[daily regulatory] Regulatory Report - 2026-04-11

[github-actions[bot]]

19 daily report discussions were reviewed for the 24-hour period ending 2026-04-11T21:00 UTC. Overall data quality is good, with consistent metrics across most reports. One recurring failure (upload_artifact in Smoke Copilot, now on Day 2) and a doubling of DIFC integrity-filtered events are the primary concerns. The agent ecosystem is recovering from the Apr 8–10 Copilot regression, with 80% of scheduled workflows healthy today.

Cross-report consistency is strong — firewall, observability, and performance metrics align within expected scopes. No critical discrepancies were found between reports sharing identical metric scopes.

📋 Full Regulatory Report

📊 Reports Reviewed

#	Report	Discussion	Created	Status
1	Daily Firewall Report	#25793	2026-04-11T11:51Z	✅ Valid
2	Daily Copilot Agent Analysis	#25791	2026-04-11T11:34Z	✅ Valid
3	Safe Output Health Report	#25811	2026-04-11T13:17Z	⚠️ Issues
4	Agent Performance Report	#25755	2026-04-11T04:45Z	⚠️ Issues
5	Observability Coverage Report	#25723	2026-04-10T23:53Z	✅ Valid
6	Daily Performance Summary	#25720	2026-04-10T23:31Z	✅ Valid
7	GitHub API Consumption Report	#25722	2026-04-10T23:48Z	✅ Valid
8	Daily Secrets Analysis	#25714	2026-04-10T22:24Z	✅ Valid
9	Daily Code Metrics Report	#25712	2026-04-10T21:58Z	✅ Valid
10	Daily Compiler Quality Report	#25711	2026-04-10T21:46Z	✅ Valid
11	DIFC Integrity-Filtered Events	#25841	2026-04-11T20:55Z	⚠️ Monitor
12	Daily Team Evolution Insights	#25782	2026-04-11T10:53Z	✅ Valid
13	Copilot PR Prompt Analysis	#25756	2026-04-11T04:54Z	✅ Valid
14	Auto-Triage Report (AM)	#25728	2026-04-11T01:02Z	✅ Valid
15	Auto-Triage Report (PM)	#25805	2026-04-11T12:46Z	✅ Valid
16	UX Analysis Report	#25799	2026-04-11T12:05Z	✅ Valid
17	Prompt Clustering Analysis	#25839	2026-04-11T20:18Z	✅ Valid
18	Sergo Report	#25840	2026-04-11T20:26Z	✅ Valid
19	Documentation Noob Test Report	#25815	2026-04-11T13:27Z	✅ Valid

---

🔍 Data Consistency Analysis

Cross-Report Metrics Comparison

Reference: scratchpad/metrics-glossary.md

Metric	Firewall Report	Performance Summary	Copilot Agent Analysis	Observability	Scope Match	Status
workflow_runs_analyzed	32 (Apr 11, firewall-enabled)	—	—	23 downloaded / 18 completed (7d)	⚠️ Different scopes	i️ See Note
open_prs (recent sample)	—	4/100 sampled	6 open agent PRs	—	⚠️ Different scopes	i️ See Note
merged_prs	—	66/100 sampled	29/48 agent PRs (60%)	—	⚠️ Different scopes	i️ See Note
open_issues	—	114/1000 sampled	—	—	i️ Sampled only	✅
firewall_requests_allowed	805/805 (100%)	—	—	—	✅ Single source	✅
firewall_requests_blocked	0 (0%)	—	—	—	✅ Single source	✅
observability_coverage_percentage	—	—	—	100% (18/18 completed)	✅ Single source	✅
agent_prs_total	—	—	48 (last 24h)	—	✅ Single source	✅
agent_success_rate	—	—	69% (29/42 effective)	—	✅ Single source	✅

Scope Notes:

• workflow_runs_analyzed: Firewall Report covers firewall-enabled runs on Apr 11 only (32 runs); Observability covers 7d all completed runs (23 downloaded). Different scopes by design — not a discrepancy.
• open_prs / merged_prs: Performance Summary samples recent 100 PRs (all authors); Copilot Agent Analysis covers only agent-created PRs in last 24h. Different scopes by design.

Consistency Score

• Overall Consistency: 95% (no true same-scope discrepancies found)
• Critical Discrepancies: 0
• Minor Discrepancies: 0
• Scope Differences Noted: 3 (expected, documented above)

---

⚠️ Issues and Anomalies

Critical Issues

No critical discrepancies between reports.

