[WHM] Workflow Health Dashboard — 2026-03-31 #23692

Overview

Score: 73/100 (↑1 from 72). 178 workflows total. Run §23796317224

Summary

P1 — Critical Issues 🚨

Smoke Update Cross-Repo PR (#23193 — open)

• Schedule runs [Custom Engine Test] Test Pull Request - Custom Engine Safe Output #438–[Custom Engine Test] Test Issue Created by Custom Engine #450: ALL failing
• Latest: run [Custom Engine Test] Test Issue Created by Custom Engine #450 failed 2026-03-31T01:01Z
• Root cause: push_repo_memory git checkout removes actions/setup from working tree
• Action: Status update comment added to [aw] Smoke Update Cross-Repo PR failed #23193

Smoke Create Cross-Repo PR (new issue #23691)

• Schedule runs [Custom Engine Test] Test Issue Created by Custom Engine #434–[Custom Engine Test] Test Issue Created by Custom Engine #450: ALL failing since 2026-03-25
• Previous issue [WHM] Smoke Create Cross-Repo PR - Persistent Schedule Failures (Systemic Bug) #23447 expired and auto-closed 2026-03-30
• Same root cause as above
• Action: New P1 tracking issue created

P2 — External Issues (Team closed as not_planned) ⚠️

Smoke Codex (#23431 — closed not_planned 2026-03-31)

• OpenAI API model access restriction
• Schedule runs still failing (Add Copilot CLI and VS Code MCP server configuration instructions #2548–Safe output messages #2560)
• Team acknowledged; no fix from this team possible

Smoke Gemini (#23399 — closed not_planned 2026-03-31)

• Exit code 41 failure pattern
• Schedule runs #651–[Custom Engine Test] Test Pull Request - Custom Engine Safe Output #663 all failing
• Team acknowledged; monitoring

✅ Recoveries

• PR Triage Agent ([aw] PR Triage Agent failed #23151, open): 5 consecutive successes don't run add-reaction on PRs of forked repos #305–309. Please close manually — update_issue cannot be called in schedule context.
• Metrics Collector: Runs MCP Network Permissions Test Results - Proxy Isolation Analysis #109–110 success. No open issue.

WHM Self-Failure Root Cause Identified

WHM runs #282–286 ALL failed in safe_outputs. Root cause:

• Agent called update_issue (3 times) to close stale issues
• WHM runs on schedule → no triggering issue → update_issue fails
• Fix: use only add_comment (not update_issue) in schedule-triggered workflows
• This run ([UPDATED] Update Issue Test - Processed by Codex #287) uses only add_comment + create_issue → should pass clean

Trends

• Overall health: 72→73 (+1)
• External API issues (Codex, Gemini) acknowledged by team as not_planned
• Systemic push_repo_memory bug (2 affected smoke tests) pending fix in compiler_yaml.go

References:

• §23796317224 — This run
• #23193 — Smoke Update Cross-Repo PR (P1, ongoing)

Generated by Workflow Health Manager - Meta-Orchestrator · ◷

• expires on Apr 1, 2026, 12:16 PM UTC