Add relay-ready task-shape signals

[sungjunlee]

Dispatch Summary

구현 완료했고 커밋했습니다.

커밋: `e036e1a91e9a15de1755de81c0c12cec6ea07a63` (`Add relay-ready task-shape signals`)

변경 내용:
- `score-readiness.js`에 additive `task_shape` 신호 추가
- 강한 task-shape 신호는 direct bypass를 막고 granularity를 낮춰 relay-ready shaping으로 유도
- `probe-readiness.js` JSON/human output에 `task_shape` 추가
- broad foundation / narrow single-subsystem regression 테스트 추가
- 문서에 “신호만 emit, semantic decomposition/leaf 생성은 AI shaping 책임” 경계 명시

검증:
- 커밋된 HEAD에서 실행:
  `node --test tests/relay-ready/scripts/scor

Score Log

• Run: issue-679-20260608002011820-c5b567e2
• Executor: codex
• Branch: issue-679-task-shape

[coderabbitai]

Warning

Review limit reached

@sungjunlee, we couldn't start this review because you've reached your PR review rate limit.

More reviews will be available in 45 minutes and 56 seconds. Learn how PR review limits work.

Your organization has run out of usage credits. Purchase more in the billing tab.

⌛ How to resolve this issue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

We recommend that you space out your commits to avoid hitting the rate limit.

🚦 How do rate limits work?

CodeRabbit enforces hourly rate limits for each developer per organization.

Our paid plans include higher PR review limits than trial, open-source, and free plans. In all cases, reviews become available again over time. During sustained high-volume PR review activity, CodeRabbit may temporarily slow when the next review becomes available.

Please see our Fair Usage Limits Policy for further information.

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: d3e78c0f-b1e3-4bad-82b9-cf0fc11b3cd8

📥 Commits

Reviewing files that changed from the base of the PR and between c50f743 and e036e1a.

📒 Files selected for processing (5)

• skills/relay-ready/references/design-v1.md
• skills/relay-ready/scripts/probe-readiness.js
• skills/relay-ready/scripts/score-readiness.js
• tests/relay-ready/scripts/probe-readiness.test.js
• tests/relay-ready/scripts/score-readiness.test.js

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch issue-679-task-shape

---

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 and usage tips.}

[sungjunlee]

Relay Review

Verdict: LGTM
Summary: Inspection pass. VERIFIED: score-readiness adds deterministic clipped task_shape signals for criteria groups, broad scope language, subsystem count, multi-stage journey, and product/foundation mix; strong signals block bypass by forcing granularity low. VERIFIED: probe-readiness preserves existing envelope fields and adds task_shape to JSON, event payload, and human output. VERIFIED: regression coverage includes a broad foundation-style request that does not bypass and a narrow multi-verb single-subsystem request that avoids decomposition signals. VERIFIED: docs state these are AI shaping signals, not semantic leaf-boundary generation. Callers/imports reviewed: probe-readiness consumes the new score field additively, run-preflight still keys off bypass/signals_summary compatibly, and run-qa-loop consumes scoreReadiness without depending on a closed signal-dimension set.
Contract: PASS
Quality Review: PASS
Quality Execution: PASS
Rounds: 1

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: e036e1a91e

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Strip fenced examples before counting criteria bullets

When a single-leaf issue includes an expected Markdown/output sample inside the Done Criteria, every - line in that fenced block is counted as a criteria bullet here. With eight sample lines this sets manyCriteria.strong, forces task_shape.strong, and blocks the readiness bypass even though the actual criteria may be only a file change plus a test; the rest of the scorer already strips fenced code for probe signals, so this should do the same before counting bullets.

Useful? React with 👍 / 👎.