t18648: GH#34281 route stale-approved external parents past stage 0b to parent close

[marcusquinn]

Summary

Follow-up to #34282. Its stale-approved terminal-close basis lives in stage 4, but the single-pass router in pulse-issue-reconcile.sh returned at stage 0b for every STALE_APPROVAL external issue, so stage 4 was never reached and #34180/#34125 stayed open after deploy. This routes only an unmutated STALE_APPROVAL parent-task issue (no NMR/persistent label) straight from stage 0b to stage 4. It never runs stages 1-3: both parents carry status:available, so a general fall-through would have exposed stage 1. Stage 4 keeps its per-mutation live gates: close needs graph-backed children, and nudges/phase filing/escalation/backfill still require VERIFIED.

Files Changed

.agents/scripts/pulse-issue-reconcile-actions.sh, .agents/scripts/pulse-issue-reconcile.sh, .agents/scripts/tests/test-pulse-issue-reconcile.sh

Runtime Testing

• Risk level: High
• Verification: runtime-verified — Runtime: ran the branch's real _pir_reconcile_issue_stages against live GitHub data for security: enforce local-only processing and stop-and-notify on failed model readiness #34125 and Automate JavaScript worktree tooling readiness in the harness with durable project policy #34180, with all writes intercepted (in-shell gh function plus PATH shim for subprocesses). Both issues went from stage 0b (STALE_APPROVAL) to stage 4, passed live trust re-read, child verification, close-contract, snapshot-stability and graph-evidence gates, and reached exactly one write: 'gh issue close ' (intercepted). No label edits, nudges, comments or stage 1-3 actions were attempted, and both issues stayed untouched (updatedAt unchanged). Tests: test-pulse-issue-reconcile.sh 36/36. The new test_stage0b_routes_stale_approved_parent_to_stage4 runs the real router body and fails without this change (stale-parent:[]). Negative cases (non-parent, NMR, NO_APPROVAL/mutated, MALFORMED) never reach any stage. test-pulse-reconcile-parent-task-subissue-graph.sh 46/46. shellcheck -x clean.

Resolves #34281

---

aidevops.sh v3.39.0 plugin for OpenCode v1.18.35 with claude-opus-5-5 spent 4h 7m and 435,436 tokens on this with the user in an interactive session.

[coderabbitai]

Important

Review skipped

Auto reviews are limited based on label configuration.

🏷️ Required labels (at least one) (1)

---

🚫 Excluded labels (none allowed) (1)

• no-review

---

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

• Configuration used: Repository: marcusquinn/aidevops/.coderabbit.yaml
• Review profile: CHILL
• Plan: Advanced
• Run ID: e945e479-5ff4-49b9-ab13-3823cffa1cc8

---

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

---

_{Comment @coderabbitai help to get the list of available commands.}

[marcusquinn]

