[daily regulatory] Regulatory Report - 2026-02-20

[github-actions[bot]]

Reviewed 21 daily report discussions generated on 2026-02-20 across all active report categories. Overall data quality is good, with most reports internally consistent and well-structured. Three critical issues require prompt attention: Issue Monster's 100% failure rate (16/16 runs wasting compute), zero MCP gateway log coverage (0/4 runs), and the Observability report being severely limited by a tool timeout (only 10 of 43 runs captured). Minor discrepancies in safe output job failures were also identified.

Cross-report metric consistency is largely sound — expected scope differences between reports are well-documented and align with the metrics glossary. The token consumption and audit reports use different time windows and run samples, which is appropriate. No metric values with identical scopes were found to diverge by more than 10%.

📋 Full Regulatory Report

📊 Reports Reviewed

#	Report	Discussion	Created	Status
1	Daily Performance Summary	#17263	19:45 UTC	✅ Valid
2	Auto-Triage (latest)	#17254	18:31 UTC	✅ Valid
3	Repository Chronicle	#17220	16:15 UTC	✅ Valid
4	Daily Team Evolution	#17219	16:10 UTC	✅ Valid
5	DeepReport Intelligence	#17203	15:30 UTC	✅ Valid
6	Copilot PR Merged Report	#17193	15:25 UTC	✅ Valid
7	Repository Quality Improvement	#17164	13:37 UTC	✅ Valid
8	Daily Code Metrics	#17136	11:49 UTC	✅ Valid
9	UX Analysis	#17129	11:37 UTC	✅ Valid
10	Daily Copilot Token Consumption	#17124	11:30 UTC	⚠️ Issues
11	Daily Status	#17115	09:27 UTC	✅ Valid
12	Documentation Noob Test	#17110	08:11 UTC	✅ Valid
13	Observability Coverage	#17108	07:56 UTC	⚠️ Issues
14	Static Analysis	#17105	06:37 UTC	✅ Valid
15	Agentic Workflow Audit	#17103	05:52 UTC	⚠️ Issues
16	Safe Output Health	#17095	04:31 UTC	⚠️ Issues
17	Daily Firewall	#17043	01:05 UTC	✅ Valid
18	Daily Compiler Code Quality	#17013	00:25 UTC	✅ Valid

🔍 Data Consistency Analysis

Cross-Report Metrics Comparison

Metric definitions and scope documentation: scratchpad/metrics-glossary.md

Metric	Performance Rpt	Daily Status	Scope	Status
open_issues	55	—	✅ Same scope	✅ N/A
merged_prs (90d)	141	—	✅ Same scope	✅ OK
open_prs	7	—	✅ Same scope	✅ OK
Issues opened today	—	76 opened, 87 closed	n/a	✅ Internally consistent
total_loc	—	—	Code Metrics only	✅ 821,420
workflow_runs_analyzed	Firewall: 30/43	Observability: 10/43	⚠️ Different coverage	⚠️ See Note
total_tokens (today)	Token: 166.5M/195 runs	Audit: 148.8M/22 runs	⚠️ Different scopes	ℹ️ Expected

Scope Notes:

• workflow_runs_analyzed: Firewall Report covered 30 of 43 runs (7d); Observability covered only 10 due to MCP logs tool timeout — both say "last 7 days" but with different tooling limits. Not a data error, but Observability data is incomplete.
• total_tokens: Token report spans 195 runs (30-day data window Feb 19–20); Audit report spans 22 sampled runs from a single morning. Different scopes by design.
• issues_analyzed: Performance report uses 1000 issues (sampled); different from issue-specific reports — expected per glossary.

Consistency Score

• Overall Consistency: 85% — metrics where same-scope data is available are consistent
• Critical Discrepancies: 0
• Minor Discrepancies: 1 (Observability incomplete data vs Firewall full coverage)

⚠️ Critical Findings

🔴 Issue Monster — 100% Failure Rate

• Affected Reports: Token Consumption (📊 Daily Copilot Token Consumption Report - 2026-02-20 #17124), Agentic Workflow Audit (Agentic Workflow Audit Report — 2026-02-20 #17103), Observability ([observability] Observability Coverage Report - 2026-02-20 #17108)
• Metric: agent_success_rate for Issue Monster workflow
• Description: Issue Monster failed in 16 out of 16 runs (100% failure rate), consuming zero tokens but wasting runner compute on each attempt.
• Impact: 16 failed runs contributed to 23 total token-less runs, distorting aggregate statistics.
• Severity: 🔴 Critical
• Recommended Action: Disable or pause Issue Monster workflow immediately; investigate root cause before re-enabling.