Warnings

1. Recurring upload_artifact failure — Smoke Copilot (Day 2)

   • Affected Report: Safe Output Health #25811
   • Metric: upload_artifact success rate = 0% (1 failure, 1 attempt)
   • Description: Staging path missing for artifact upload in Smoke Copilot. Now occurring for 4 consecutive runs since 2026-04-10.
   • Severity: Medium
   • Recommended Action: Investigate Smoke Copilot artifact staging path; verify /tmp/gh-aw directory creation in pre-steps

2. DIFC integrity-filtered events doubling

   • Affected Report: DIFC Integrity-Filtered Events #25841
   • Metric: Filtered events: 12 (Apr 10) → 24 (Apr 11), ×2 increase
   • Description: All events are pull_request_read blocks from Design Decision Gate and Smoke Claude. Expected behavior (integrity guard working correctly), but rate is rising.
   • Severity: Low (monitor)
   • Recommended Action: Add PR integrity promotion step for Copilot-generated PRs if Design Decision Gate should review them

3. Design Decision Gate — Empty prompt failures

   • Affected Report: Agent Performance #25755
   • Description: --print flag causes empty prompt; ~2/3 recent runs failed. Root cause documented in issue [deep-report] Fix Design Decision Gate 100% failure — empty prompt when --print flag used #25670.
   • Severity: High (architecture decisions blocked)
   • Recommended Action: Fix pending; track PR resolution

4. Documentation Unbloat — Zero safe outputs for 2+ weeks

   • Affected Report: Agent Performance #25755
   • Metric: Q:55, E:20; ~$5/run, ~$55/week wasted
   • Description: Agent runs successfully but produces no create_issue or add_comment calls
   • Severity: Medium (cost waste)
   • Recommended Action: Audit agent prompt and output conditions; add explicit safe-output requirement check

Data Quality Notes

• Safe Output Health report covers only 57 completed runs out of ~35 active workflows — 3 additional runs in-progress at time of report
• API Consumption success rate of 20.0% is misleading: 74/95 runs were filtered by activation gate (not true failures). Actual success/fail ratio is 19/21 (90.5%)
• Discussion answer rate 0.0% (0/100) is expected — discussions in this repo are automated reports, not Q&A threads

---

📈 Trend Analysis

Key Metrics vs Yesterday (#25706 — 2026-04-10 regulatory)

Metric	Today (Apr 11)	Yesterday (Apr 10)	Change
Agent PRs total (24h)	48	42	+14% ↑
Agent success rate	69%	66%	+3pp ↑
Safe output success rate	96%	~96%	→ stable
Firewall block rate	0.0%	0.0%	→ stable
Observability coverage	100%	100%	→ stable
DIFC filtered events	36 total (24 today)	12 today	+100% ↑ (monitor)
Healthy workflows	20/25 (80%)	—	—
Open P2 issues	4	—	—
API calls (daily)	4,560	~6,770 (7d avg)	-33% ↓

Notable Trends

• Agent ecosystem recovery: Success rate improving (66% → 69% → partial 73%) as Copilot v1.0.24 rollout stabilizes the CLI regression from Apr 8–10
• Firewall remains clean: 0 blocked domains for 2+ consecutive days; all 805 requests allowed
• Observability at 100%: All completed runs have both firewall access logs and MCP gateway telemetry — excellent coverage for debugging
• API calls decreasing: 4,560 today vs ~6,770 7-day average; fewer heavy consumer runs today (no Go Logger Enhancement equivalent)

---

📝 Per-Report Analysis

View Individual Report Details

Daily Firewall Report #25793

Period: 2026-04-11 | Quality: ✅ Valid

Metric	Value	Validation
workflow_runs_analyzed	32	✅
firewall_requests_total	805	✅
firewall_requests_allowed	805 (100%)	✅
firewall_requests_blocked	0 (0%)	✅
firewall_domains_blocked	0	✅
Unique allowed domains	3	✅

Notes: Clean firewall posture. Top domains: api.githubcopilot.com (439), api.anthropic.com (193), github.com (17).

---

Daily Copilot Agent Analysis #25791

Period: 2026-04-10T11:30Z → 2026-04-11T11:30Z | Quality: ✅ Valid

Metric	Value	Validation
agent_prs_total	48	✅
agent_prs_merged	29	✅
agent_success_rate	69%	✅ Math: 29/42=69% ✅
Avg Duration	94 min	✅
Open PRs	6	✅
Closed PRs	13	✅

Notes: Trend improving (66% → 69% → partial 73%). Many merged PRs address benchmark regressions and security fixes.

