[detection-analysis] Detection Analysis Report — 2026-07-30

[github-actions[bot]]

Summary

• Window evaluated: last 24 full hours (UTC), 2026-07-29T23:40Z → 2026-07-30T23:40Z
• Total runs analyzed: 220
• Detection-enabled runs: 145 (65.9%)
• Regular runs: 75 (34.1%)
• Misconfigured workflows found: 9

Warning

9 workflows show detection-configuration issues in this window — 6 with inconsistent gh-aw-detection settings across runs of the same workflow, and 3 audit/report-named workflows without detection enabled. See details below.

Comparison Chart

Regular runs averaged 61,055 tokens at a 96.0% success rate (3 failures); detection runs averaged 102,289 tokens at a 91.7% success rate (12 failures). Detection-enabled runs cost roughly 1.7x the tokens of regular runs on average, and run at a slightly lower success rate — consistent with the added threat-detection step surfacing more agent-side errors rather than being a regression itself.

Misconfigured Workflows

Workflow	Misconfiguration Type	Run Count	Recommended Fix
Matt Pocock Skills Reviewer	inconsistent gh-aw-detection setting across runs within 24h window	13	Pin gh-aw-detection to a single value in frontmatter; inconsistent settings suggest a compiled/.lock.yml drift or partial rollout.
Design Decision Gate 🏗️	inconsistent gh-aw-detection setting across runs within 24h window	10	Pin gh-aw-detection to a single value in frontmatter; inconsistent settings suggest a compiled/.lock.yml drift or partial rollout.
Impeccable Skills Reviewer	inconsistent gh-aw-detection setting across runs within 24h window	10	Pin gh-aw-detection to a single value in frontmatter; inconsistent settings suggest a compiled/.lock.yml drift or partial rollout.
PR Code Quality Reviewer	inconsistent gh-aw-detection setting across runs within 24h window	11	Pin gh-aw-detection to a single value in frontmatter; inconsistent settings suggest a compiled/.lock.yml drift or partial rollout.
PR Description Updater	inconsistent gh-aw-detection setting across runs within 24h window	7	Pin gh-aw-detection to a single value in frontmatter; inconsistent settings suggest a compiled/.lock.yml drift or partial rollout.
PR Sous Chef	inconsistent gh-aw-detection setting across runs within 24h window	16	Pin gh-aw-detection to a single value in frontmatter; inconsistent settings suggest a compiled/.lock.yml drift or partial rollout.
Deep Report	name suggests audit/monitoring role but gh-aw-detection is not consistently enabled	1	Add gh-aw-detection: true to the workflow frontmatter, or add a documented repository-level opt-out if intentional.
Daily Security Observability Report	name suggests audit/monitoring role but gh-aw-detection is not consistently enabled	1	Add gh-aw-detection: true to the workflow frontmatter, or add a documented repository-level opt-out if intentional.
Daily Regulatory Report Generator	name suggests audit/monitoring role but gh-aw-detection is not consistently enabled	1	Add gh-aw-detection: true to the workflow frontmatter, or add a documented repository-level opt-out if intentional.

Note: rule 1 (gh-aw-detection: false on a workflow with >3 runs in the last 7 days) could only be evaluated against the 24h log window fetched by this run — any workflow with >3 runs in 24h also has >3 in 7d, so no false positives are possible under this proxy, but a workflow averaging ≤3 runs/day that still exceeds 3 runs over 7 days would not be caught here. No such cases were found among detection-disabled workflows in this window. Rule 3 (detection job failures) found no failed detection jobs in this window.

View All Run Metrics

Workflows with more than one run in the window (16 of 116 total workflows seen; the remaining 100 each ran once):

Workflow	Runs	Detection (any run)	Success Rate	Avg Tokens	Engine(s)
PR Sous Chef	16	Yes	87.5%	47,050	pi
Matt Pocock Skills Reviewer	13	Yes	100.0%	15,232	copilot
Issue Monster	12	Yes	100.0%	29,898	pi
PR Code Quality Reviewer	11	Yes	81.8%	31,538	copilot
Design Decision Gate 🏗️	10	Yes	90.0%	12,359	claude
Test Quality Sentinel	10	Yes	100.0%	15,233	copilot
Impeccable Skills Reviewer	10	Yes	100.0%	10,367	copilot
AI Moderator	8	Yes	100.0%	120,478	codex
PR Description Updater	7	Yes	100.0%	12,620	copilot
Auto-Triage Issues	5	Yes	100.0%	15,051	pi
Contribution Check	4	Yes	100.0%	23,303	copilot
PR Triage Agent	3	No	100.0%	25,677	copilot
Discussion Task Miner - Code Quality Improvement Agent	3	No	100.0%	23,815	copilot
[aw] Failure Investigator (6h)	3	Yes	100.0%	34,728	claude
Code Scanning Fixer	3	Yes	100.0%	8,390	copilot
Avenger	2	Yes	100.0%	11,472	claude

View Historical Trend

28 days of history available in cache-memory (trending/detection-metrics/history.jsonl).

Recommendations

1. Pin gh-aw-detection consistently for PR Sous Chef, Matt Pocock Skills Reviewer, Design Decision Gate, Impeccable Skills Reviewer, PR Code Quality Reviewer, and PR Description Updater — each alternated between enabled/disabled across runs in the same 24h window, which points to a .lock.yml recompilation drift (frontmatter edited mid-day) rather than a deliberate rollout. Recompile once from the current source and verify all subsequent runs pick up the same flag consistently.
2. Add gh-aw-detection: true to Deep Report, Daily Security Observability Report, and Daily Regulatory Report Generator, or add an explicit documented opt-out (as already exists for Daily Agentic Workflow AIC Usage Audit) if detection is intentionally not applicable.
3. No detection-job failures or 7-day high-frequency detection gaps were observed this window — re-run this audit after wider historical scrape (7d) to fully validate rule 1 coverage for lower-frequency workflows.

Generated by 🔎 Detection Analysis Report · agent · 144 AIC · ⌖ 25 AIC · ⊞ 7.2K · ◷

• expires on Aug 2, 2026, 3:46 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 #49680.