Normalize setup-js caught-error logging to use getErrorMessage

[Copilot]

This change addresses the ESLint cluster in actions/setup/js where caught errors were still stringified with String(err) / String(error) despite getErrorMessage() already being available in scope. The update makes error-message formatting consistent across the affected setup scripts without expanding scope beyond actions/setup/js.

• Scope

  • Replaced eligible caught-error stringification in 38 actions/setup/js modules called out by the issue.
  • Kept the change limited to sites where getErrorMessage(...) was already imported or otherwise available.

• What changed

  • Updated thrown error messages that wrapped filesystem, patch, gateway, eval, and artifact failures.
  • Updated warning / failure logging paths in command-dispatch and entrypoint code where caught errors were still falling back to String(...).

• Behavioral effect

  • Standardizes error output on the shared helper used elsewhere in the setup-js surface.
  • Avoids raw String(...) coercion in caught-error paths, aligning these files with the repository’s preferred error-message handling.

• Representative change

  throw new Error(`Failed to read file ${patchFilePath}: ${getErrorMessage(err)}`, { cause: err });

• Areas touched

  • Slash-command routing and PR dispatch helpers
  • Safe-outputs setup and loaders
  • Eval / runtime import helpers
  • Artifact, patch, manifest, and repo-memory utilities
  • MCP gateway and related setup scripts

• Fixes [eslint-monster] actions/setup/js: normalize caught-error logging to getErrorMessage #49667

---

Generated by 👨‍🍳 PR Sous Chef · gpt54 · 10.4 AIC · ⌖ 6.45 AIC · ⊞ 8.2K · ◷
_{Comment /souschef to run again}

---

Generated by 👨‍🍳 PR Sous Chef · gpt54 · 25.2 AIC · ⌖ 10.7 AIC · ⊞ 8.2K · ◷
_{Comment /souschef to run again}

[Copilot]

Pull request overview

Normalizes caught-error formatting across actions/setup/js using the shared getErrorMessage() helper.

Changes:

• Replaces String(err) and String(error) in 38 setup modules.
• Standardizes filesystem, API, gateway, artifact, and command-routing errors.
• Preserves existing causes, control flow, and scope.

Show a summary per file

File	Description
actions/setup/js/apply_safe_outputs_replay.cjs	Normalizes replay filesystem errors.
actions/setup/js/apply_samples.cjs	Normalizes sample replay failures.
actions/setup/js/artifact_client.cjs	Normalizes artifact directory errors.
actions/setup/js/build_checkout_manifest.cjs	Normalizes manifest filesystem errors.
actions/setup/js/check_workflow_recompile_needed.cjs	Normalizes file-read errors.
actions/setup/js/convert_gateway_config_shared.cjs	Normalizes gateway output errors.
actions/setup/js/create_code_scanning_alert.cjs	Normalizes SARIF write errors.
actions/setup/js/create_pull_request.cjs	Normalizes branch and patch errors.
actions/setup/js/frontmatter_hash_pure.cjs	Normalizes frontmatter read errors.
actions/setup/js/generate_aw_info.cjs	Normalizes info-file errors.
actions/setup/js/generate_git_bundle.cjs	Normalizes bundle directory errors.
actions/setup/js/generate_git_patch.cjs	Normalizes patch filesystem errors.
actions/setup/js/generate_safe_outputs_tools.cjs	Normalizes tool-generation write errors.
actions/setup/js/github_rate_limit_logger.cjs	Normalizes logger directory errors.
actions/setup/js/handle_agent_failure.cjs	Normalizes template-read errors.
actions/setup/js/handle_detection_runs.cjs	Normalizes detection template errors.
actions/setup/js/handle_noop_message.cjs	Normalizes no-op template errors.
actions/setup/js/install_frontmatter_skills.cjs	Normalizes skill installation errors.
actions/setup/js/load_experiment_state_from_repo.cjs	Normalizes experiment-state errors.
actions/setup/js/mcp_scripts_config_loader.cjs	Normalizes MCP config read errors.
actions/setup/js/merge_remote_agent_github_folder.cjs	Normalizes merge and entrypoint errors.
actions/setup/js/messages_core.cjs	Normalizes message-template errors.
actions/setup/js/mount_mcp_as_cli.cjs	Normalizes MCP CLI directory errors.
actions/setup/js/parse_mcp_gateway_log.cjs	Normalizes gateway log errors.
actions/setup/js/patch_awf_chroot_config.cjs	Normalizes chroot config errors.
actions/setup/js/pick_experiment.cjs	Normalizes experiment persistence errors.
actions/setup/js/push_repo_memory.cjs	Normalizes repo-memory read errors.
actions/setup/js/push_to_pull_request_branch.cjs	Normalizes patch-read errors.
actions/setup/js/route_slash_command.cjs	Normalizes routing warnings and failures.
actions/setup/js/run_evals.cjs	Normalizes evaluation filesystem errors.
actions/setup/js/runtime_import.cjs	Normalizes runtime cache errors.
actions/setup/js/safe_outputs_config.cjs	Normalizes output-directory errors.
actions/setup/js/safe_outputs_handlers.cjs	Normalizes handler filesystem errors.
actions/setup/js/safe_outputs_tools_loader.cjs	Normalizes output append errors.
actions/setup/js/setup_comment_memory_files.cjs	Normalizes comment-memory errors.
actions/setup/js/setup_threat_detection.cjs	Normalizes threat-detection errors.
actions/setup/js/start_mcp_gateway.cjs	Normalizes gateway filesystem errors.
actions/setup/js/upload_artifact.cjs	Normalizes artifact staging errors.