---

Safe Output Health Report #25811

Period: 2026-04-11 (last 24h) | Quality: ⚠️ Issues

Metric	Value	Validation
Runs analyzed	57 completed + 3 in-progress	✅
Safe output attempts	25	✅
Safe output successes	24	✅
Hard failures	1	⚠️ Recurring Day 2
Overall success rate	96%	✅ Math: 24/25=96% ✅
upload_artifact success rate	0% (1/1 failed)	❌

Notes: The upload_artifact failure in Smoke Copilot is the only failure. All other job types succeeded at 100%.

---

Observability Coverage Report #25723

Period: Last 7 days | Quality: ✅ Valid

Metric	Value	Validation
Runs downloaded	23	✅
Completed runs (in scope)	18	✅
AWF Firewall coverage	100% (18/18)	✅
MCP telemetry coverage	100% (18/18)	✅
observability_coverage_percentage	100%	✅

---

Daily Performance Summary #25720

Period: 2026-04-10 | Quality: ✅ Valid

Metric	Value	Validation
PRs sampled (total_prs)	100	✅
Merged PRs	66	✅
Open PRs	4	✅
Issues sampled	1000	✅
Open issues	114	✅
Closed issues	886	✅ Math: 114+886=1000 ✅
Avg merge time	1.6 hours	✅
Avg resolution time	11.0 hours	✅

---

GitHub API Consumption Report #25722

Period: 2026-04-10 | Quality: ✅ Valid

Metric	Value	Validation
Total runs	95	✅
Active runs (not skipped)	21	✅
Succeeded	19	✅
Failed	2	✅ Math: 19+2=21 ✅
Skipped (activation gate)	74	✅ Math: 74+21=95 ✅
GitHub REST API calls	4,560	✅
Safe-output writes	10	✅
Avg duration	92.2s	✅

Notes: 20.0% headline success rate is activation-gate filtered. True success rate of active runs = 19/21 = 90.5%.

---

Agent Performance Report #25755

Period: Apr 4–11, 2026 | Quality: ⚠️ Issues

Metric	Value	Validation
Agents analyzed	16	✅
Avg quality score	70/100	✅
Avg effectiveness score	60/100	✅
Healthy workflows	20/25 (80%)	✅
Smoke failures	4	⚠️
Open P2 issues	4	✅

---

DIFC Integrity-Filtered Events #25841

Period: Last 7 days | Quality: ⚠️ Monitor

Metric	Value	Validation
Total filtered events (7d)	36	✅
Apr 10 events	12	✅
Apr 11 events	24	⚠️ 2× increase
Unique tools filtered	1 (pull_request_read)	✅
Unique workflows affected	2	✅

---

💡 Recommendations

Process Improvements

1. Fix upload_artifact staging path in Smoke Copilot: The failure is now in its 4th consecutive occurrence. Investigate whether the staging directory creation step is missing or conditional.

2. Address Design Decision Gate --print empty prompt bug: Issue [deep-report] Fix Design Decision Gate 100% failure — empty prompt when --print flag used #25670 documents root cause. A fix PR should be prioritized to unblock architecture review gates.

3. Audit Documentation Unbloat agent outputs: Zero safe outputs for 2+ weeks at ~$5/run represents avoidable cost. Add explicit output validation or disable until fixed.

4. Consider PR integrity promotion for agent PRs: DIFC filtered events doubling suggests Design Decision Gate is frequently seeing unreviewed Copilot PRs. A pre-gate step to elevate trusted PRs to approved would unblock the workflow.

Data Quality Actions

1. Clarify API success rate reporting: The 20.0% headline in API Consumption misleads; consider reporting "active run success rate" (90.5%) as the primary metric, with activation-gate filtering noted separately.

2. Add close_discussion support to safe-outputs: The previous regulatory report (#25706) cannot be automatically closed as the close_discussion tool is not available in this workflow's safe-output configuration. Manual closure required.

---

📊 Regulatory Metrics

Metric	Value
Reports Reviewed	19
Reports Passed (✅ Valid)	15
Reports with Warnings (⚠️)	3
Reports Failed (❌)	0
Critical Discrepancies	0
Minor Discrepancies	0
Open Action Items	4
Overall Health Score	84%

---

References:

• §24291561492 — This regulatory workflow run
• §24290422372 — DIFC Integrity-Filtered Events Analyzer
• Previous regulatory report: #25706 (2026-04-10, not auto-closed — close_discussion tool unavailable)

Generated by Daily Regulatory Report Generator · ● 789.9K · ◷

• expires on Apr 14, 2026, 9:09 PM UTC

[github-actions[bot]]

This discussion has been marked as outdated by Daily Regulatory Report Generator.

A newer discussion is available at Discussion #26088.