revert(ship): restore readiness reviewer over launch task runner

[baleen37]

Summary

• Revert plugins/me/skills/ship/ to the readiness-reviewer form at 5003c42 (parent of feat(ship): rework /ship as launch task runner #635), reproducing the pre-rework SKILL.md and references/ship-checklist.md.
• Drop references/curation-rules.md and references/launch-playbook.md introduced by feat(ship): rework /ship as launch task runner #635.
• Update tests/me/me-specific.bats to expect ship-checklist.md instead of the removed reference files.

Why

#635 reworked /ship into a phased launch task runner (SCOPE → GATE → CLASSIFY → PLAN → EXECUTE → WRAP) with TaskCreate/Update loops. After using it on a real one-line OpenSearch fix, the friction was clear:

1. PLAN preview is unactionable on its own — task titles plus a playbook section ref. Judging whether each task applies requires opening the playbook every time, which defeats the "scannable preview" goal.
2. standard default task set is web/API biased — "feature flag", "user-facing copy", "staging deploy" don't map cleanly onto batch/ETL jobs (env-separation is the flag, alpha cluster is staging). Adding more lanes per domain doesn't scale.
3. SCOPE is rigid — "can't read diff → stop" is too strict for workdir-only or brand-new branches where the change is plainly in git status.
4. EXECUTE confirm protocol is opaque — "user confirm OR provide evidence OR skip with reason" is enforced one task at a time but never surfaced as options to the user, so each task becomes a free-form back-and-forth.

The readiness-reviewer form sidesteps all four: no task creation, no per-class default task lists, no per-task confirm loop. Output is a single Ready / Conditionally ready / Not ready report against four areas, which is what most launches actually need.

If we later want classification or task-tracking back, it should be reintroduced as an explicit add-on, not as the default /ship shape.

Test plan

• bats tests/me/me-specific.bats — me: ship skill exists with required files now checks ship-checklist.md
• Manual: invoke /ship against a small change and confirm the output is the four-section readiness report (Decision / Blocking / Warnings / Readiness by area / Next actions), not a TaskCreate flow

Summary by CodeRabbit

• Documentation
  • Redesigned the shipping review skill to function as a readiness gate rather than interactive task executor.
  • Updated reference documentation with comprehensive criteria for assessing pre-launch, rollout, rollback, and monitoring readiness.
  • Introduced standardized decision language (Ready/Conditionally Ready/Not Ready) for consistent shipping readiness reporting.

[coderabbitai]

Caution

Review failed

The pull request is closed.

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: baee6056-476a-4481-baac-c7f5fa4e61f6

📥 Commits

Reviewing files that changed from the base of the PR and between f1a11f2 and 060fa73.

📒 Files selected for processing (5)

• plugins/me/skills/ship/SKILL.md
• plugins/me/skills/ship/references/curation-rules.md
• plugins/me/skills/ship/references/launch-playbook.md
• plugins/me/skills/ship/references/ship-checklist.md
• tests/me/me-specific.bats

---

📝 Walkthrough

Walkthrough

The /ship skill shifts from an interactive task-execution workflow to a non-executing launch gate. Three reference files are updated: removing curation-rules.md and launch-playbook.md, adding ship-checklist.md. The test suite is updated accordingly.

Changes

Cohort / File(s)	Summary
Ship Skill Specification plugins/me/skills/ship/SKILL.md	Specification restructured from task-executing launch runner (with CLASSIFY, PLAN, EXECUTE phases) to non-executing readiness gate. Removes all tool permissions and task-creation/execution workflow; defines readiness decision criteria (Ready/Conditionally ready/Not ready) and reporting sections for blockers/warnings by area (pre-launch, rollout, rollback, monitoring).
Ship Skill Reference Files plugins/me/skills/ship/references/curation-rules.md, plugins/me/skills/ship/references/launch-playbook.md, plugins/me/skills/ship/references/ship-checklist.md	Deletes two guidance files (curation-rules, launch-playbook) used in classification/planning phases. Introduces new static checklist documenting concrete sectioned criteria for assessing change readiness with example conditions mapping to readiness decisions and standardized decision language.
Test Update tests/me/me-specific.bats	Updates required reference file list for /ship skill test: replaces curation-rules.md and launch-playbook.md with ship-checklist.md.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Possibly related PRs

• refactor(skills): redesign qa verification and tighten ralph wording #619: Introduced/defined the /ship skill and /qa boundary; this PR modifies that skill's documentation and boundary clarifications.
• feat(ship): add shipping readiness skill #618: Touches the same skill files (SKILL.md, ship-checklist.md) and converts the /ship readiness-gate behavior.
• feat(ship): rework /ship as launch task runner #635: Directly opposes these changes—reverses the phased launch-runner behavior by removing the same reference files that this PR adds back or restructures.

Poem

🐰 Hop, hop, the ship is ready—
No more tasks to execute, no phase so steady,
Now we gate and measure readiness with care,
A checklist of checklists, fit and fair! ✨

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch refactor/revert-ship-to-readiness-reviewer

---

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.}