[workflow-health] PR Triage Agent: push_repo_memory job failing (post-cleanup error)

[github-actions]

Problem

PR Triage Agent started failing on 2026-03-26 with 2 consecutive failures in the push_repo_memory job.

• Run #289 — failed 2026-03-26T06:21
• Run #288 — failed 2026-03-26T00:12
• Run #287 — ✅ success 2026-03-25T18:16

Root Cause

The push_repo_memory job switches the workspace branch to memory/pr-triage to commit and push memory files. After the push succeeds, the GitHub Actions post-job cleanup for the ./actions/setup local action fails because the workspace is now on the memory/pr-triage branch (which doesn't have the actions/ directory).

Error in post-cleanup:

##[error]Can't find 'action.yml', 'action.yaml' or 'Dockerfile' under 
'/home/runner/work/gh-aw/gh-aw/actions/setup'. Did you forget to run 
actions/checkout before running your local action?

Important: The actual memory push succeeds — data is correctly written to memory/pr-triage branch. The failure only affects the post-job cleanup phase.

Evidence

From run #289 logs (push_repo_memory job):

• ✅ Checked out branch memory/pr-triage
• ✅ Committed and pushed 3 files (pr-triage-latest.json, pr-triage-summary.md, README.md)
• ❌ Post-job cleanup fails: can't find actions/setup

Impact

• PR Triage Agent reports are still being generated (the agent job succeeds)
• Repo memory is being saved successfully
• But the overall workflow run is marked as failure due to the cleanup error

Suggested Fix

Options:

1. Add a checkout step at the end of push_repo_memory to restore the workspace to the main branch before cleanup
2. Use persist-credentials: false + sparse checkout on main as the last step to restore actions/setup
3. Set continue-on-error: true on steps that might leave the workspace on the memory branch

The fix should be implemented in the workflow compiler's push_repo_memory job generation, so all workflows that use repo-memory benefit from the fix.

Priority

P1 — New regression, 2 consecutive failures starting 2026-03-26T00:12

Detected by §23593331592 — Workflow Health Manager

Note

🔒 Integrity filter blocked 3 items

The following items were blocked because they don't meet the GitHub integrity level.

• #21703 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #18407 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #21784 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".

To allow these resources, lower min-integrity in your GitHub frontmatter:

tools:
  github:
    min-integrity: approved  # merged | approved | unapproved | none

Generated by Workflow Health Manager - Meta-Orchestrator · ◷

• expires on Mar 27, 2026, 12:14 PM UTC

[github-actions]

🍪 Issue Monster has assigned this to Copilot!

I've identified this issue as a good candidate for automated resolution and assigned it to the Copilot coding agent.

The Copilot coding agent will analyze the issue and create a pull request with the fix.

Om nom nom! 🍪

🍪 Om nom nom by Issue Monster · ◷

[copilot-swe-agent]

The agent encountered an error and was unable to start working on this issue: This may be caused by a repository ruleset violation. See granting bypass permissions for the agent, or please contact support if the issue persists.

[github-actions]

Status Update — 2026-03-27T12:03Z | §23645263363

Escalation: 6 consecutive failures (was 2 when issue was filed)

Run	Time	Result
#288	2026-03-26T00:12	❌
#289	2026-03-26T06:21	❌
#290	2026-03-26T12:13	❌
#291	2026-03-26T18:17	❌
#292	2026-03-27T00:11	❌
#293	2026-03-27T06:22	❌

Pattern Confirmed Systemic

The push_repo_memory → Post Setup Scripts failure is confirmed to affect multiple workflows:

• PR Triage Agent: 6/6 failures in push_repo_memory
• Metrics Collector (run #106): also failed with push_repo_memory → Post Setup Scripts failure

Root cause remains: after pushing to the memory/* branch, the workspace is left on that branch. When GitHub Actions runs post-job cleanup for ./actions/setup, it can't find action.yml because the memory branch doesn't have the actions/ directory.

Auto-Fix Status

Issue Monster assigned this to Copilot on 2026-03-26T12:34, but Copilot's agent reported:

"This may be caused by a repository ruleset violation."

The fix requires adding a checkout-restore step to the push_repo_memory job generator in pkg/workflow/compiler_yaml.go. Manual intervention needed.

Recommended Fix

In the compiled push_repo_memory job, add a final step to restore the workspace branch to main before cleanup runs:

- name: Restore workspace branch
  if: always()
  run: git checkout main || git checkout -

This should be implemented in the workflow compiler so all workflows with repo-memory benefit automatically.

Generated by Workflow Health Manager - Meta-Orchestrator · ◷

[github-actions]

This issue was automatically closed because it expired on 2026-03-27T12:14:04.829Z.

Closed by Workflow