[daily regulatory] Regulatory Report - 2026-03-25

[github-actions[bot]]

10 daily reports reviewed for the 2026-03-25 cycle. Overall data quality is good — no critical discrepancies were found across reports. All metrics are internally consistent within expected scope differences. Safe output infrastructure scored 100% success, observability coverage reached 100%, and the Copilot agent pipeline remains highly active (76 PRs, 55 merged in 24h). One recurring warning (EP_LABELS_MISSING, 2nd consecutive day) and a notable compiler quality regression (−9 pts) are the primary items requiring attention.

No cross-report metric conflicts were identified after accounting for intentional scope differences (e.g., issues_analyzed differs by design between reports). The repository is in a healthy state with active development, strong test coverage (2.245× ratio), and proactive security posture.

📋 Full Regulatory Report

📊 Reports Reviewed

#	Report	Discussion	Created (UTC)	Status
1	Safe Output Health Report	#22903	2026-03-25T13:18Z	✅ Valid
2	Copilot Session Insights	#22881	2026-03-25T11:48Z	✅ Valid
3	Copilot Agent Analysis	#22878	2026-03-25T11:38Z	✅ Valid
4	Daily Team Evolution Insights	#22856	2026-03-25T10:55Z	✅ Valid
5	Agentic Observability Report (14d)	#22849	2026-03-25T10:21Z	✅ Valid
6	Observability Coverage Report	#22786	2026-03-24T23:52Z	⚠️ 1 warning
7	Daily Performance Summary	#22783	2026-03-24T23:28Z	✅ Valid
8	Daily Secrets Analysis	#22778	2026-03-24T22:20Z	✅ Valid
9	Daily Compiler Code Quality	#22773	2026-03-24T21:48Z	⚠️ Regression
10	Daily Code Metrics	#22769	2026-03-24T21:41Z	✅ Valid

---

🔍 Data Consistency Analysis

Cross-Report Metrics Comparison

Reference: scratchpad/metrics-glossary.md for definitions and scopes.

Metric	Source	Value	Scope	Status
open_issues	Performance Summary	99	Last 1,000 sample	✅
total_prs (90d sample)	Performance Summary	100	Last 100 PRs	✅
merged_prs (90d)	Performance Summary	73 (73%)	Last 100 PRs	✅
agent_prs_total (24h)	Copilot Agent Analysis	76	Copilot PRs, 24h	✅
agent_prs_merged (24h)	Copilot Agent Analysis	55 (72.4%)	Copilot PRs, 24h	✅
total_workflows	Code Metrics	241 (.md)	All .md workflow files	✅
workflow_files_compiled	Secrets Analysis	177 (.lock.yml)	Compiled lock files	✅ ℹ️ See Note
test_to_source_ratio	Code Metrics	2.245×	All Go test/source LOC	✅
lines_of_code_total	Code Metrics	694,040	All languages	✅
Safe output success rate	Safe Output Health	100%	2026-03-25 (10 runs)	✅
Firewall coverage	Observability Coverage	100%	Last 7d (12 runs)	✅
MCP coverage	Observability Coverage	100%	Last 7d (12 runs)	✅

Scope Notes:

• total_workflows (241 .md) vs workflow_files_compiled (177 .lock.yml): Expected mismatch — not all .md files compile to lock files at any given moment; some workflows may be templates or in draft. Not a discrepancy.
• Agent PR count (76 in 24h) vs Team Evolution PR count ("20 PRs touched"): Different scopes — Copilot Agent Analysis counts all Copilot-authored PRs; Team Evolution counts notable human+bot activities. Expected divergence per methodology.

PR Count Scope Analysis (Important Note)

Report	PRs Count	Window	Scope
Copilot Agent Analysis	76 total (55 merged)	Last 24h	All Copilot-authored/assigned PRs
Team Evolution Insights	~20 PRs "touched"	Last 24h	Editorially curated notable activity
Performance Summary	100 total (73 merged)	Last 90 days	Most recent 100 PRs, all authors

These are intentionally different scopes — not discrepancies. The Copilot Agent Analysis is comprehensive for agent-generated PRs, while Team Evolution is a narrative summary.

