivecommons-hive]

Reviewed — no findings from this perspective.

— hive: agent=ci-maintainer 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 (intent-alignment): FixExclusions re-reports the same not-hive-authored head every cycle despite doc comment #14882 (folded into an existing issue) — Review backlog (intent-alignment): DispatchPlan.FixExclusions field doc still promises no silent planner branch (Review backlog (intent-alignment): FixExclusions re-reports the same not-hive-authored head every cycle despite doc comment #14882)

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]

Reviewed (docs-currency) — approve, one low advisory.

Read at head 73c3062. The behavior change in withholdFix (src/pkg/review/dispatch.go:1399-1425) and the reworded auditWithheldReviewFixes doc (src/cmd/hive/main.go:11146-11149) now agree, and TestWithheldFixExclusionReportedOncePerHead pins it. Two adjacent comments still describe the old every-cycle contract:

• src/pkg/review/dispatch.go:485-488 — the FixExclusions field doc says it holds "changes-requested PRs whose fix kick was deliberately not added to FixKicks this cycle". After this change an already-withheld not-hive-authored head's kick is still not added, but it no longer appears in FixExclusions (and so not in the fix_excluded / fix_excluded_reasons summary at main.go:10908-10909). Worth a clause like the one on WithheldFixes (dispatch.go:481-483): refusals already in State.Withheld are not repeated.
• src/cmd/hive/main.go:11151-11153 — "One WARN per changes_requested head that ends the cycle without a fix kick" is now only true for non-policy exclusions; an already-withheld head ends the cycle without a kick and without a line.

Neither blocks: the linked #14882/#14883 asked for exactly this dedup, and the changelog fragment (changelog.d/fixed-withheld-fix-exclusion-once.md) states the new behavior accurately.

— hive: agent=reviewer-2 backend=copilot model=claude-fable-5.1 copilot=1.0.88

[hivecommons-hive]

Style/maintainability — fix is correct and minimal; one advisory, non-blocking.

• src/pkg/review/fix_push_gate_test.go:273-290 — TestWithheldFixExclusionReportedOncePerHead re-builds the same two-cycle scenario that TestHumanPRFixWithheldWhenFixHumanPRsOff already runs at lines 71-80 (comment there: "Reported once per head: the next cycle finds the state entry and produces nothing new to audit"). The only new assertion is len(second.FixExclusions) != 0; adding that one check to the existing again block would keep the once-per-head contract pinned in one place. The new test also sets pr.HiveAttributed = true (line 275) with no comment on why — humanPR("sha1") alone already reaches withholdFix (as the existing test shows), so the flag reads as load-bearing when it is not. Every other test in this file carries a doc comment; this one does not.

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

[hivecommons-hive]

Evidence bundle `#1488773c3062` sha256:307a5cc0211a183b51a124f821cc6f45e8bedaf92b5e2e412bd6c2f100583902 — download from the dashboard or `hivectl review evidence #14887`.

[github-actions]

Thank you for your contribution! Your PR has been merged.

We'd love to hear how your experience was: share feedback