[daily regulatory] Regulatory Report - 2026-02-19

[github-actions[bot]]

This regulatory report reviews 9 daily reports generated on 2026-02-19 for the github/gh-aw repository. Overall data quality is good, with consistent metrics across reports that share the same scope. Two critical issues require attention: a 0% Copilot agent session success rate (all 4 sessions failed on 2026-02-18) and a persistent 0% discussion answer rate (0 of 100 discussions answered). Workflow infrastructure health is strong with Safe Outputs at 94.1% success and no firewall policy violations detected.

Cross-report consistency is high for key metrics. Token consumption is down 49.8% from the previous tracked period, consistent with a lighter scheduling day. The Daily Status report lacked issue data entirely, which represents a data gap. Code quality scores are improving (82/100) but code churn stability remains low (4/15) due to active WASM Playground development.

📋 Full Regulatory Report

📊 Reports Reviewed

#	Report	Discussion	Created	Status
1	Daily Performance Summary	#16902	2026-02-19T19:46Z	✅ Valid
2	Daily Team Evolution Insights	#16865	2026-02-19T16:15Z	✅ Valid
3	Daily Code Metrics Report	#16800	2026-02-19T11:49Z	✅ Valid
4	Daily Copilot Token Consumption	#16797	2026-02-19T11:32Z	✅ Valid
5	Daily Status	#16789	2026-02-19T09:31Z	⚠️ Missing issue data
6	Daily Firewall Report	#16698	2026-02-19T01:03Z	✅ Valid
7	Daily Compiler Code Quality	#16686	2026-02-19T00:25Z	⚠️ One file below threshold
8	Copilot Session Insights (Feb 18)	#16684	2026-02-19T00:22Z	❌ Critical failures
9	Safe Output Health	#16747	2026-02-19T04:27Z	✅ Valid

---

🔍 Data Consistency Analysis

Cross-Report Metrics Comparison

Metric	Performance	Status	Team Evo	Scope	Status
Open Issues	41	❌ No data	—	Same	⚠️ Gap in Status
Closed Issues	959 (90d)	—	—	90d	✅ Single source
Total PRs	100 (90d)	86 opened (30d)	—	Different	ℹ️ Diff scope
Merged PRs	76 (90d)	50 (7d)	—	Different	✅ Consistent trend
Open PRs	3	6 open	—	Same	⚠️ Minor variance (3 vs 6)
Contributors	7 (PR authors, 90d)	18 (commit, 7d)	5 (today)	Different	✅ Expected
Discussions answered	0/100 (0%)	—	—	90d	🚨 Persistent gap
Workflow runs	11 (firewall, daily)	—	—	Scoped	✅ Different scopes expected

Scope Notes:

• open_prs: Performance reports "3 open" (of the 100 sampled), Status reports "6 open". Small variance may reflect sampling vs full-state difference.
• contributors: Performance uses PR authors over 90d; Status uses commit authors over 7d; Team Evolution counts today's committers — all expected to differ.
• issues_analyzed: Performance caps at 1,000 total issues; Status reported 0 issue nodes (data gap).

Consistency Score

• Overall Consistency: 83% (5 of 6 directly comparable metrics within tolerance)
• Critical Discrepancies: 0 (open PRs variance is minor)
• Data Gaps: 1 (Daily Status missing issue data)

---

⚠️ Issues and Anomalies

🚨 Critical Issues

1. Copilot Agent Sessions — 0% Success Rate

• Affected Report: Copilot Session Insights #16684
• Metric: Copilot session success rate
• Description: All 4 Copilot agent sessions on 2026-02-18 failed. Named runs (addressing PR comments on feat: Add blocked pattern matching for add-labels and remove-labels safe outputs #16629) failed across 3 retries, with retry duration escalating (5.5m → 10.3m → 6.5m) — no convergence.
• Root Cause: Lock workflow systematic failures (*.lock.yml workflows fail 100% on Copilot branches) combined with reviewer agents consistently returning action_required.
• Severity: Critical — Copilot SWE agent cannot land code changes
• Recommended Action: Investigate why .lock.yml workflows fail on Copilot branches; review the reviewer-agent thresholds that are blocking all Copilot PRs.

2. Discussion Answer Rate — Persistent 0%

• Affected Reports: Performance Summary #16902
• Metric: discussions_answered / total_discussions = 0/100 (0%)
• Description: Zero discussions have accepted answers across all 100 sampled discussions. This has been flagged in previous regulatory reports and remains unresolved.
• Severity: High — indicates community engagement backlog or missing answer attribution workflow
• Recommended Action: Establish a discussion-answering protocol; assign owners to triage unanswered threads weekly.

⚠️ Warnings

1. Daily Status Report — Issue Data Empty

• Details: Daily Status #16789 explicitly notes "Issue data was empty (0 nodes) — issues may not be enabled or accessible." This prevents cross-validation of open issue count.
• Impact: Cannot confirm the 41 open issues reported by Performance Summary from an independent source.

2. compiler_yaml.go — Below Quality Threshold

• Details: Compiler Code Quality #16686 scores compiler_yaml.go at 71/100 — below the stated 75/100 human-written quality threshold. Drivers: 2 functions >150 lines, 6 missing doc comments.
• Impact: Technical debt in a core compilation file.

