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
Window evaluated: last 24 full hours (UTC), 2026-09-09T23:32:53Z → 2026-09-10T23:03:48Z
Total runs analyzed: 432 (across 164 distinct workflows)
Detection-enabled runs: 431 (99.8%)
Regular runs (gh-aw-detection: false): 1 (0.2%)
Misconfigured workflows found: 0
Note
No gh-aw-detection-related misconfigurations were found in this window. The one workflow with detection explicitly disabled ("Daily GitHub Docs SEO Optimizer") only ran once in the last 24h, so it does not cross the >3-runs/7-days threshold used to flag a stale opt-out.
Comparison Chart
Sample sizes are heavily imbalanced (1 regular run vs 431 detection-enabled runs), so the success-rate and token comparisons are not statistically meaningful this cycle — they're included for continuity with the trend chart below.
Misconfigured Workflows
No misconfigured workflows detected in this window.
View All Run Metrics
Aggregate:
Metric
Regular
Detection-enabled
Run count
1
431
Success rate
100.0%
76.3%
Failure count
0
102
Avg tokens
83,750
105,099
Top workflows by run count (25 of 164):
Workflow
Runs
Status
Success rate
Avg tokens
Engine(s)
PR Sous Chef
80
Detection
98.8%
112,961
pi
Issue Monster
45
Detection
0.0%
0
pi
AI Moderator
19
Detection
0.0%
0
pi
Test Quality Sentinel
15
Detection
100.0%
8,265
copilot
Impeccable Skills Reviewer
14
Detection
85.7%
32,075
copilot
Matt Pocock Skills Reviewer
14
Detection
64.3%
48,492
copilot
PR Code Quality Reviewer
14
Detection
64.3%
837,894
copilot
Design Decision Gate 🏗️
14
Detection
100.0%
165,202
pi
Ponytail Reviewer
13
Detection
100.0%
80,132
codex
Daily Go Test Parallelizer
12
Detection
100.0%
136,182
codex
Auto-Triage Issues
10
Detection
100.0%
314,348
codex
Avenger
9
Detection
77.8%
293,868
codex
Contribution Check
6
Detection
100.0%
18,612
copilot
[aw] Failure Investigator (6h)
5
Detection
100.0%
8,077
claude
PR Triage Agent
4
Detection
75.0%
42,561
copilot
Deep Report
4
Detection
100.0%
70,310
claude
Code Scanning Fixer
4
Detection
0.0%
18,018
copilot
Daily Credit Limit Test
2
Detection
100.0%
0
pi
Release
2
Detection
100.0%
12,082
copilot
Bot Detection
2
Detection
100.0%
522,790
codex
Daily Documentation Healer
1
Detection
100.0%
27,028
claude
Daily Observability Report for AWF Firewall and MCP Gateway
1
Detection
100.0%
250,025
codex
Daily Model Inventory Checker
1
Detection
100.0%
75,745
copilot
Outcome Collector
1
Detection
100.0%
181,578
codex
Documentation Unbloat
1
Detection
100.0%
293,006
pi
Detection-job health: of the 406 runs where a detection job actually executed, 402 completed successfully and 4 were skipped — none failed, so no rule-2 misconfigurations were flagged.
View Historical Trend
9 days of history now recorded in cache-memory (trending/detection-metrics/history.jsonl). Regular-run volume has stayed near zero throughout; detection-enabled run volume has grown from ~250/day to ~430/day, while detection success rate has drifted down from ~92% to ~76% over the same period.
Recommendations
No detection-configuration action needed this cycle — the gh-aw-detection feature is enabled by default across essentially the entire fleet (99.8% of runs), and no detection job failures or inconsistent per-workflow settings were observed.
Investigate two unrelated but notable failure clusters (out of scope for this detection audit, but worth a look): "Issue Monster" (45/45 runs failed, pi engine, 0 tokens consumed) and "AI Moderator" (19/19 runs failed, pi engine, 0 tokens consumed). Both fail before any tokens are used, suggesting a startup/config error rather than an agent-logic failure.
Watch the declining detection success-rate trend (92% → 76% over 9 days) — while not a gh-aw-detection misconfiguration per se, it's worth correlating against the recent token-usage growth to rule out timeouts or rate limiting as the agent workload scales.
Warning
Firewall blocked 1 domain
The following domain was blocked by the firewall during workflow execution:
api.anthropic.com
To allow these domains, add them to the network.allowed list in your workflow frontmatter:
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
gh-aw-detection: false): 1 (0.2%)Note
No
gh-aw-detection-related misconfigurations were found in this window. The one workflow with detection explicitly disabled ("Daily GitHub Docs SEO Optimizer") only ran once in the last 24h, so it does not cross the >3-runs/7-days threshold used to flag a stale opt-out.Comparison Chart
Sample sizes are heavily imbalanced (1 regular run vs 431 detection-enabled runs), so the success-rate and token comparisons are not statistically meaningful this cycle — they're included for continuity with the trend chart below.
Misconfigured Workflows
No misconfigured workflows detected in this window.
View All Run Metrics
Aggregate:
Top workflows by run count (25 of 164):
Detection-job health: of the 406 runs where a
detectionjob actually executed, 402 completed successfully and 4 were skipped — none failed, so no rule-2 misconfigurations were flagged.View Historical Trend
9 days of history now recorded in cache-memory (
trending/detection-metrics/history.jsonl). Regular-run volume has stayed near zero throughout; detection-enabled run volume has grown from ~250/day to ~430/day, while detection success rate has drifted down from ~92% to ~76% over the same period.Recommendations
gh-aw-detectionfeature is enabled by default across essentially the entire fleet (99.8% of runs), and no detection job failures or inconsistent per-workflow settings were observed.piengine, 0 tokens consumed) and "AI Moderator" (19/19 runs failed,piengine, 0 tokens consumed). Both fail before any tokens are used, suggesting a startup/config error rather than an agent-logic failure.gh-aw-detectionmisconfiguration per se, it's worth correlating against the recent token-usage growth to rule out timeouts or rate limiting as the agent workload scales.Warning
Firewall blocked 1 domain
The following domain was blocked by the firewall during workflow execution:
api.anthropic.comTo allow these domains, add them to the
network.allowedlist in your workflow frontmatter:See Network Configuration for more information.
All reactions