You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
[detection-analysis] Detection Analysis Report — Last 24h (2026-11-10)
#76630
Window evaluated: last 24 full hours (UTC) — 2026-11-09T23:11:25Z to 2026-11-10T23:11:25Z
Total runs observed in window: 1,045 across 161 distinct workflows
Fully classified (detection-feature data available): 547 runs (52.3% of observed runs) — the remainder (498 runs) only had shallow GitHub Actions metadata downloaded before the 1 GB log-collection storage budget was reached, so their gh-aw-detection status could not be determined. All percentages below are computed over the 547 fully-classified runs.
Detection-enabled runs: 542 (99.1%)
Regular runs (gh-aw-detection: false): 5 (0.9%)
Misconfigured workflows found: 3
Warning
3 workflows had a genuinely failing detection job (not merely a skipped/absent job) in this window. See details below.
Comparison Chart
Metric
Regular Runs
Detection Runs
Run count
5
542
Success rate
100.0%
77.7%
Failure count
0
121
Avg token usage
30,284
170,374
Note: the "Regular" group (5 runs, all one-off conformance/smoke tests) is too small to draw a reliable success-rate comparison against the 542 detection-enabled runs; the 100% regular success rate reflects low sample size, not a systematic quality difference.
Misconfigured Workflows
All 3 flagged workflows share the same root cause: the detection job's "Install AWF binary" step failed, cascading to skip the remaining detection steps. In each case the agent job itself completed successfully — this is a detection-job-specific failure, not an agent/trigger issue.
Workflow
Misconfiguration Type
Run Count
Recommended Fix
Daily Documentation Updater
detection job failing (Install AWF binary step)
1
Investigate why AWF binary install fails in the detection job for this workflow; likely a shared infra/version-pinning issue — see §42721686333
No workflow exceeded the "gh-aw-detection: false with >3 runs" threshold (rule 1), and no workflow alternated between detection-enabled and detection-disabled within the window (rule 3). Caveat: rule 1 is specified against a 7-day run history; only the 24h window was collected (log-collection was scoped to the last 24 hours per the audit budget), so this is a bounded proxy, not a full 7-day check.
View All Run Metrics (161 workflows, full-detail runs only)
Based on 42 daily snapshots (90-day retention applied, none dropped this run). Detection-enabled run volume and success rate have been broadly stable day over day; today's detection success rate (77.7%) is on the lower end of the recent range, driven mainly by Auditor (0% success, 22 runs — a collect_failure_evidence step failure unrelated to the detection feature itself) and Daily Go Test Parallelizer (0% success, 10 runs).
Recommendations
Fix the shared Install AWF binary failure in the detection job for Daily Documentation Updater, Daily Storify, and Daily Documentation Healer — same failure signature across all three, so a single infra/version-pinning fix likely resolves all three. Start from §42721686333.
No gh-aw-detection opt-outs need action — the 5 regular runs observed are all single-run conformance/smoke-test workflows (Daily GitHub Docs SEO Optimizer, Engine Conformance Copilot/Pi/OpenCode, Smoke Copilot Sub Agents), none exceeding the >3-runs misconfiguration threshold.
Re-run this audit with a longer collection budget or incremental caching to classify the 498 runs that only received shallow metadata in this window (52.3% coverage achieved); this would also allow a true 7-day check for rule 1 instead of the 24h proxy used here.
Separately from detection-feature scope: Auditor and Daily Go Test Parallelizer both show 0% success across double-digit run counts in this window — worth a dedicated reliability investigation, though their failures are in the agent/collect_failure_evidence jobs, not detection.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
Summary
2026-11-09T23:11:25Zto2026-11-10T23:11:25Zgh-aw-detectionstatus could not be determined. All percentages below are computed over the 547 fully-classified runs.gh-aw-detection: false): 5 (0.9%)Warning
3 workflows had a genuinely failing
detectionjob (not merely a skipped/absent job) in this window. See details below.Comparison Chart
Note: the "Regular" group (5 runs, all one-off conformance/smoke tests) is too small to draw a reliable success-rate comparison against the 542 detection-enabled runs; the 100% regular success rate reflects low sample size, not a systematic quality difference.
Misconfigured Workflows
All 3 flagged workflows share the same root cause: the
detectionjob's "Install AWF binary" step failed, cascading to skip the remaining detection steps. In each case theagentjob itself completed successfully — this is a detection-job-specific failure, not an agent/trigger issue.Install AWF binarystep)Install AWF binarystep)Install AWF binarystep)No workflow exceeded the "
gh-aw-detection: falsewith >3 runs" threshold (rule 1), and no workflow alternated between detection-enabled and detection-disabled within the window (rule 3). Caveat: rule 1 is specified against a 7-day run history; only the 24h window was collected (log-collection was scoped to the last 24 hours per the audit budget), so this is a bounded proxy, not a full 7-day check.View All Run Metrics (161 workflows, full-detail runs only)
View Historical Trend
Based on 42 daily snapshots (90-day retention applied, none dropped this run). Detection-enabled run volume and success rate have been broadly stable day over day; today's detection success rate (77.7%) is on the lower end of the recent range, driven mainly by
Auditor(0% success, 22 runs — acollect_failure_evidencestep failure unrelated to the detection feature itself) andDaily Go Test Parallelizer(0% success, 10 runs).Recommendations
Install AWF binaryfailure in thedetectionjob forDaily Documentation Updater,Daily Storify, andDaily Documentation Healer— same failure signature across all three, so a single infra/version-pinning fix likely resolves all three. Start from §42721686333.gh-aw-detectionopt-outs need action — the 5 regular runs observed are all single-run conformance/smoke-test workflows (Daily GitHub Docs SEO Optimizer,Engine Conformance Copilot/Pi/OpenCode,Smoke Copilot Sub Agents), none exceeding the >3-runs misconfiguration threshold.AuditorandDaily Go Test Parallelizerboth show 0% success across double-digit run counts in this window — worth a dedicated reliability investigation, though their failures are in theagent/collect_failure_evidencejobs, notdetection.References:
All reactions