Skip to content

fix(ci): publish main dashboards through pull requests - #889

Draft
vincentkoc wants to merge 1 commit into
mainfrom
fix/pr-only-dashboard-publishers-20261122
Draft

vincentkoc wants to merge 1 commit into
mainfrom
fix/pr-only-dashboard-publishers-20261122

Conversation

@vincentkoc

Copy link
Copy Markdown
Member

Summary

  • Main dashboard writers currently push generated reports directly to the protected branch. Requiring pull requests prevents those refreshes from publishing.
  • Route both main publication paths through one publisher that retains the complete generated payload, opens a pull request, uses a normal exact-head squash merge, and verifies the actual merged README/report tree.
  • Preserve Check and Dependabot validation chains, exact generated-SHA Default Track status, and the original validation-run link. Latest-track publication waits for the real required check and never borrows that attestation.

Fixture impact

No fixture entries, submodule pins, or contract/smoke behavior change. Beta and development retain their existing publication paths. Main report jobs share noncancelling serialization; unresolved report branches or PRs hold later main refreshes instead of losing their payload.

Verification

  • 38 focused tests pass: 15 Git-backed publisher fixtures and 23 workflow contracts. Coverage includes real squash trees, stale sources, merge races, refused required checks, lost API responses, orphan/closed PRs, and exact status ownership.

  • JavaScript syntax, git diff --check, Actionlint syntax/expression checks, and standalone ShellCheck/Bash syntax for all 10 publisher shell blocks pass.

  • Production code adds 274 net lines for the shared publication/recovery boundary; tests add 272 and documentation adds 4. The reusable nonmain shell body is unchanged apart from indentation.

  • Fresh exact-diff autoreview and independent semantic review found no blocking findings.

Notes

An automation-created PR can require workflow approval. A retained PR is incomplete publication until its real required checks pass and it merges; the helper never bypasses that rule. Main report commits no longer request skipped CI.

No manual workflow dispatch, deployment, live runtime proof, credential change, or repository-policy change was performed. Local validation was limited to the changed publisher and workflow contracts; the full fixture corpus was not run locally.

@clawsweeper

clawsweeper Bot commented Nov 21, 2026 •

Copy link
Copy Markdown
Contributor

🦞👀
ClawSweeper picked this up.

Pull request received. I will update this pull request when review starts.

ClawSweeper review complete

ClawSweeper finished reviewing this revision. The review result is being finalized.

View the workflow run.

@vincentkoc
vincentkoc force-pushed the fix/pr-only-dashboard-publishers-20261122 branch from bebdb21 to 4a96f94 Compare November 21, 2026 16:08
@clawsweeper clawsweeper Bot added clawsweeper:intake:bug Preliminary submission classification; full source review pending. clawsweeper:intake:needs-repro Preliminary submission classification; full source review pending. P2 Normal priority bug or improvement with limited blast radius. merge-risk: 🚨 automation 🚨 Merging this PR could break CI, automerge, proof capture, label sync, or automation. rating: 🦪 silver shellfish Thin PR readiness signal; proof, validation, or implementation needs work. status: ⏳ waiting on author ClawSweeper has contributor-facing work open and is waiting for author action. and removed clawsweeper:intake:bug Preliminary submission classification; full source review pending. clawsweeper:intake:needs-repro Preliminary submission classification; full source review pending. labels Nov 21, 2026
@clawsweeper

clawsweeper Bot commented Nov 21, 2026

Copy link
Copy Markdown
Contributor

Codex review: needs changes before merge.

What this changes

Route main dashboard updates through a shared pull-request publisher that retains reports and verifies their merged contents.

Example: A successful main dashboard run generates updated README.md and reports.

  • Before: The workflow pushes directly to main and is rejected when pull requests are required.
  • After: The workflow retains a report branch, opens a PR, attempts an exact-head squash merge, and verifies the merged payload.

Review scores

Measure Result What it means
Overall readiness 🦪 silver shellfish (2/6) Useful publication repair with meaningful Git-backed coverage, but a deterministic concurrency defect blocks merging.
Proof confidence 🦐 gold shrimp (3/6) This CI automation change requires no shipped application proof; the supplied Git-backed fixtures exercise publication with mocked GitHub responses, and the maintainer explicitly reports no hosted publication run.
Patch quality 🦪 silver shellfish (2/6) 1 actionable review finding remain.

Product

Kind: Bug fix · Worth it: Yes · Fix scope: Complete
User problem: Generated main dashboards cannot publish when branch protection requires pull requests.
Reason: The maintainer-authored direction repairs both existing main writers while explicitly preserving report payloads and ordinary merge protection.

Merge readiness

⛔ Needs changes before merge - 1 item remains

Keep this maintainer-authored PR open: the publication repair remains necessary, but skipped matrix jobs can cancel pending main dashboard refreshes.

Priority: P2
Reviewed head: 4a96f947b5f0c29607f360aa8d862b3ce4d39c2a

Before merge

  • Keep unselected matrix jobs out of the main writer group (P2) - Every latest matrix leg now uses the shared main-writer group, even when this is a scheduled development run or a manual beta/development request. Selection happens only after job concurrency admission. If a main writer is active and a requested latest refresh or reusable publisher is pending, the next scheduled run's unselected latest leg replaces that pending job: cancel-in-progress=false protects the running job, not the single pending slot. This cancels real publication work without producing a replacement. The same skipped-matrix interference motivated fix(ci): protect track dashboard refreshes #65. Restrict shared-group membership to selected main-writing legs and retain isolated groups for unselected legs.