🔴 MCP Gateway Log Coverage — 0%

• Affected Reports: Observability Coverage ([observability] Observability Coverage Report - 2026-02-20 #17108)
• Metric: observability_coverage_percentage for MCP gateway
• Description: Of 4 MCP-enabled runs analyzed, 0 had gateway.jsonl logs present. This means MCP interactions are completely unauditable.
• Missing MCP runs: Workflow Health Manager - Meta-Orchestrator, Go Fan, Auto-Triage Issues, Static Analysis Report
• Severity: 🔴 Critical
• Recommended Action: Investigate why MCP gateway logs are not being written to artifacts; verify log collection step is wired up for MCP-enabled workflows.

🟡 Observability Severely Under-Sampled

• Affected Reports: Observability Coverage ([observability] Observability Coverage Report - 2026-02-20 #17108)
• Metric: workflow_runs_analyzed
• Description: Observability only captured 10 of 43 known firewall-enabled runs due to MCP logs tool timeout/limit. Firewall report captured 30/43. This 67% gap means true coverage metrics are unknown.
• Severity: 🟡 High
• Recommended Action: Increase count parameter or implement pagination in the Observability workflow; resolve MCP logs tool timeout.

🟡 Token Report — Ambiguous Analysis Window

• Affected Reports: Daily Copilot Token Consumption (📊 Daily Copilot Token Consumption Report - 2026-02-20 #17124)
• Description: Report headline says "last 30 days" but data footnote states "Analysis Period: 2026-02-19 to 2026-02-20 (2-day window in this dataset)". The actual data coverage is 2 days, not 30 days. This discrepancy could mislead consumers of this report.
• Severity: 🟡 Medium
• Recommended Action: Align report title/summary with actual data window; clarify whether "30 days" refers to the intent or actual coverage.

📊 Detailed Per-Report Analysis

Daily Performance Summary (#17263)

Time Period: Last 90 days (2025-11-22 through 2026-02-20)
Quality: ✅ Valid

Metric	Value	Validation
total_prs	200 (sampled)	✅
merged_prs	141	✅
open_prs	7	✅
total_issues	1,000 (sampled)	✅
closed_issues	945	✅
open_issues	55	✅
total_discussions	100	✅
Discussions answered	0 (0.0%)	⚠️ Zero answer rate
Avg merge time	1.2 hours	✅
Unique contributors	10	✅

Notes: Math checks pass (141 merged + 7 open = 148; 52 closed without merge — plausible). Discussion answer rate of 0% is a recurring concern; recommend assigning a triage owner for discussions.

---

Daily Copilot Token Consumption (#17124)

Time Period: 2026-02-19 to 2026-02-20 (described as "30-day window")
Quality: ⚠️ Issues (ambiguous window)

Metric	Value	Validation
Total tokens today	166,507,574	✅
Workflow runs	195	✅
Unique workflows	50	✅
Avg tokens/run	853,885	✅ Math: 166.5M/195 = 854K ✅
Token trend vs yesterday	+39.5%	✅ (166M vs 119M)
Runs with no token data	23	⚠️ 11.8% waste rate
Issue Monster failures	16	🔴 100% failure

---

Observability Coverage (#17108)

Time Period: Last 7 days (partial — 10/43 runs)
Quality: ⚠️ Issues (incomplete coverage)

Component	Runs Analyzed	Logs Present	Coverage
AWF Firewall (access.log)	6	3	50% 🔴
MCP Gateway (gateway.jsonl)	4	0	0% 🔴

---

Daily Firewall Report (#17043)

Time Period: Last 7 days (30 of 43 runs with data)
Quality: ✅ Valid

Metric	Value	Validation
workflow_runs_analyzed	30/43	✅
firewall_requests_total	1,169	✅
firewall_requests_allowed	621 (53.1%)	✅
firewall_requests_blocked	548 (46.9%)	✅ Math: 621+548=1169 ✅
firewall_domains_blocked	3 unique	✅
Block rate	46.9%	⚠️ High — mostly system calls

Notes: 531/548 blocked requests (96.9%) are internal system calls (- category). Only 17 blocks are external (github.com, codeload.github.com), all from Changeset Generator doing Go module fetches.

