chore(ci): keep PR #827 repair workflow (still reopenable); fix exact-head-path-policy CI gate

[seonghobae]

Update (2026-08-31): original removal reverted

This PR no longer removes the PR #827 repair workflow. Devin Review flagged that PR #827 remains reopenable — it is closed/merged: false and locked: false, and its head branch fix/opencode-rust-coverage-runtime-boundary-main still exists on origin. Reopening keeps the original PR number, so github.event.pull_request.number == 827 could fire again. The original "GitHub never reuses PR numbers" reasoning below only covers new PRs, not reopening an existing closed one — so the removal was incorrect and has been reverted (byte-identical to pre-removal). See the review-thread discussion on this PR for the full evidence chain.

This PR now carries two small, unrelated fixes instead:

• The CHANGELOG.md/gap-baseline entries documenting why the workflow is being kept, not removed.
• A port of the exact-head-path-policy root-cause fix (scripts/ci/test_strix_quick_gate.sh's awk job-block-boundary bug, tracked in fix(ci): bound required-workflow-bootstrap awk extraction to its own job #1506) onto this branch, needed because that CI check runs per-PR-head rather than from the base branch.

Net diff against main is now small (+14/-1) — see the PR's Files Changed tab for the actual current contents rather than the original description below, which is retained for historical context.

---

Original PR description (historical — see update above)

Hollow-path finding

Org-wide audit finding (responding to "빈 깡통 경로 너무 많다.." — too many empty-can/hollow-shell code paths):

.github/workflows/repair-pr827-coderabbit-comments.yml is gated on:

if: >-
  github.event.pull_request.number == 827 &&
  github.event.pull_request.head.repo.full_name == github.repository &&
  github.event.pull_request.head.ref == 'fix/opencode-rust-coverage-runtime-boundary-main' &&
  github.event.pull_request.head.user.login != 'github-actions[bot]'

Independently re-verified against live GitHub state before touching anything: PR #827 is state: closed, merged: false (confirmed via pull_request_read). GitHub never reuses PR numbers, so github.event.pull_request.number == 827 can never be true again for any future event — the condition is permanently false and the job is dead code.

scripts/ci/repair_pr827_coderabbit_comments.py (460 lines) had exactly one real-world caller: this workflow. Its only other reference in the repo was test_bounded_repair_driver_runs_against_a_staged_fixture in tests/test_materialize_base_python_requirements.py — a synthetic fixture that copies the script into a tmp dir and exercises its internal helper functions purely to keep scripts/ci's coverage number at the required 100%, not because anything live calls it.

Confirmed no other file in the repo references either path (workflow-glob contract tests, PR_GOVERNANCE_AUDIT.md, docs/org-required-workflow-rollout.md, and the doctoring doc it wrote to were all checked — none mention it), and no hardcoded workflow-file-count assertion exists that this deletion would break. This repo already has precedent for banning this exact category of dead PR-specific writer workflow (test_no_pr_specific_writer_workflow_remains forbids repair-pr787*.yml etc.), so removing rather than special-casing this one is consistent with existing convention.

Fix (superseded — see update above)

• Delete .github/workflows/repair-pr827-coderabbit-comments.yml (reverted — kept)
• Delete scripts/ci/repair_pr827_coderabbit_comments.py (reverted — kept)
• Remove the now-purposeless test_bounded_repair_driver_runs_against_a_staged_fixture fixture test (reverted — kept)
• Add a CHANGELOG.md entry under ## [Unreleased]

🤖 Generated with Claude Code

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 691cdb42-2710-408c-87b1-53fe6b0a9c19

---

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.}

[devin-ai-integration]

Devin Review found 1 new potential issue.

[seonghobae]

exact-head-path-policy failure: pre-existing main bug, now fixed on this branch too

This PR's exact-head-path-policy check (Strix Changed Path Quality CI) started failing today, but not because of anything in this PR's own diff.

Root cause: scripts/ci/test_strix_quick_gate.sh's assert_opencode_review_uses_codegraph_and_contextual_orchestrator extracted the required-workflow-bootstrap: job block from .github/workflows/opencode-review.yml with awk '/^ required-workflow-bootstrap:$/,/^[^ ]/'. Every job key in that workflow is indented 2 spaces, never column 0, so the /^[^ ]/ end pattern never matched until EOF — the "block" captured was actually the rest of the entire jobs section. An already-merged, unrelated commit (4a5dfd8, PR #1497) added a step-level if: line inside a later, different job; the broken awk swept it into this job's "block" and the assertion wrongly failed. Reproduced identically against unmodified origin/main before any change.