Findings

  • [P2] Keep unselected matrix jobs out of the main writer group — .github/workflows/track-dashboard.yml:31

Tests

  • Low-value test test/ci-workflows.test.mjs: The added shared-publisher test pins the literal concurrency expression, including its defective unselected-job grouping; replace those expression assertions with selected-versus-unselected group cases.
Agent review details

How this fits together

Check, Dependabot, and track-dashboard workflows generate compatibility reports; the publisher sends their README and report changes to main through GitHub pull requests.

flowchart TD
  A[Check and Dependabot validation] --> C[Main writer concurrency group]
  B[Latest track refresh] --> C
  C --> D[Generate README and reports]
  D --> E[Retain branch and open PR]
  E --> F[Normal squash merge]
  F --> G[Verify merged report tree]
Loading

Technical review

Best possible solution:

Keep the shared publisher and place only selected main-writing jobs in its concurrency group, preserving separate groups for unselected matrix legs.

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

Source establishes the direct-push publication path and the introduced concurrency failure: an active main writer, a pending requested main refresh, and the next scheduled development run are sufficient; no scheduler reproduction was executed.

Is this the best way to solve the issue?

A single publisher preserves payload and attestation ownership appropriately, but its shared concurrency group must exclude matrix jobs that will perform no main publication.

Full review comments:

  • [P2] Keep unselected matrix jobs out of the main writer group — .github/workflows/track-dashboard.yml:31
    Every latest matrix leg now uses the shared main-writer group, even when this is a scheduled development run or a manual beta/development request. Selection happens only after job concurrency admission. If a main writer is active and a requested latest refresh or reusable publisher is pending, the next scheduled run's unselected latest leg replaces that pending job: cancel-in-progress=false protects the running job, not the single pending slot. This cancels real publication work without producing a replacement. The same skipped-matrix interference motivated fix(ci): protect track dashboard refreshes #65. Restrict shared-group membership to selected main-writing legs and retain isolated groups for unselected legs.
    Confidence: 0.99

Overall correctness: patch is incorrect
Overall confidence: 0.96

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning medium; reviewed against 3009ed5399b4.

Provenance checked

Testing

Proof path: in-process harness.

Security

None.

Evidence

Acceptance criteria:

  • node --test test/ci-workflows.test.mjs test/publish-dashboard.test.mjs
  • actionlint .github/workflows/check.yml .github/workflows/dependabot-auto-merge.yml .github/workflows/publish-dashboard.yml .github/workflows/track-dashboard.yml
  • git diff --check

What I checked:

Likely related people:

  • Vincent Koc: Raw commit 8c17bd0 adds .github/workflows/publish-dashboard.yml:21 relative to its recorded parents. This identifies author metadata, not feature responsibility or a PR merger. (role: source-line author; confidence: high; commits: 8c17bd018daf; files: .github/workflows/publish-dashboard.yml)

Preflight source

Source unknown. Source comparison at the observed PR snapshot only. Completion is self-reported; body, policy, tests, execution, and report bytes are not verified.

Review metrics

Metric Value Why it matters
Change distribution Automation +302/-28; tests +276/-4; documentation +4/-0 The shared publisher accounts for most production growth; its local Git fixtures cover distinct payload and receipt failures.

Labels

Label changes:

  • add P2: Dashboard publication is a bounded repository automation repair with a concrete pending-job cancellation defect.
  • add merge-risk: 🚨 automation: The changed concurrency key lets unselected scheduled matrix jobs cancel pending dashboard publication jobs.
  • add rating: 🦪 silver shellfish: Overall readiness is 🦪 silver shellfish; proof is 🦐 gold shrimp and patch quality is 🦪 silver shellfish.
  • add status: ⏳ waiting on author: ClawSweeper has contributor-facing work open and is waiting for author action.

Label justifications:

  • P2: Dashboard publication is a bounded repository automation repair with a concrete pending-job cancellation defect.
  • merge-risk: 🚨 automation: The changed concurrency key lets unselected scheduled matrix jobs cancel pending dashboard publication jobs.
  • rating: 🦪 silver shellfish: Overall readiness is 🦪 silver shellfish; proof is 🦐 gold shrimp and patch quality is 🦪 silver shellfish.
  • status: ⏳ waiting on author: ClawSweeper has contributor-facing work open and is waiting for author action.

Rating scale

6/6 🦀 challenger crab · 5/6 🦞 diamond lobster · 4/6 🐚 platinum hermit · 3/6 🦐 gold shrimp · 2/6 🦪 silver shellfish · 1/6 🧂 unranked krab. Overall follows the weaker of proof and patch quality; ✨ marks media proof (a screenshot, video, or linked artifact) that directly shows the changed behavior.

Workflow

ClawSweeper edits this one comment on every review. Comment @clawsweeper re-review for a fresh review only. Repair and merge execution lanes are retired.

Reviewed November 21, 2026, 11:23 AM ET / 16:23 UTC.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

merge-risk: 🚨 automation 🚨 Merging this PR could break CI, automerge, proof capture, label sync, or automation. P2 Normal priority bug or improvement with limited blast radius. rating: 🦪 silver shellfish Thin PR readiness signal; proof, validation, or implementation needs work. status: ⏳ waiting on author ClawSweeper has contributor-facing work open and is waiting for author action.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant