Skip to content

[Fleet Monitor] petry-projects/TalkTerm β€” .github/workflows/ci.ymlΒ #470

Closed

Description

@don-petry

🟑 WARNING: petry-projects/TalkTerm / .github/workflows/ci.yml

Metric Value
Failure Rate 11.8% (2 / 17 runs)
Status WARNING
p50 Duration 33s
p95 Duration 150s
Successful Runs 15
Cancelled Runs 0

Workflow: .github/workflows/ci.yml

Last updated by Fleet Monitor run on 2026-09-01.
Threshold: failure rate > 10% over the monitored window.

Activity

  1. added
    dev-leadFor dev-lead agent pickup
    fleet-trackerTracked by the Actions Fleet Monitor
    health-checkAutomated health check result
    on Sep 1, 2026
  2. don-petry commented on Sep 1, 2026

    @don-petry
    ContributorAuthor

    Dev-Lead Implementation Plan

    Issue: #470 β€” [Fleet Monitor] petry-projects/TalkTerm β€” .github/workflows/ci.yml

    Investigation

    The 2 failing runs in the monitored window were both on the feature branch
    worktree-implement-sprint-1, in the quality matrix job at the
    npm run test:coverage step:

    ERROR: Coverage for branches (89.73%) does not meet global threshold (90%)
    

    That is legitimate quality-gate enforcement, not workflow flakiness β€” the CI
    correctly failed branch code that missed the 90% branch-coverage threshold. That
    gate will not be weakened.

    Scope

    Following the repo's established pattern (Fleet Monitor #380 concurrency, #430 yq
    checksum, #438 curl --retry β€” each a durable invariant locked by a regression
    guard), harden ci.yml against the remaining CI-inflating failure class it does
    not yet defend: jobs with no timeout-minutes. Without a job timeout, a
    stalled step (a curl that connects then hangs β€” --retry does not cover stalls β€”
    or any hung process) runs to GitHub's 6-hour default before being killed and
    counted as a timeout failure. p95 duration is 150s, so any multi-minute run is
    anomalous. timeout-minutes is already a convention in 4 sibling workflows
    (sonarcloud, copilot-setup-steps, initiative-driver, feature-ideation).

    Implementation Checklist

    • .github/workflows/ci.yml β€” add timeout-minutes to both jobs (secret-scan, workflow-tests)
    • scripts/test-ci-workflow.sh β€” add Check 7: every job must declare a positive-integer timeout-minutes
    • scripts/test-ci-workflow.test.sh β€” add fixtures/cases: reject a job missing timeout-minutes, accept jobs that declare it

    Tests to Write

    • Case: workflow with a job missing timeout-minutes is REJECTED
    • Case: workflow whose every job declares timeout-minutes is ACCEPTED
    • Case: timeout-minutes: 0 (invalid) is REJECTED
    • Re-run existing guard against the real ci.yml after the fix β€” passes

    Test Command

    bash scripts/test-ci-workflow.sh && bash scripts/test-ci-workflow.test.sh (full guard suite via scripts/test-*-workflow*.sh)

  3. don-petry commented on Sep 1, 2026

    @don-petry
    ContributorAuthor

    Dev-Lead: Implementation Complete

    Investigation summary

    The 2 failing runs in the monitored window were not workflow flakiness. Both were
    on the feature branch worktree-implement-sprint-1, in the quality matrix job at
    npm run test:coverage:

    ERROR: Coverage for branches (89.73%) does not meet global threshold (90%)
    

    That is the CI correctly enforcing the 90% branch-coverage gate against branch code β€”
    legitimate quality-gate enforcement, not a defect in ci.yml. That gate was not
    weakened. Instead, following the repo's established pattern (Fleet Monitor #380 concurrency,
    #430 yq checksum, #438 curl --retry β€” each a durable invariant locked by a regression
    guard), I hardened ci.yml against the remaining CI-inflating failure class it did not yet
    defend: jobs with no timeout-minutes.

    Plan Execution

    Why timeout-minutes

    Without a job timeout, a step that stalls rather than errors β€” a curl that connects then
    hangs (which --retry does not cover), or any wedged process β€” runs to GitHub's 6-hour
    default before being killed and counted as a timeout failure, inflating the failure rate and
    burning runner minutes. p95 for this workflow is ~150s, so timeout-minutes: 10 is ample
    headroom and turns a hang into a fast, obvious failure. This mirrors the convention already
    used by sonarcloud.yml, copilot-setup-steps.yml, initiative-driver.yml and
    feature-ideation.yml.

    Test Results

    === test-ci-workflow ===
    PASS: 'yq' is available
    PASS: .github/workflows/ci.yml exists
    PASS: 'concurrency.group' block present in .github/workflows/ci.yml
    PASS: concurrency group keyed on github.ref in .github/workflows/ci.yml
    PASS: 'concurrency.cancel-in-progress' is unconditionally true in .github/workflows/ci.yml
    PASS: 'Install yq' extracts the SHA-256 (column 19) in .github/workflows/ci.yml
    PASS: every 'curl' download retries on transient failures in .github/workflows/ci.yml
    PASS: job 'secret-scan' declares timeout-minutes (10) in .github/workflows/ci.yml
    PASS: job 'workflow-tests' declares timeout-minutes (10) in .github/workflows/ci.yml
    
    All checks passed.
    
    --- test-ci-workflow.test.sh ---
    ok   - buggy yq extraction is rejected
    ok   - correct yq extraction (^yq_linux_amd64 / column 19) is accepted
    ok   - workflow without an Install yq step passes (check skipped)
    ok   - workflow missing a concurrency block is rejected
    ok   - decoy $19 in unnamed step does not rescue buggy Install yq
    ok   - correct Install yq with following unnamed step is accepted
    ok   - curl download without --retry is rejected
    ok   - curl downloads with --retry (multi-line continuation) are accepted
    ok   - curl without --retry chained with one that has --retry is rejected
    ok   - every job declaring timeout-minutes is accepted
    ok   - a job missing timeout-minutes is rejected
    ok   - timeout-minutes: 0 is rejected
    
    All tests passed.
    

    Full workflow-tests job suite (all 7 scripts/test-*-workflow*.sh) also passes.

    Files Changed

    • .github/workflows/ci.yml: added timeout-minutes: 10 to both jobs with explanatory comments
    • scripts/test-ci-workflow.sh: added Check 7 (every job declares a positive-integer timeout-minutes); refreshed header docstring
    • scripts/test-ci-workflow.test.sh: added fixtures + Cases 10–12 (accept when present, reject when missing, reject timeout-minutes: 0); backfilled timeout-minutes into existing fixtures so each still isolates its intended defect

    Notes

    The observed 2 failures are a per-branch coverage issue for that PR's author to resolve, not a
    ci.yml defect. Edge cases covered by the guard: missing value, non-integer/expression values,
    and zero are all rejected; hyphenated job names are handled via yq env() lookup. No git
    operations performed β€” left to the CI workflow.

  4. added a commit that references this issue on Sep 1, 2026
  5. added 4 commits that reference this issue on Sep 1, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    dev-leadFor dev-lead agent pickupfleet-trackerTracked by the Actions Fleet Monitorhealth-checkAutomated health check result

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions