Repository navigation
improve(ci): split native platform checks across PRs and main - #89
Conversation
|
🦞👀 Pull request received. I will update this pull request when review starts. ClawSweeper review completeClawSweeper finished reviewing this revision. The review result is being finalized. |
|
Codex review: needs maintainer review before merge. What this changesThis PR runs native macOS and Windows checks on Node 24 for PRs and Node 22 for main, retaining all Ubuntu checks and full manual coverage. Example: A pull request targeting main triggers CI.
Review scores
ProductKind: Maintenance · Worth it: Yes · Fix scope: Complete Merge readiness✅ Ready for maintainer review Keep this maintainer-authored draft open; the matrix implements the stated coverage tradeoff correctly, hosted PR checks pass, and no actionable defect was found. Priority: P3 Before mergeNone. FindingsNone. Agent review detailsHow this fits togetherGitHub Actions turns PR, main-push, and manual events into OS/Node jobs that install frozen dependencies and run the package quality gate. flowchart TD
A[GitHub event] --> B[Event-aware matrix]
B --> C[Ubuntu on three Node versions]
B --> D[Native platforms on selected Node versions]
C --> E[Frozen dependency installation]
D --> E
E --> F[Full package quality gate]
Technical reviewBest possible solution: Retain the compact event-aware matrix and matching documentation with the explicitly accepted post-merge Node 22 coverage tradeoff. Do we have a high-confidence way to reproduce the issue? No runtime bug is claimed; baseline CI confirms the former seven-job matrix, and hosted PR execution confirms the new five-job selection. Is this the best way to solve the issue? The existing matrix owner handles the split directly, retains full checks for every selected pair, and preserves all seven combinations for manual dispatch. AGENTS.md: found and applied where relevant. Codex review notes: model internal, reasoning medium; reviewed against 6fefc0bb9054. Provenance checked
TestingProof path: shipped entry point. SecurityNone. EvidenceWhat I checked:
Likely related people:
Preflight sourceSource 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
LabelsLabel changes:
Label justifications:
Rating scale6/6 🦀 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. WorkflowClawSweeper edits this one comment on every review. Comment Reviewed November 20, 2026, 4:10 AM ET / 09:10 UTC. |
…/test-fixtures * commit 'be1709487ea73b28c2ff08b1e96403d93de8af65': improve(ci): split native platform checks across PRs and main (#89) improve(ci): use the prebuilt actionlint image (#88) refactor(ci): skip ineligible dispatch events before runner allocation (#94) improve(ci): right-size runners for GitHub metadata jobs (#92) improve(ci): reuse pnpm packages across lockfile updates (#91) refactor(router): consolidate match reuse and path normalization (#90) # Conflicts: # test/router-navigation.test.ts
…e-mappings * commit 'be1709487ea73b28c2ff08b1e96403d93de8af65': improve(ci): split native platform checks across PRs and main (#89) improve(ci): use the prebuilt actionlint image (#88) refactor(ci): skip ineligible dispatch events before runner allocation (#94) improve(ci): right-size runners for GitHub metadata jobs (#92) improve(ci): reuse pnpm packages across lockfile updates (#91) refactor(router): consolidate match reuse and path normalization (#90) # Conflicts: # src/matches.ts # src/router.ts
…cription-disposal * commit 'be1709487ea73b28c2ff08b1e96403d93de8af65': improve(ci): split native platform checks across PRs and main (#89) improve(ci): use the prebuilt actionlint image (#88) refactor(ci): skip ineligible dispatch events before runner allocation (#94) improve(ci): right-size runners for GitHub metadata jobs (#92) improve(ci): reuse pnpm packages across lockfile updates (#91) refactor(router): consolidate match reuse and path normalization (#90) # Conflicts: # src/matches.ts
…d-revalidation * commit 'be1709487ea73b28c2ff08b1e96403d93de8af65': improve(ci): split native platform checks across PRs and main (#89) improve(ci): use the prebuilt actionlint image (#88) refactor(ci): skip ineligible dispatch events before runner allocation (#94) improve(ci): right-size runners for GitHub metadata jobs (#92) improve(ci): reuse pnpm packages across lockfile updates (#91) refactor(router): consolidate match reuse and path normalization (#90) # Conflicts: # src/router.ts
What Problem This Solves
A PR and its merge each scheduled both Node 22 and Node 24 on Windows and macOS. In the preceding 25 CI runs all seven package checks passed on the first attempt.
User Impact
PRs exercise both native platforms on Node 24. Main exercises both on Node 22. Ubuntu runs the full check on all three Node versions on both events. A manual run still tests all seven pairs together. A Node 22 issue specific to macOS or Windows can now first surface after merge.
Why This Change Was Made
Five package jobs per automatic run cover every OS and Node major. Keeping every OS on main also seeds default-branch pnpm caches for PRs. On the baseline main run, the two Node 24 native jobs used 119 runner-seconds together.
Evidence
Local matrix expansion produced five distinct jobs on PRs, five on pushes, and seven on manual dispatch. Actionlint, YAML parsing, and
git diff --checkpassed. Every selected platform runs the fullpnpm run check. Baseline CI