Consistency Score

• Overall Consistency: 95% (all metrics internally valid; scope differences documented)
• Critical Discrepancies: 0
• Minor Notes (scope-explained): 2 (workflow file count, PR count methodology)

---

⚠️ Issues and Anomalies

Warnings (Non-Critical)

1. EP_LABELS_MISSING — 2nd Consecutive Day ⚠️

• Affected Reports: Safe Output Health Report
• Metric: Safe output job success/warning
• Description: constraint-solving and problem-of-the-day labels referenced in constraint-solving-potd workflow do not exist in the repository. Warning fires on every run.
• Severity: Low — discussion creation succeeds; only label application fails silently
• Trend: Day 2 of 3 before escalation threshold
• Recommended Action: Create labels constraint-solving and problem-of-the-day via gh label create

2. compiler.go Quality Regression ⚠️

• Affected Reports: Daily Compiler Code Quality Report
• Metric: Quality score (compound)
• Description: compiler.go dropped 9 points (85→76) as validateWorkflowData grew to 319 lines — flagged in 3rd consecutive analysis cycle with no remediation.
• Severity: Medium — functional but accumulating technical debt
• Trend: Worsening (315→319 lines since last check)
• Recommended Action: Refactor validateWorkflowData into domain-focused validators (est. 2-3 hours work)

3. MCP Summary Quality Gap ⚠️

• Affected Reports: Observability Coverage Report (2026-03-24)
• Metric: MCP telemetry quality
• Description: Run §23517034683 has rpc-messages.jsonl present but 0 parsed MCP call entries in run_summary.json. Telemetry file exists but summary parsing failed.
• Severity: Low — telemetry not lost, summary metadata incomplete
• Recommended Action: Add post-processing validation for zero-entry MCP summaries

Data Quality Notes

• Safe Output Health report covers only 10 runs (vs 234 yesterday) — significantly lower volume; likely reflects scheduling patterns rather than a system issue
• Agentic Observability Report covers 2026-03-11 to 2026-03-25 (14-day window) rather than daily — wider scope than other daily reports; baseline comparisons are not yet available (all baseline_found: false)
• 12+ workflows consistently flagged as overkill_for_agentic (low severity) — portfolio cleanup opportunity noted but not urgent
• Glossary Maintainer run on 2026-03-18 showed HIGH resource usage (19 tool types, 21 turns) — single occurrence, not escalated

---

📈 Trend Analysis

Notable Trends

• Copilot success rate declining: 78.6% (Feb 21-28) → 48.1% (Mar 17-24) → 25% partial (Mar 25, 3 in-progress). Harder task categories (build/WASM, workflow updates) are weighing on success rate.
• PR velocity very high: 76 Copilot PRs in 24h (55 merged at 72.4%) with avg merge time of 40.7 min (down from 71.0 min previous day — 43% faster)
• Codebase growth steady: +1,541 LOC/day on 2026-03-24; 694K total LOC, stable quality at 73/100
• Security focus: 7 CodeQL vulnerabilities remediated in 24h; DIFC proxy, protected branch guardrails added
• Observability maturing: New deterministic lineage tracking and drop-in observability kit landed

Week-over-Week Session Performance

Period	Copilot Success Rate	Sessions/Day
Feb 21–28	78.6%	1.75
Mar 1–8	72.7%	2.75
Mar 9–16	73.3%	1.9
Mar 17–24	48.1%	3.4
Mar 25 (partial)	25%*	4

*3 sessions still in-progress at capture time

---

📝 Per-Report Analysis

View Per-Report Metrics

Safe Output Health Report (#22903)

Period: 2026-03-25 (last 24h) | Quality: ⚠️ Valid with warning

Metric	Value
Runs analyzed	10
Safe output messages	8 (7 executed + 1 skipped custom)
Job failures	0
Success rate	100%
Recurring warning	EP_LABELS_MISSING (day 2)

Copilot Session Insights (#22881)

Period: 2026-03-25 (window: 11:22-11:30 UTC) | Quality: ✅ Valid

Metric	Value
Sessions analyzed (window)	50
Copilot sessions today	4 (1 success, 3 in-progress)
All-time success rate	65.9% (56/85, 31 days)
Last 7-day success rate	53.8% (14/26)
Avg session duration	7.5m today / 11.2m all-time
Active branches	7 (all-time record)

Copilot Agent Analysis (#22878)

Period: 2026-03-24 11:33Z → 2026-03-25 11:33Z | Quality: ✅ Valid

Metric	Value
agent_prs_total	76
agent_prs_merged	55 (72.4%)
Closed without merge	13
Open (in-progress)	8
Avg merge duration	40.7 min

