[Fleet Monitor] petry-projects/google-app-scripts — .github/workflows/coverage.yml #500

Activity

1. don-petry

   added

   health-check

   dev-leadFor dev-lead agent pickupFor dev-lead agent pickup

   fleet-tracker

   on Jul 29, 2026
2. don-petry commented on Jul 29, 2026

   

   don-petry

   on Jul 29, 2026

   CollaboratorAuthor

   More actions

   Dev-Lead Implementation Plan

   Issue: #500 — [Fleet Monitor] petry-projects/google-app-scripts — .github/workflows/coverage.yml

   Investigation Summary

   Over the last 100 coverage.yml runs: 98 success, 2 action_required, 0 genuine test failures. The DEGRADED signal (2/9 = 22.2%) comes from GitHub's approval-gate (action_required) runs on bot-authored PRs being counted as failures — not from flaky tests or coverage regressions. Rapid successive pushes (dev-lead commits, ci.yml autofix commits, dependabot rebases) queue redundant coverage.yml runs that accumulate in the gated state.

   Scope

   Add a top-level concurrency group to coverage.yml (keyed on ${{ github.workflow }}-${{ github.ref }}, cancel-in-progress: true) so superseded runs are cancelled rather than lingering — the same fix pattern previously applied to this workflow (#390) and ci.yml (#457).

   Implementation Checklist

   • .github/workflows/coverage.yml — add concurrency block after permissions

   Tests to Write

   • None — CI-config-only change to a workflow YAML with no JS/GAS logic surface and no workflow-YAML test harness in this repo (Phase 2 TDD exception, per [Fleet Monitor] petry-projects/google-app-scripts — .github/workflows/coverage.yml #390 precedent). Verified via existing full suite + lint.

   Test Command

   npm test -- --coverage && node scripts/check-coverage.js
3. don-petry commented on Jul 29, 2026

   

   don-petry

   on Jul 29, 2026

   CollaboratorAuthor

   More actions

   Dev-Lead: Implementation Complete

   Plan Execution

   • Investigated 100 coverage.yml runs: 98 success, 2 action_required, 0 genuine test failures — DEGRADED signal is approval-gated bot-PR runs counted as failures, not flaky tests
   • Added concurrency block to .github/workflows/coverage.yml (matches [Fleet Monitor] petry-projects/google-app-scripts — .github/workflows/coverage.yml #390/[Fleet Monitor] petry-projects/google-app-scripts — .github/workflows/ci.yml #457 precedent)
   • Verified full test suite, coverage gate, and dev-lead lint all pass

   Test Results

   Test Suites: 12 passed, 12 total
   Tests:       301 passed, 301 total
   Snapshots:   0 total
   Time:        6.067 s
   
   📊 Coverage Report:
   ✅ statements     : 98.91% (999/1010) [required: 95%]
   ✅ branches       : 90.00% (558/620) [required: 85%]
   ✅ functions      : 98.00% (98/100) [required: 95%]
   ✅ lines          : 100% (0/0) [required: 100%]
   ✅ All coverage metrics meet requirements.
   
   [lint] all checks passed  (.dev-lead/scripts/dev-lead-lint.sh → exit 0)
   

   Files Changed

   • .github/workflows/coverage.yml: added top-level concurrency group keyed on
     ${{ github.workflow }}-${{ github.ref }} with cancel-in-progress: true, so rapid
     successive pushes (dev-lead commits, ci.yml autofix commits, dependabot rebases) cancel
     superseded runs instead of accumulating in the action_required approval-gate state that
     the monitor counts as failures.

   Notes

   • Root cause: the 22.2% "failure rate" was 2 action_required runs (GitHub's approval
     gate on bot-authored PRs where jobs never executed) — not test flakiness or coverage
     regressions. cancel-in-progress reduces redundant gated runs by cancelling superseded
     ones (cancelled runs are excluded from the monitor's failure rate).
   • Per-PR isolation: for PRs github.ref is refs/pull/N/merge (unique per PR), so
     distinct PRs never cancel each other; only older commits of the same PR are superseded.
     Latest-commit runs are never cancelled, so no required check is lost.
   • Phase 2 (TDD) exception: CI-config-only change to a workflow YAML — no JS/GAS logic
     surface and no workflow-YAML test harness in this repo. Verified via existing full suite +
     lint (no regressions), consistent with the [Fleet Monitor] petry-projects/google-app-scripts — .github/workflows/coverage.yml #390 resolution.
   • Pre-existing, unrelated: npm run check reports prettier formatting warnings in 4
     unrelated workflow files (add-to-project.yml, auto-rebase.yml,
     dependabot-automerge.yml, pr-auto-review.yml) that exist on main. Left untouched per
     the "do not modify unrelated files" constraint. coverage.yml itself is prettier-clean.
4. don-petry commented on Aug 1, 2026

   

   don-petry

   on Aug 1, 2026

   CollaboratorAuthor

   More actions

   ✅ Auto-resolved by Fleet Monitor. This workflow's failure rate has been below the 10% threshold (or the workflow was removed) since 2026-07-29 (3 days) — the monitor stopped flagging it. Closing. It will reopen automatically if the workflow crosses the threshold again.
5. don-petry

   closed this as completedon Aug 1, 2026