Recompile workflow lock files with updated MCP environment variables

[Copilot]

Regenerated all 65 workflow lock files to incorporate updated MCP server configuration that includes GITHUB_REPOSITORY and GITHUB_SERVER_URL environment variables.

Changes

• Ran make recompile to regenerate all .lock.yml files from their .md sources
• Updated MCP server environment variable passthrough in all workflows to include repository context variables

# Before
"env": {
  "GH_AW_SAFE_OUTPUTS_CONFIG": "${GH_AW_SAFE_OUTPUTS_CONFIG}",
  "GH_AW_ASSETS_BRANCH": "${GH_AW_ASSETS_BRANCH}"
}

# After
"env": {
  "GH_AW_SAFE_OUTPUTS_CONFIG": "${GH_AW_SAFE_OUTPUTS_CONFIG}",
  "GH_AW_ASSETS_BRANCH": "${GH_AW_ASSETS_BRANCH}",
  "GITHUB_REPOSITORY": "${GITHUB_REPOSITORY}",
  "GITHUB_SERVER_URL": "${GITHUB_SERVER_URL}"
}

All 65 workflows compiled successfully with 0 errors.

Original prompt

Recompile

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

[Copilot]

Pull Request Overview

This PR regenerates all 65 GitHub Actions workflow lock files to include additional environment variables (GITHUB_REPOSITORY and GITHUB_SERVER_URL) for MCP server configuration.

• Automated recompilation of all workflow lock files using make recompile
• Addition of two GitHub context environment variables to MCP server configurations
• Consistent environment variable passing across all workflow files

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[github-actions]

Agentic Changeset Generator triggered by this pull request.

[github-actions]

🔍 Smoke Test Investigation - Run #61

Summary

The Smoke OpenCode workflow failed due to a transient Anthropic API error (AI_APICallError). This is a well-documented recurring pattern - the 12th occurrence since October 22, 2025. OpenCode successfully initialized (installed plugins, configured MCP servers) but failed when attempting the first Anthropic API call.

Failure Details

• Run: 18988563486
• Commit: 1b0d76e
• Trigger: schedule
• Duration: 1.2m
• Failed Job: agent (43s)

Root Cause Analysis

Pattern: OPENCODE_ANTHROPIC_API_ERROR

• Category: AI Engine - API Failure
• Severity: High
• Is Transient: Yes ✅
• Is Flaky: Yes ✅
• Occurrence Count: 12 occurrences since Oct 22, 2025
• Affects: ~25% of OpenCode smoke test runs

What Happened

1. OpenCode initialized successfully (plugins installed, MCP servers configured)
2. Session created and prompt processed
3. API call to (redacted)/v1/messages failed with AI_APICallError
4. Agent completed with 0 turns, 0 token usage
5. No safe-outputs or artifacts created
6. Downstream jobs skipped/failed

Root Causes (from pattern analysis)

1. Anthropic API temporarily unavailable (likelihood: high)
   • Evidence: Transient nature, resolves on retry
2. API rate limiting during concurrent requests (likelihood: medium)
   • Evidence: Often occurs during scheduled runs when multiple workflows execute
3. Network connectivity issues (likelihood: low)
   • Evidence: Other API calls succeed in same workflow

Historical Context

Related Issues: #2852, #2772, #2730, #2143, #2121, #2604, #2307

Recent Occurrences:

• 2025-10-31 (redacted) - Run 18964376017 ([smoke-detector] 🔍 Smoke Test Investigation - Smoke OpenCode Run #55: OpenCode Agent Execution Failure (Recurring) #2852, closed as "not planned")
• 2025-10-30 (redacted) - Run 18931623375 ([smoke-detector] 🔍 Smoke Test Investigation - Smoke OpenCode Run #51: OpenCode Agent Execution Failure #2772, closed as "not planned")
• 2025-10-30 (redacted) - Run 18940217849 (cached)
• 2025-10-30 (redacted) - Run 18926079635 (cached)
• 2025-10-29 (redacted) - Run 18907236182 (cached)

Recommended Actions

🔴 High Priority (Recommended 12 times, still unimplemented)

1. Add retry logic with exponential backoff

- name: Run OpenCode (with retry)
  uses: nick-fields/retry@v3
  with:
    timeout_minutes: 5
    max_attempts: 3
    retry_wait_seconds: 30
    exponential_backoff: true
    command: |
      # OpenCode execution command

Impact: Would prevent ~90% of these transient failures

2. Make create_issue job conditional

create_issue:
  needs: [agent]
  if: needs.agent.result == 'success'

Impact: Prevents cascading failures, clearer error signals

🟡 Medium Priority

3. Add API health check

• Query Anthropic API status before running agent
• Skip with clear message if API is down

Prevention Strategies

1. Retry Logic: 3 attempts, 30s delay, 2x backoff → 90%+ success rate improvement
2. Conditional Jobs: Only run downstream jobs when agent succeeds
3. Health Checks: Pre-flight checks for OpenCode CLI and Anthropic API
4. Stagger Scheduled Runs: Reduce concurrent API load during schedule triggers
5. Circuit Breaker: Implement for repeated failures

Assessment: Related to PR #2918?

No ❌ - This failure is unrelated to the PR changes.

Evidence:

• This is a well-established recurring transient pattern (12 occurrences)
• Pattern documented since Oct 22, 2025
• Occurs randomly during scheduled runs regardless of code changes
• PR Recompile workflow lock files with updated MCP environment variables #2918 only recompiled workflow lock files with MCP environment variables
• Failure occurs at Anthropic API call, not workflow configuration

Conclusion: The timing is coincidental. This is an infrastructure/API issue, not a code bug.

---

**Investigation (redacted)

• Pattern Database: /tmp/gh-aw/cache-memory/patterns/opencode_anthropic_api_error.json
• Investigation Record: /tmp/gh-aw/cache-memory/investigations/2025-11-01-18988563486.json
• Investigator: Smoke Detector

AI generated by Smoke Detector - Smoke Test Failure Investigator