Repository navigation
fix(ci): batch main CodeQL scans weekly - #33
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 changesAdds weekly Monday CodeQL scans of main while retaining pull-request and manual analysis. Example: A change reaches main on Tuesday.
Review scores
ProductKind: Maintenance · Worth it: Yes · Fix scope: Complete Merge readiness✅ Ready for maintainer review No actionable code defect found at the requested head; source landing is consistent with the maintainer’s explicit direction, while the scanning ownership cutover remains separately authorized. Priority: P3 Before mergeNone. FindingsNone. Agent review detailsHow this fits togetherGitHub Actions sends repository source through Actions and JavaScript/TypeScript CodeQL analysis to GitHub code scanning, independently of Worker publication and deployment. flowchart TD
A[Pull request] --> D[CodeQL workflow]
B[Monday schedule] --> D
C[Manual dispatch] --> D
D --> E[Actions and JavaScript analysis]
E --> F[GitHub code scanning]
G[Existing Deploy workflow] --> H[Cloudflare Worker]
Technical reviewBest possible solution: Land this focused workflow, then activate weekly scanning through the separately authorized continuity-or-rollback handoff. Do we have a high-confidence way to reproduce the issue? This is an operational cadence change rather than a reported runtime defect; hosted execution confirms the candidate reaches the existing default-setup ownership boundary. Is this the best way to solve the issue? A single standard CodeQL workflow implements the accepted cadence without a custom delta-skip mechanism or changes to deployment ownership. AGENTS.md: not found in the target repository. Codex review notes: model internal, reasoning medium; reviewed against 34be4fb9b844. Provenance checked
TestingProof path: shipped entry point. SecurityNone. EvidenceWhat I checked:
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. 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 HistoryReview history (1 earlier review cycle)
Reviewed November 14, 2026, 1:07 AM ET / 06:07 UTC (Revision 2). |
…l-weekly-20261103 * commit '34be4fb9b8447a7f9fcdb24c17bd15fc23bb5678': chore(telemetry): refresh released public vocabulary chore(telemetry): refresh released public vocabulary chore(telemetry): refresh released public vocabulary chore(telemetry): refresh released public vocabulary
|
@clawsweeper re-review Please review exact head Normal source landing is authorized after this exact-head review clears. The draft will be marked ready at that point. The updated body retains the separate post-merge same-SHA full historical alert/analysis/settings baseline, reviewed canonical rollback, and 20-minute continuity-or-rollback gate. Default setup remains enabled; advanced processing has not passed and no settings cutover is authorized by this comment. |
|
🦞👀 Re-review progress:
|
Main pushes repeat CodeQL work in this quiet repository. Batch full Actions and JavaScript/TypeScript default analysis weekly on Mondays at 05:37 UTC while retaining full PR/manual scans. Scheduled scans do not cancel running scans; only superseded PR scans cancel. The one-line root
.ci-policy.mdand changelog explain why. Weekly full scans cover accumulated changes without introducing a separate delta-skip implementation.Existing PR validation and serialized main deployment remain unchanged. A normal main merge may invoke the existing released-vocabulary publisher and deployment. The publisher commits only its two vocabulary files, preserving this workflow. Preserve attached security configuration 268171, which allows advanced scanning.
Repository maintainers own explicit language-matrix additions, pinned action updates and checking/re-enabling scheduled Actions after 60 days of repository inactivity. Scheduled results do not guarantee exact release-SHA security proof.
The old vocabulary blocker is resolved by the existing daily producer, not a manual product refresh: producer 42952311331 successfully generated, checked, published and deployed the released vocabulary. This branch now includes main
34be4fb9b8447a7f9fcdb24c17bd15fc23bb5678; only its two generated vocabulary files changed. The resulting PR diff remains the workflow, one-line rationale and changelog.Validation at
7885e456792aeb818774dd5ecdd641fce5853493: actionlint, diff checks, deterministic vocabulary validation and latest-release coverage forv2026.11.3(847611dcaa3bf0e2783335f961dd1fd0580e4d9f) passed. Fresh full-context P0–P3 autoreview and independent publication review found no blocking findings. Exact-head native Deploy/check passed, including the latest-release gate and Worker dry-run. Both managed-default language analyses passed. The candidate workflow completed 17 Actions and 85 JavaScript security queries; each job reported only the expected default-ownership SARIF rejection as an error. This verifies candidate extraction/query execution, not advanced processing or full main coverage. Fresh hosted review at this exact head found no actionable findings or before-merge items; independent maintainer review also cleared source landing. The separately controlled scanning ownership cutover remains pending.Default setup remains enabled. Candidate extraction/security-query completion is distinct from successful processing: advanced SARIF processing is expected to reject while default setup owns scanning, and that rejection is not a successful scan. Do not rerun the known ownership failure as coverage proof or weaken upload behavior to make the migration appear green.
Handoff: require successful native CI, successful managed-default PR analyses, completed candidate extraction/security queries with only the expected ownership rejection, and refreshed exact-head review before normal source merge. After merge, capture a fresh baseline tied to the same merged main SHA: default settings, attached security configuration/features/rules, every alert ID/state/configuration, all historical alert instances and full analysis history. Review the exact canonical rollback before any settings change. Only then disable default setup and dispatch full main analysis, verifying processing for all languages/categories and alert continuity within a 20-minute operator window; otherwise restore the captured default configuration and verify rollback. PR delta results do not establish full main coverage. No ownership cutover has happened yet.