fix: use target default branch in repair plans#269
Conversation
|
Codex review: needs maintainer review before merge. Reviewed June 9, 2026, 7:16 PM ET / 23:16 UTC. Summary Reproducibility: yes. from source inspection: current Review metrics: 3 noteworthy metrics.
Merge readiness Overall follows the weaker of proof and patch quality, so missing proof can cap an otherwise strong patch. Rank-up moves:
Risk before merge
Maintainer options:
Next step before merge
Security Review detailsBest possible solution: Land the branch-preservation fix after maintainers explicitly accept the workflow input compatibility tradeoff, or revise the input surface before merge if manual Do we have a high-confidence way to reproduce the issue? Yes, from source inspection: current Is this the best way to solve the issue? Yes for the core bug: preserving AGENTS.md: found and applied where relevant. Codex review notes: model gpt-5.5, reasoning high; reviewed against 205a27a9d70d. Label changesLabel justifications:
Evidence reviewedWhat I checked:
Likely related people:
What the crustacean ranks mean
Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics. How this review workflow works
|
|
@clawsweeper re-review |
|
🦞🧹 I asked ClawSweeper to review this item again. Re-review progress:
|
|
@clawsweeper re-review Follow-up pushed in 3da99c9: |
|
🦞🧹 I asked ClawSweeper to review this item again. |
|
@clawsweeper re-review Follow-up pushed in 9c987f3: fixed the second |
|
🦞🧹 I asked ClawSweeper to review this item again. Re-review progress:
|
|
@clawsweeper re-review Follow-up pushed in 7af72ed: restored |
|
🦞🧹 I asked ClawSweeper to review this item again. Re-review progress:
|
|
@clawsweeper re-review Follow-up pushed in 5b2df8c: fixed the empty |
|
🦞🧹 I asked ClawSweeper to review this item again. Re-review progress:
|
Summary
main.target_branchthrough repair comment-router dispatch, event review payload resolution, and bothsweep.ymlmanual/scheduled planning fallbacks.apply_min_age_minutesworkflow input; to keep the workflow_dispatch input count under GitHub's limit, retire only the manualapply_progress_everylogging-cadence input and keep the internal default at10.target_branchargument so omitted branch dispatches stay omitted instead of becomingtrue.Review follow-up
3da99c9137added the manualtarget_branchinput and first fallback path.9c987f3e24fixed the secondsweep.ymlplanning resolver that still ignoredgithub.event.inputs.target_branch.7af72ed163restores minute-level apply throttling by bringing backapply_min_age_minutes, preserving it through continuation dispatches, and dropping onlyapply_progress_everyfrom the manual input surface.5b2df8c102appends--target-branchonly when non-empty in both repair comment-router jobs, with config coverage for the omitted-branch default.sweep.ymltarget-branch resolver blocks to readgithub.event.inputs.target_branchunderworkflow_dispatchbefore falling back toclient_payloadormain, verifies the dispatch input count stays under GitHub's limit, and verifies the router only passes--target-branchwhen non-empty.Live proof
Direct GitHub Actions workflow dispatch from this account is still blocked by repository permissions:
Positive non-main-default checkout proof against the real OpenClaw target repo:
Validation
pnpm run format && pnpm run build:repair && node --test test/repair/workflow-sparse-checkout.test.ts test/clawsweeper.test.ts --test-name-pattern 'sweep workflow preserves|workflow_dispatch input count|review continuations'(371 tests passed)pnpm run format && pnpm run build:repair && node --test test/repair/workflow-sparse-checkout.test.ts test/repair/comment-router-config.test.ts test/clawsweeper.test.ts --test-name-pattern 'target branch|comment router|workflow_dispatch input count|review continuations'(373 tests passed)pnpm run check(passed on Nodev24.15.0: active surface, limits, build, lint, unit, repair, changed coverage, full coverage, format)