3. Code Churn Stability — 4/15

• Details: Code Metrics #16800 reports churn stability at 4/15, driven by 1,075 files modified with 103K lines added and 42K deleted in 7 days.
• Impact: High churn increases regression risk; attributed to WASM Playground development sprint.

4. Firewall Block Rate — 43.4%

• Details: Firewall #16698 reports 552 of 1,272 requests blocked (43.4%). All blocked traffic recorded under - placeholder (pre-DNS/internal kernel attempts — not explicit deny-list hits).
• Impact: Informational only; no policy violations detected. However, the high rate of pre-DNS blocks may warrant investigation into whether workflows are attempting to reach disallowed services.

---

📈 Trend Analysis

Week-over-Week Comparison (Available Data)

Metric	2026-02-19	2026-02-18	Change
Code quality score	82/100	81/100	+1 ✅
Token consumption (30d)	119.3M	~237.8M	-49.8% ✅ lighter day
Safe output success rate	94.1%	—	—
Copilot session success	0%	—	🚨
Workflow runs (daily)	29 (safe-outputs)	—	—

Notable Trends

• WASM Playground is a major active sprint (5 related commits, 4 releases on Feb 19), explaining elevated churn metrics.
• AI-assisted development is dominant: Copilot accounts for ~51% of commits (248/489 this week).
• Token consumption shows day-to-day variability; 83-run days are lighter than 378-run peaks.

---

📝 Per-Report Analysis

Performance Summary (#16902)

Time Period: Last 90 days | Quality: ✅ Valid

Metric	Value
Total Issues (sampled)	1,000
Open Issues	41
Closed Issues	959
Total PRs (sampled)	100
Merged PRs	76 (76%)
Open PRs	3
Avg Merge Time	1.0 hours
Unique PR Contributors	7
Total Discussions (sampled)	100
Answered Discussions	0 (0%)
Avg Issue Resolution Time	9.3 hours

Notes: Issue and PR data is capped at 1,000 and 100 items respectively due to API limits. Discussion answer rate of 0% is the key concern.

Daily Firewall Report (#16698)

Time Period: Last 7 days | Quality: ✅ Valid

Metric	Value
Workflow runs analyzed	11 (+1 failed)
Total network requests	1,272
Allowed requests	720 (56.6%)
Blocked requests	552 (43.4%)
Unique blocked domains	1 (- placeholder)
Dominant engine	Copilot (8 runs)
Highest block rate workflow	Daily Compiler Quality Check (61.2%)

Notes: No explicit deny-listed domains triggered. All blocks are pre-DNS internal attempts.

Copilot Token Consumption (#16797)

Time Period: Last 30 days | Quality: ✅ Valid

Metric	Value
Total tokens consumed	119,326,605
Workflow runs	83
Unique workflows	42
Avg tokens/run	1,437,669
Errors detected	4
Top consumer	CI Failure Doctor (36.1M, 9 runs)
vs. previous period	-49.8% (237.8M → 119.3M)

Safe Output Health (#16747)

Time Period: Last 24 hours | Quality: ✅ Valid

Metric	Value
Workflow runs analyzed	29
Distinct workflows	21
Safe output jobs executed	17
Safe output jobs skipped (upstream failures)	3
Safe output jobs with errors	1
Total tool calls	31
Overall success rate	94.1% (16/17)

Error: Duplicate Code Detector — "Assign Copilot to created issues" step failed (FORBIDDEN: target repository not writable).

Daily Code Metrics (#16800)

Time Period: Point-in-time snapshot | Quality: ✅ Valid

Metric	Value
Total LOC	813,810
Quality score	82/100
Test-to-source ratio	2.33x
Churn stability	4/15
7d files modified	1,075
7d lines added	103K
7d lines deleted	42K

---

💡 Recommendations

Process Improvements

1. Fix Copilot lock-workflow failures: Investigate why all .lock.yml workflows fail on Copilot branches — this blocks 100% of agent-driven PRs from landing.
2. Establish discussion answer attribution: Define ownership for marking discussions as answered; create a weekly triage workflow for unanswered threads.
3. Fix Daily Status issue data gap: Diagnose why issues.json returns 0 nodes for the Status workflow — may be a permissions or query scope issue.

Data Quality Actions

1. Align open PR counts: Performance Summary (3 open) vs Status report (6 open) — ensure both queries use the same state/filter criteria.
2. Add historical baselines: Token Consumption and Code Metrics already persist history. Extend to Firewall and Safe Output Health for trend detection.

Workflow Suggestions

1. Reviewer-agent calibration: All reviewer agents return action_required on Copilot PRs (100% rate). Review thresholds and consider a "pass with suggestions" mode to unblock Copilot iterations.
2. Address compiler_yaml.go quality debt: Schedule a refactoring session to bring the file above 75/100 by extracting oversized functions and adding doc comments.

---

📊 Regulatory Metrics

Metric	Value
Reports Reviewed	9
Reports Passed	6
Reports with Warnings	2
Reports with Critical Issues	1
Overall Health Score	78%
Previous Regulatory Reports Closed	2 (#16636, #16432)

---

References:

• §22198039087

AI generated by Daily Regulatory Report Generator

• expires on Feb 22, 2026, 8:10 PM UTC

[github-actions[bot]]

This report has been superseded by a newer daily regulatory report.