Review details

Tip

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

• Files reviewed: 38/38 changed files
• Comments generated: 0
• Review effort level: Balanced

[github-actions]

Great work on normalizing the error-handling patterns across actions/setup/js! 🎯 This addresses a real consistency issue flagged by ESLint, and the scope is well-defined (38 files, single concern).

One thing that would strengthen this PR:

• Add tests for the error paths — the changes refactor caught-error logging in critical paths like artifact handling, safe-outputs setup, and MCP gateway configuration. Consider adding unit tests that exercise these error-handling branches and verify that getErrorMessage() produces the expected output for different error types (filesystem errors, network errors, etc.).

The PR is otherwise ready to go — good description, focused changes, and aligned with the project's error-handling conventions.

Add unit tests for error-handling paths in actions/setup/js modules affected by the getErrorMessage normalization.

Focus on:
1. Filesystem error handling (apply_samples.cjs, build_checkout_manifest.cjs, etc.)
2. Network/API error handling (artifact_client.cjs, create_pull_request.cjs, github_rate_limit_logger.cjs)
3. Eval/import error handling (runtime_import.cjs)
4. MCP gateway error handling (start_mcp_gateway.cjs)

Each test should:
- Trigger an error condition
- Verify getErrorMessage() is called and produces readable output
- Check that error messages include the helper format rather than raw String(err)

Generated by ✅ Contribution Check · auto · 49.9 AIC · ⌖ 5.03 AIC · ⊞ 8.7K · ◷

[gh-aw-bot]

@copilot quick triage for this PR:

• Please investigate this PR and move it toward merge readiness.
• There are no failing compact checks listed right now, but the PR still looks unstabilized; please re-run the pr-finisher skill, refresh the branch if needed, and summarize any remaining maintainer blockers on the current head.
• If all blockers are maintainer-only, say so explicitly in your reply.

Run details: https://github.com/github/gh-aw/actions/runs/30727625239

Generated by 👨‍🍳 PR Sous Chef · gpt54 · 10.4 AIC · ⌖ 6.45 AIC · ⊞ 8.2K · ◷
_{Comment /souschef to run again}

[Copilot]

@copilot quick triage for this PR:

• Please investigate this PR and move it toward merge readiness.
• There are no failing compact checks listed right now, but the PR still looks unstabilized; please ...

Re-ran triage on the current head (29d5b01).

• No unresolved PR review threads or review comments are open against this change.
• The branch is already refreshed with main; I didn’t find another branch-update action to take.
• Local validation: make fmt and make lint pass.
• Local unit validation is not fully green on the current head: TEST_UNIT_RUN_FULL=1 make test-unit fails in pkg/cli/TestAWFailureInvestigatorPrefetchUsesRunLevelFailures, which is checking .github/workflows/aw-failure-investigator.md / pkg/cli content outside this PR’s actions/setup/js scope.
• GitHub-side, the CJS runs for this PR head are action_required with no jobs started, so CI on the refreshed head still needs maintainer-side approval/re-trigger to produce a fresh signal. The linked Sous Chef run itself had no failed jobs.

So: not all remaining blockers are maintainer-only. The remaining non-maintainer blocker I found is the unrelated local pkg/cli unit-test failure on the current head; otherwise this PR has no open in-scope review/thread blockers.

[gh-aw-bot]

@copilot quick triage for this PR:

• No unresolved PR review threads are visible, but your latest update still reports an unrelated local make test-unit failure on the current head. Please re-run the pr-finisher skill, confirm whether that failure is still present, and summarize whether any blockers remain in-scope for this PR versus maintainer-only.
• Branch refresh: GitHub branch update was requested from this run.
• If all remaining blockers are maintainer-only, say so explicitly in your reply.

Run details: https://github.com/github/gh-aw/actions/runs/30731801766

Generated by 👨‍🍳 PR Sous Chef · gpt54 · 25.2 AIC · ⌖ 10.7 AIC · ⊞ 8.2K · ◷
_{Comment /souschef to run again}

[Copilot]

@copilot quick triage for this PR:

• No unresolved PR review threads are visible, but your latest update still reports an unrelated local make test-unit failure on the current head. Please re-run th...

Re-ran triage on the current head (29d5b01).

• I unshallowed /home/runner/work/gh-aw/gh-aw, fetched origin/main, and re-ran make test-unit; the earlier local failure is no longer present. On the current head it exits cleanly (No changed Go files; skipping impacted Go tests.).
• make fmt and make lint also pass on the current head.
• There are still no unresolved PR review threads or review comments in scope for this PR.
• The requested branch refresh did not produce a new branch head; local and remote are both still 29d5b01.
• The referenced Sous Chef run failed in safe_outputs, but the log shows workflow-side branch-update/review-dismissal issues (There are no new commits on the base branch and stale review-dismissal 404s), not an in-scope blocker caused by this PR’s actions/setup/js changes.

So yes: all remaining blockers are maintainer-only. I do not see any remaining in-scope PR blockers on the current head; what remains is GitHub-side workflow/check state and any maintainer re-trigger/approval needed to get fresh CI on this head.

[github-actions]

🔍 Triage Summary

Field	Value
Category	refactor
Risk	low
Priority	medium
Score	40/100 (impact 15 + urgency 10 + quality 15)
Recommended action	batch_review
Batch	eslint-monster

Normalizes error logging via getErrorMessage across setup-js. CI green, reviewer commented (no blocking issues).

Automated triage — see the triage report for full context.

Generated by 🔧 PR Triage Agent · auto · 40.3 AIC · ⌖ 4.01 AIC · ⊞ 8K · ◷