[detection-analysis] Detection Analysis Report — Last 24h (2026-07-09)

[github-actions[bot]]

Summary

• Window evaluated: last 24 full hours (UTC) — 2026-07-08T13:29:12Z → 2026-07-09T13:29:12Z
• Total runs analyzed: 64
• Detection-enabled runs: 28 (43.8%)
• Regular runs: 36 (56.2%)
• Misconfigured workflows found: 3

Warning

3 workflow(s) flagged as misconfigured for the detection feature in the last 24h.

Comparison Chart

Metric	Regular Runs	Detection Runs
Total runs	36	28
Success rate	44.4%	60.7%
Avg tokens/run	1093K	1540K
Failure count	20	11
Misconfigured	—	3

Misconfigured Workflows

Workflow	Misconfiguration type	Runs	Recommended fix
Bot Detection	2: Detection absent on audit/analysis/report workflow	1 (ex §29002110299)	Add gh-aw-detection: true — the name matches an audit/analysis/report/detector pattern that should run threat detection.
Copilot Agent Prompt Clustering Analysis	2: Detection absent on audit/analysis/report workflow	1 (ex §29012331375)	Add gh-aw-detection: true — the name matches an audit/analysis/report/detector pattern that should run threat detection.
Duplicate Code Detector	2: Detection absent on audit/analysis/report workflow	1 (ex §28998745648)	Add gh-aw-detection: true — the name matches an audit/analysis/report/detector pattern that should run threat detection.

All three are Type 2 (detection feature absent on a workflow whose name matches an audit/analysis/report/detector pattern). No Type 1 (explicit-disable on active workflow), Type 3 (detection-job failure), or Type 4 (inconsistent state) misconfigurations were found — see notes below.

Misconfiguration-check details

• Type 1 (explicit disable on active workflow): Smoke CI sets gh-aw-detection: false (2 runs), but did not exceed the >3-runs-in-7-days activity threshold in the fetched window, so it was not flagged.
• Type 3 (detection job failed): the detection job ran 50× success, 8× cancelled, 0× failed. The 8 cancellations are downstream of the agent job failing/cancelling (the detection job depends on it), not genuine detection errors — so none were flagged.
• Type 4 (inconsistent state): no workflow alternated between enabled and disabled within the window.

View All Run Metrics

Workflow	Detection	Runs	Success	Avg tokens
AI Moderator	enabled	3	2/3 (66.7%)	49K
Auto-Triage Issues	enabled	2	2/2 (100.0%)	37K
Bot Detection	absent	1	1/1 (100.0%)	70K
CLI Version Checker	enabled	1	1/1 (100.0%)	1246K
Code Scanning Fixer	enabled	1	1/1 (100.0%)	66K
Constraint Solving — Problem of the Day	enabled	1	0/1 (0.0%)	0K
Contribution Check	enabled	2	2/2 (100.0%)	886K
Copilot Agent Prompt Clustering Analysis	absent	1	0/1 (0.0%)	0K
Copilot Session Insights	enabled	1	1/1 (100.0%)	3972K
Daily AgentRx Trace Optimizer	enabled	1	1/1 (100.0%)	4621K
Daily Cli Tools Tester	enabled	1	1/1 (100.0%)	2073K
Daily Elixir Credo Snippet Audit	enabled	1	0/1 (0.0%)	0K
Daily Hippo Learn	enabled	1	1/1 (100.0%)	648K
Daily MCP Tool Concurrency Analysis	enabled	1	0/1 (0.0%)	381K
Daily Max Ai Credits Test	enabled	1	0/1 (0.0%)	0K
Daily News	enabled	1	1/1 (100.0%)	407K
Daily Rendering Scripts Verifier	enabled	1	1/1 (100.0%)	4216K
Daily Safe Outputs Conformance Checker	absent	1	1/1 (100.0%)	330K
Daily Sub-Agent Model Resolution Audit	enabled	1	0/1 (0.0%)	3587K
Daily Syntax Error Quality Check	absent	1	1/1 (100.0%)	148K
Daily Workflow Updater	absent	1	1/1 (100.0%)	877K
Design Decision Gate 🏗️	absent	2	0/2 (0.0%)	0K
Dev	absent	1	1/1 (100.0%)	1752K
Duplicate Code Detector	absent	1	1/1 (100.0%)	182K
ESLint Miner	absent	1	0/1 (0.0%)	0K
ESLint Refiner	absent	1	1/1 (100.0%)	1853K
Functional Pragmatist	absent	1	0/1 (0.0%)	409K
GitHub API Consumption Report Agent	enabled	1	0/1 (0.0%)	2601K
GitHub Remote MCP Authentication Test	enabled	1	0/1 (0.0%)	0K
Go Fan	absent	1	1/1 (100.0%)	1123K
Impeccable Skills Reviewer	absent	3	0/3 (0.0%)	1470K
Instructions Janitor	absent	1	1/1 (100.0%)	5676K
Matt Pocock Skills Reviewer	absent	2	0/2 (0.0%)	0K
Multi-Device Docs Tester	enabled	1	1/1 (100.0%)	236K
PR Code Quality Reviewer	absent	3	0/3 (0.0%)	1441K
PR Description Updater	absent	1	0/1 (0.0%)	103K
PR Sous Chef	absent	4	3/4 (75.0%)	144K
PR Triage Agent	absent	2	1/2 (50.0%)	1663K
Package Specification Enforcer	absent	1	0/1 (0.0%)	0K
Schema Consistency Checker	absent	1	1/1 (100.0%)	384K
Smoke CI	disabled	2	0/2 (0.0%)	0K
Sub-Issue Closer	absent	1	1/1 (100.0%)	894K
Team Status	absent	1	0/1 (0.0%)	0K
Terminal Stylist	absent	1	1/1 (100.0%)	3424K
Test Quality Sentinel	enabled	3	0/3 (0.0%)	0K
[aw] Failure Investigator (6h)	enabled	1	1/1 (100.0%)	2187K
daily-experiment-report	enabled	1	1/1 (100.0%)	4107K

View Historical Trend

Regular vs. detection-enabled runs per day over the last 30 days (13 days of history available), with the detection success-rate band shaded.

Recommendations

1. Enable detection on the three flagged analysis/detector workflows (Bot Detection, Copilot Agent Prompt Clustering Analysis, Duplicate Code Detector) — add gh-aw-detection: true. These are exactly the workflow archetypes (bot/duplicate detection, prompt-cluster analysis) where prompt-injection/threat detection is most valuable.
2. Review Smoke CI's explicit gh-aw-detection: false — confirm the opt-out is intentional; if it becomes an active workflow (>3 runs/week) it will be flagged.
3. Detection had a higher success rate than regular runs this window (60.7% vs 44.4%); the overall ~50% failure rate across all workflows is worth a separate reliability review, but it is not attributable to the detection feature.
4. Watch cancelled detection jobs — 8 detection jobs were cancelled because their upstream agent job failed. Reducing agent-job failures will also reduce wasted detection-job scheduling.

Warning

Firewall blocked 1 domain

The following domain was blocked by the firewall during workflow execution:

• awmgmcpg

To allow these domains, add them to the network.allowed list in your workflow frontmatter:

network:
  allowed:
    - defaults
    - "awmgmcpg"

See Network Configuration for more information.

Generated by 🔎 Detection Analysis Report · 237.8 AIC · ⌖ 22.9 AIC · ⊞ 7.1K · ◷

• expires on Jul 12, 2026, 6:00 AM UTC-08:00

[github-actions[bot]]

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

A newer discussion is available at Discussion #44734.