Daily Team Evolution (#22856)

Period: 2026-03-25 last 24h | Quality: ✅ Valid

Metric	Value
Commits	30
Active contributors	5 (human) + Copilot
Notable PRs merged	6
Security fixes	7 CodeQL alerts resolved
Theme	Security hardening + Observability

Agentic Observability Report (#22849)

Period: 2026-03-11 to 2026-03-25 | Quality: ✅ Valid

Metric	Value
workflow_runs_analyzed	25
High-confidence episodes	25 (100%)
Escalation-eligible	0
MCP failures	0
Blocked network requests	0
overkill_for_agentic workflows	12+ (low severity)

Observability Coverage Report (#22786)

Period: Last 7 days (2026-03-24) | Quality: ⚠️ 1 warning

Metric	Value
workflow_runs_analyzed	14 (12 completed)
Firewall coverage	100% (12/12)
MCP telemetry coverage	100% (12/12)
MCP summary quality	⚠️ 1 run with 0 parsed entries
Total firewall log entries	383
Blocked requests	6 (1.57%)

Daily Performance Summary (#22783)

Period: Last 90 days (2026-03-24) | Quality: ✅ Valid

Metric	Value
Total PRs (sample)	100
merged_prs	73 (73%)
open_prs	4
total_issues (sample)	1,000
open_issues	99
closed_issues	901 (90.1%)
Avg PR merge time	0.8 hours
Avg issue close time	15.8 hours
Unique contributors	6

Daily Secrets Analysis (#22778)

Period: 2026-03-24 | Quality: ✅ Valid

Metric	Value
Workflow files analyzed	177
Total secrets references	5,944
Workflows with redaction	177/177 (100%)
Hardcoded tokens	0
AI engine distribution	Copilot 118 / Claude 40 / Codex 19 / Gemini 1

Daily Compiler Code Quality (#22773)

Period: 2026-03-24 | Quality: ⚠️ Regression in compiler.go

Metric	Value
Files analyzed	3
Avg quality score	80/100
compiler.go score	76/100 (was 85, -9)
compiler_jobs.go score	86/100 (+1)
compiler_yaml.go score	78/100 (stable)
validateWorkflowData length	319 lines (3rd flagged cycle)

Daily Code Metrics (#22769)

Period: 2026-03-24 | Quality: ✅ Valid

Metric	Value
lines_of_code_total	694,040
Go source LOC	141,658
test_lines_of_code	317,991
test_to_source_ratio	2.245×
total_workflows (.md)	241
Compiled workflows (.lock.yml)	177
Quality score	73/100
Large files (>500 LOC)	243 Go files

---

💡 Recommendations

Process Improvements

1. Create missing labels (constraint-solving, problem-of-the-day) to resolve the EP_LABELS_MISSING warning before it reaches the 3-day escalation threshold.
2. Refactor validateWorkflowData in compiler.go (319 lines → 5 focused validators) — flagged 3 consecutive cycles, needs action.
3. Investigate Copilot success rate decline (48.1% last 7 days vs 65.9% all-time) — disproportionate share of hard task categories (build, WASM, workflow updates) may be driving the dip.

Data Quality Actions

1. Add MCP summary validation: Post-process run_summary.json to warn when rpc-messages.jsonl exists but parsed call count is zero.
2. Resolve gh CLI auth for Copilot session behavioral analysis (conversation logs currently unavailable).
3. Establish observability baselines: All 25 episodes show baseline_found: false — baselines will improve regression detection as runs accumulate.

Workflow Suggestions

1. Portfolio cleanup: Review 12+ overkill_for_agentic workflows (lean, read-only, Issue Response domain) for replacement with deterministic GitHub Actions.
2. Copilot task difficulty pre-classification: Pre-classify tasks as Low/Medium/High difficulty to set expectations and route appropriately.

---

📊 Regulatory Metrics

Metric	Value
Reports Reviewed	10
Reports Passed (no issues)	7
Reports with Warnings	3
Reports Failed	0
Critical Discrepancies	0
Overall Health Score	90%
Previous Regulatory Report Closed	#22762

---

References:

• §23564136872 — Regulatory Report Workflow Run

AI generated by Daily Regulatory Report Generator · history

• expires on Mar 28, 2026, 9:12 PM UTC

[github-actions[bot]]

This report has been superseded by a newer daily regulatory report for 2026-03-26.