[detection-analysis] Detection Analysis Report — 2026-08-11

[github-actions[bot]]

Summary

• Window evaluated: last 24 full hours (UTC), ending ~2026-08-11T23:24:00Z
• Total runs analyzed with usable log data: 11 (of 44 run directories discovered in the pre-downloaded log set — see data-completeness note below)
• Detection-enabled runs: 10 (90.9%)
• Regular runs: 1 (9.1%)
• Misconfigured workflows found: 8

Warning

8 scheduled workflows whose names match audit/report/detector/analyzer keywords do not have gh-aw-detection: true in their frontmatter.

Comparison Chart

• Regular runs: 1, success rate 100.0%, avg tokens 15,696
• Detection runs: 10, success rate 90.0%, avg tokens 17,079

Misconfigured Workflows

These are flagged under Rule 2 (name contains audit/analyzer/report/detector/monitor/inspector but lacks gh-aw-detection: true), verified directly against each workflow's frontmatter in .github/workflows/. All eight run on a daily or weekday schedule trigger, so they plausibly executed within the analyzed 24h window even though none of their runs happened to land in this run's log sample (see caveat below — example_run_id is therefore not available for these).

Workflow	Misconfiguration Type	Schedule	Recommended Fix
daily-action-setup-security-audit	Rule 2: name match, no gh-aw-detection: true	daily	Add gh-aw-detection: true to frontmatter
daily-firewall-report	Rule 2: name match, no gh-aw-detection: true	daily	Add gh-aw-detection: true to frontmatter
daily-regression-audit-kiro	Rule 2: name match, no gh-aw-detection: true	daily	Add gh-aw-detection: true to frontmatter
daily-schema-audit-cursor	Rule 2: name match, no gh-aw-detection: true	daily	Add gh-aw-detection: true to frontmatter
daily-sentrux-report	Rule 2: name match, no gh-aw-detection: true	daily	Add gh-aw-detection: true to frontmatter
designer-drift-audit	Rule 2: name match, no gh-aw-detection: true	daily (weekdays)	Add gh-aw-detection: true to frontmatter
go-pattern-detector	Rule 2: name match, no gh-aw-detection: true	daily (weekdays)	Add gh-aw-detection: true to frontmatter
static-analysis-report	Rule 2: name match, no gh-aw-detection: true	daily	Add gh-aw-detection: true to frontmatter

Not flagged: agentic-token-audit matches the name pattern and also lacks gh-aw-detection: true, but it is a documented, source-managed mirror of githubnext/agentic-ops and is excluded per the standing repository opt-out.

Excluded from this pass (on-demand only, no schedule, no evidence of running in-window): objective-impact-report, repo-audit-analyzer, video-analyzer (all workflow_dispatch-only). Excluded (weekly cadence, unlikely to fall in this specific 24h window): github-mcp-tools-report (Sunday), mcp-inspector (Monday), org-health-report (Monday), ubuntu-image-analyzer (weekly, day unspecified).

No violations of Rule 1 (>3 runs/7 days with detection off — insufficient history in a 24h window to assess), Rule 3 (detection steps failing on a detection-enabled run — all 10 detection-enabled runs had empty-but-present usage/detection/token_usage.jsonl, consistent with normal no-op detection rather than failure), or Rule 4 (alternating detection state within 24h — no workflow in the observed sample ran more than once with different gh-aw-detection values) were found in the 11-run log sample.

View All Run Metrics

Workflow	Run	Engine	Detection	Conclusion	Tokens
PR Code Quality Reviewer	§31543322294	pi	Yes	success	0
Design Decision Gate 🏗️	§31543322301	claude	Yes	failure	13,983
Ponytail Reviewer	§31543322357	copilot	Yes	success	5,987
Matt Pocock Skills Reviewer	§31543322378	copilot	Yes	success	18,162
Impeccable Skills Reviewer	§31543322462	copilot	Yes	success	9,294
Test Quality Sentinel	§31543322465	copilot	Yes	success	7,547
PR Sous Chef	§31543567557	pi	Yes	success	48,021
Auto-Triage Issues	§31544565977	pi	Yes	success	12,058
Squad Plan	§31544608280	copilot	No	success	15,696
Daily Reliability Review	§31545174905	opencode	Yes	success	11,070
PR Sous Chef	§31545290203	pi	Yes	success	44,665

Design Decision Gate was the only failure in the sample. ErrorCount/MissingToolCount/MissingDataCount were all 0 and the run produced one push_to_pull_request_branch safe-output, consistent with the gate intentionally rejecting the PR rather than a technical or detection-related failure.

Data-completeness caveat: the pre-downloaded log set at /tmp/gh-aw/aw-mcp/logs/ contained 44 run directories, but only 11 had a populated aw_info.json/run_summary.json. The other 33 (including this report's own in-flight run, 31546135660) were empty — most likely runs still in progress at download time or whose artifacts had not yet been generated/retained. Metrics above reflect only the 11 runs with usable data; they should be read as a sample of the window, not an exhaustive census.

View Historical Trend

No trend chart is available yet. This is the first recorded data point in /tmp/gh-aw/cache-memory/trending/detection/history.jsonl — a trend chart will be generated automatically once at least one more day of history has accumulated.

Recommendations

1. Add gh-aw-detection: true to the 8 flagged daily/weekday-scheduled workflows listed above so their AI-credit usage is captured by the detection pipeline like their peers.
2. Re-run this analysis once artifact retention/timing allows a fuller log sample — 75% of discovered run directories in this window had no downloadable data, which limits confidence in Rule 1 (frequency-based) and Rule 4 (alternation) checks.
3. No action needed on Design Decision Gate's failure — it appears to be an intentional gate rejection, not a detection or infrastructure fault.

References:

• §31543322301 — Design Decision Gate (only failure in sample)
• §31543567557 — PR Sous Chef (highest token usage, 48,021)
• §31544608280 — Squad Plan (only regular/non-detection run in sample)

Generated by 🔎 Detection Analysis Report · agent · 165.6 AIC · ⌖ 49.2 AIC · ⊞ 6.4K · ◷

• expires on Aug 14, 2026, 3:43 PM UTC-08:00

[github-actions[bot]]

This discussion has been marked as outdated by Detection Analysis Report.

A newer discussion is available at Discussion #52384.