Trigger-type check: .github/workflows/strix-changed-path-quality-ci.yml's exact-head-path-policy job runs on plain pull_request: (not pull_request_target:) and explicitly checks out github.event.pull_request.head.sha — i.e. it runs this PR's own branch copy of test_strix_quick_gate.sh, not main's. So merging the root-cause fix into main alone would not resolve this PR's check.

Root-cause fix: #1506 (targets main, full write-up and validation there).

What I did here: ported the identical one-line awk fix onto this branch directly (commit 4323e27), validated from a clean isolated clone: bash scripts/ci/test_strix_quick_gate.sh → test_strix_quick_gate: PASS (was FAIL before the fix); coverage run -m pytest tests -q → 2125 passed, 1 skipped, 21 subtests; coverage report --show-missing → 100% on scripts/ci/; interrogate → 100%; bash -n clean.

Note (out of scope for this comment, flagging for visibility): this branch's current head (63c210e, before my port) is a commit titled Revert "Remove permanently-unreachable PR #827 repair workflow", and both .github/workflows/repair-pr827-coderabbit-comments.yml and scripts/ci/repair_pr827_coderabbit_comments.py are present on the branch as of that commit — i.e. the removal this PR's title/description describes does not currently appear to be in effect on the branch. I did not touch that; only the one-line test_strix_quick_gate.sh awk fix was ported here. Worth a look before merge.

A re-run of the failed check should now pass (for exact-head-path-policy specifically).

🤖 Generated with Claude Code

---

Generated by Claude Code

[seonghobae]

Contextual-Orchestrator를 같이 손보든 어쩌든 해결하세요. Bypass merge 필요하면 가능 (chicken and eggs 상황이라면) + NVIDIA NIM 만 쓰는 건 허용하지 않아요. Contextual-Orchestrator를 쓰세요. Timeout은 적어도 3시간으로 잡으세요. 120초 같은 건 당황스럽군요.

[seonghobae]

opencode-review failed on this head (2a77d26) with the same structural race documented on several PRs this session: it polls for a current-head opencode-agent verdict for up to 90 minutes and none landed in that window, so it exited non-zero — not caused by this PR's diff. Re-queued the failed job.

On the NIM/Contextual-Orchestrator/timeout note above: doesn't apply to this PR's scope — this PR only keeps the reopenable PR #827 repair workflow (reverting an incorrect removal) and ports a one-line test_strix_quick_gate.sh awk fix; no NIM calls or timeouts touched here.

---

Generated by Claude Code

[seonghobae]

Contextual-Orchestrator와 관계한 것들을 같이 손보든 어쩌든 해결하세요. Bypass merge 필요하면 가능 (chicken and eggs 상황이라면) + NVIDIA NIM 만 쓰는 건 허용하지 않아요. Contextual-Orchestrator를 쓰세요. Timeout은 적어도 3시간으로 잡으세요. 120초 같은 건 당황스럽군요. Opencode와 Noema 는 Coderabbitai 및 Devin 수준으로 실제로 리뷰를 하게 하시오. Strix도 보안 리뷰를 꼼꼼하게 하도록 하시오. 특히 보안 리뷰는 전체 코드로 수행하는 것입니다. Contextual-Orchestrator는 실시간으로 빠르면서 능력이 좋은 모델에 요청을 보내어 시간을 당기시오. @opencode-agent 라고 부르면 호출되는 기능도 인터넷 가이드에는 /oc 라고 나와있기 때문에 이 점도 확인해 보는 게 좋겠습니다.

[seonghobae]

Superseded after fresh 2026-09-01 control-plane revalidation. The current code delta in this PR is the same required-workflow-bootstrap/pipefail repair carried more completely by #1506, where the fix is factored into a helper and protected by a realistic 20,000-line pipe-buffer regression. The only other unique current delta here records keeping the PR #827 one-shot repair workflow. Fresh main inspection plus #1531 show that workflow still creates a run for every synchronize/reopened/ready_for_review event and only filters PR #827 at job time, while #827 is closed and unmerged. The standing product directive requires purpose-complete self-modifying/source-fix workflows to be removed rather than kept as event-driven dormant machinery. I am therefore converging on #1506 for the CI-gate repair and a separate focused removal of the PR #827 source-fix workflow, rather than preserving this duplicate branch. No predecessor checks/reviews are transferred.