Completion Summary

• What: Follow-up to t18648: GH#34281 close approved external parent trackers when sub-issue children complete #34282. Its stale-approved terminal-close basis lives in stage 4, but the single-pass router in pulse-issue-reconcile.sh returned at stage 0b for every STALE_APPROVAL external issue, so stage 4 was never reached and Automate JavaScript worktree tooling readiness in the harness with durable project policy #34180/security: enforce local-only processing and stop-and-notify on failed model readiness #34125 stayed open after deploy. This routes only an unmutated STALE_APPROVAL parent-task issue (no NMR/persistent label) straight from stage 0b to stage 4. It never runs stages 1-3: both parents carry status:available, so a general fall-through would have exposed stage 1. Stage 4 keeps its per-mutation live gates: close needs graph-backed children, and nudges/phase filing/escalation/backfill still require VERIFIED.
• Issue: t18648: Approved external parent-task issues never auto-close after children complete... #34281
• Files changed: .agents/scripts/pulse-issue-reconcile-actions.sh, .agents/scripts/pulse-issue-reconcile.sh, .agents/scripts/tests/test-pulse-issue-reconcile.sh
• Testing: Runtime: ran the branch's real _pir_reconcile_issue_stages against live GitHub data for security: enforce local-only processing and stop-and-notify on failed model readiness #34125 and Automate JavaScript worktree tooling readiness in the harness with durable project policy #34180, with all writes intercepted (in-shell gh function plus PATH shim for subprocesses). Both issues went from stage 0b (STALE_APPROVAL) to stage 4, passed live trust re-read, child verification, close-contract, snapshot-stability and graph-evidence gates, and reached exactly one write: 'gh issue close ' (intercepted). No label edits, nudges, comments or stage 1-3 actions were attempted, and both issues stayed untouched (updatedAt unchanged). Tests: test-pulse-issue-reconcile.sh 36/36. The new test_stage0b_routes_stale_approved_parent_to_stage4 runs the real router body and fails without this change (stale-parent:[]). Negative cases (non-parent, NMR, NO_APPROVAL/mutated, MALFORMED) never reach any stage. test-pulse-reconcile-parent-task-subissue-graph.sh 46/46. shellcheck -x clean.
• Key decisions: Routed at the stage-0b call site instead of changing _action_reconcile_external_issue_gate's return contract, so every other stage stays blocked for stale approvals. Reopened t18648: Approved external parent-task issues never auto-close after children complete... #34281 because its acceptance (parents actually close) was not met by t18648: GH#34281 close approved external parent trackers when sub-issue children complete #34282 alone.

---

aidevops.sh v3.39.0 plugin for OpenCode v1.18.35 with claude-opus-5-5 spent 4h 7m and 435,436 tokens on this with the user in an interactive session.

[codacy-production]

Up to standards ✅

🟢 Issues 0 issues

Results:
0 new issues

View in Codacy

_{NEW Get contextual insights on your PRs based on Codacy's metrics, along with PR and Jira context, without leaving GitHub. Enable AI reviewer
TIP This summary will be updated as you push new changes.}

[github-actions]

Qlty Smell Regression Gate

✅ No change — smell count unchanged.

Metric	Base (8833a8f)	Head (6607a98)	Delta
Total smells	65	65	+0

[github-actions]

Shell Function Complexity Regression Gate

✅ No regression — no new function(s) >100 lines.

Metric	Base (8833a8f)	Head (6607a98)
Total function(s) >100 lines	0	0

[github-actions]

Bash 3.2 Compatibility Regression Gate

✅ No regression — no new bash 3.2-incompatible construct(s).

Metric	Base (8833a8f)	Head (6607a98)
Total bash 3.2-incompatible construct(s)	0	0

[github-actions]

🔍 Code Quality Report

�[0;35m[MONITOR]�[0m Code Review Monitoring Report

SonarCloud: 0 bugs, 0 vulnerabilities, 1 code smells

Sat Oct 10 21:42:24 UTC 2026: Code review monitoring started
Sat Oct 10 21:42:25 UTC 2026: SonarCloud - Bugs: 0, Vulnerabilities: 0, Code Smells: 1
Sat Oct 10 21:44:59 UTC 2026: Codacy analysis completed - issues found (exit 2)

📈 Current Quality Metrics

• BUGS: 0
• CODE SMELLS: 1
• VULNERABILITIES: 0

Generated on: Sat Oct 10 21:45:00 UTC 2026

---

Generated by AI DevOps Framework Code Review Monitoring

[github-actions]

Shell Nesting Depth Regression Gate

✅ No regression — no new file(s) with nesting depth >8.

Metric	Base (8833a8f)	Head (6607a98)
Total file(s) with nesting depth >8	0	0

[github-actions]

File Size Regression Gate

✅ No regression — no new non-README Markdown file(s) >1000 lines at root / >500 elsewhere.

Metric	Base (8833a8f)	Head (6607a98)
Total non-README Markdown file(s) >1000 lines at root / >500 elsewhere	0	0