---

Agentic Workflow Audit (#17103)

Time Period: Recent runs (22 analyzed)
Quality: ⚠️ Issues (failures noted)

Metric	Value	Validation
Runs analyzed	22	✅
Success rate	90.5% (19/21 completed)	✅
Total tokens	148.8M	✅
Total cost	$1.83	✅ (Claude only)
Failures	2 (Issue Monster, Smoke Gemini)	⚠️

---

Safe Output Health (#17095)

Time Period: 2026-02-20
Quality: ⚠️ Issues (minor failures)

Job Type	Total Sent	Failures	Status
add_comment	39	2	⚠️ Permission denied
update_project	30	2	⚠️ View limit + cascading temp ID
update_issue	1	1	⚠️ Context mismatch
All others	124	0	✅

---

Daily Code Metrics (#17136)

Time Period: Snapshot 2026-02-20
Quality: ✅ Valid

Metric	Value	Validation
lines_of_code_total	821,420	✅
Total files	2,539	✅

---

Daily Compiler Code Quality (#17013)

Time Period: Current codebase (commit a280a6b)
Quality: ✅ Valid

File	Quality Score	Rating
compiler.go (575 LOC)	78/100	Good
compiler_yaml.go (714 LOC)	79/100	Good
compiler_jobs.go (532 LOC)	84/100	Good

All 3 files rated "Good". No "Poor" or "Needs Work" ratings. Average: 80/100.

---

Daily Status (#17115)

Quality: ✅ Valid

Metric	Value
Releases (last 7 days)	23
Latest release	v0.47.2
Commits (last 30 days)	1,857
Issues opened today	76
Issues closed today	87
PRs opened today	17
PRs merged today	11

💡 Recommendations

Process Improvements

1. Disable Issue Monster: 100% failure rate (16/16 runs) with zero useful output is pure compute waste. Pause immediately and fix before re-enabling.
2. Fix MCP gateway logging: 0% coverage for MCP gateway logs means all MCP-enabled workflow interactions are unauditable. This is a compliance gap.
3. Resolve Observability tool timeout: Implement pagination or increase limits in the logs download step — 67% of runs are not being analyzed.

Data Quality Actions

1. Token Report window: Clarify the data window label. "Last 30 days" header is misleading when actual data is 2-day coverage — this should be corrected to prevent misinterpretation.
2. Discussion answer rate: 0% answer rate (0/100 discussions) is a persistent finding. Assign a rotating triage owner to address community questions.
3. Safe Output failures: Investigate update_project cascading temp ID issue — this type of failure can cause silent data loss in project tracking workflows.

Workflow Suggestions

1. Changeset Generator network permissions: Add github.com and codeload.github.com to the allowed domains list; Go module operations legitimately require these for go get/go mod tidy.
2. Persist observability baseline: Enable daily metrics caching in both Firewall and Observability reports for week-over-week trend comparison.

📊 Regulatory Metrics

Metric	Value
Reports Reviewed	21
Reports Passed	14
Reports with Issues	4
Reports Not Analyzed (scope)	3
Critical Issues	2
High Issues	1
Medium Issues	1
Overall Health Score	76%

---

Metric definitions: scratchpad/metrics-glossary.md

References: §22239173571

AI generated by Daily Regulatory Report Generator

• expires on Feb 23, 2026, 8:08 PM UTC

[github-actions[bot]]

🤖 Smoke test agent was here! Beep boop! 🚀

Just dropped by to say the Copilot smoke test is running and everything looks great so far. Keep up the amazing discussions! ✨

📰 BREAKING: Report filed by Smoke Copilot for issue #17245

[github-actions[bot]]

This report has been superseded by a newer daily regulatory report for 2026-02-21.