ship
Finalizes tests/docs/changelog and merges the PR to make it deployment-ready.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
Finalizes tests/docs/changelog and merges the PR to make it deployment-ready.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Pre-meeting B2B account research. Produces a structured account brief (company overview, recent news, tech stack, decision structure hypothesis, hypothesized pain points, approach strategy). Use before /discovery-prep.
Generates a Discovery meeting plan from an account brief — SPIN questions, MEDDIC checklist, hypothesized pain points with verification questions, objection prep, time-boxed agenda. Run after /account-brief, before the meeting.
Generates post-meeting follow-up assets — customer-facing email draft, internal action list, CRM update fields, account_brief updates, champion status tracking. Run after any sales meeting.
Converts sales-written meeting notes (discovery type) into a PRD draft for PoC building, and updates account_brief.md with newly learned facts. The bridge between sales discovery and engineering PoC. Run after a discovery meeting, before /kickoff.
Generates a closing proposal from demo meeting notes + the actual PoC + account brief. Business-value first; technical specs in appendix. Run after a demo meeting, before contract negotiation.
Analyzes and critiques business viability. Use after /brainstorm and before /prd.
| name | ship |
| description | Finalizes tests/docs/changelog and merges the PR to make it deployment-ready. |
| disable-model-invocation | true |
| allowed-tools | Read, Glob, Grep, Write, Edit, Task, Bash(bash scripts/checkpoint.sh *), Bash(bash scripts/wt_cleanup.sh *), Bash(bash scripts/registry_edit.sh *), Bash(bash scripts/flock_edit.sh *), Bash(bash scripts/worktree.sh *), Bash(python3 scripts/*), Bash(git *), Bash(gh *), Bash(pytest *), Bash(npm *), Bash(bash ${CLAUDE_PLUGIN_ROOT}/scripts/*), Bash(python3 ${CLAUDE_PLUGIN_ROOT}/scripts/*) |
Kit root: ${CLAUDE_PLUGIN_ROOT}
scripts/ dir): prefix every kit script command with it, e.g.
bash <kit-root>/scripts/checkpoint.sh …. Absolute paths also work from worktrees.${…} placeholder above → standalone layout: run commands as written.Run these checks silently at the start. Use results to adapt behavior:
[ -f issues.md ] — if true, this project uses the sprint system. Respect issue numbering and STATUS.md.[ -f docs/sprint_state.md ] — if true and Status shows running, a sprint is active. Be aware of parallel work in worktrees.[ -f docs/prd_digest.md ] — if true, read it for quick project context before starting.gh auth status before any GitHub operation.Every phase has a checkpoint. Run the verification command and check the exit code.
ADVISORY: line (advisory gate): report the gap, self-correct, continue.
Standard prefix:bash scripts/checkpoint.sh
Append --skill <name> --phase <phase> --issue <ID> for the specific check.
checkpoint.sh resolves the main repo root internally, so the command stays
a single prefix-matchable form (safe to allowlist as Bash(bash scripts/checkpoint.sh *)).
Pipeline skills operate in git worktrees to isolate changes from main.
WT="$(bash scripts/wt_setup.sh <branch>)" — creates the
worktree via scripts/worktree.sh create and writes .claude-kit/freeze-dir.txt
inside it in a single step.bash scripts/worktree.sh rootbash scripts/wt_cleanup.sh <branch> — cd's to main root
inside a subshell, then removes the worktree (never leaves CWD dangling).
All file operations happen inside $WT/. Shared files live on main only.Shared files (issues.md, STATUS.md, CHANGELOG.md) are managed on main only.
Always use registry_edit.sh for concurrent-safe writes — it resolves the
main repo root internally and delegates to flock_edit.sh:
bash scripts/registry_edit.sh issues.md -- bash -c '<update command>'
Never commit these files to feature branches. Steps:
CHECKPOINT — MANDATORY — NEVER SKIP Run:
bash scripts/checkpoint.sh --skill ship --phase checks --issue $ARGUMENTSIf exit code ≠ 0: STOP immediately and report the failure. Do NOT proceed.
3.5) Ask documenter subagent to review and update project documentation:
Update shared files via the registry wrapper:
bash scripts/registry_edit.sh STATUS.md -- bash -c '<update command>'
gh pr merge (merge/squash per repo rules) and delete branch.CHECKPOINT — MANDATORY — NEVER SKIP Run:
bash scripts/checkpoint.sh --skill ship --phase merge --issue $ARGUMENTSIf exit code ≠ 0: STOP immediately and report the failure. Do NOT proceed.
wt_cleanup.sh wrapper — it cd's to the main repo root and
removes the worktree in a single subshell, so the caller's CWD is
never left inside a deleted directory:
bash scripts/wt_cleanup.sh <branch>
CHECKPOINT — MANDATORY — NEVER SKIP Run:
bash scripts/checkpoint.sh --skill ship --phase cleanup --issue $ARGUMENTSIf exit code ≠ 0: STOP immediately and report the failure. Do NOT proceed.
git checkout main && git pull
pytest -q --tb=shortnpm testverify_gates.py runs automatically via checkpoint
(e2e-web, e2e-mobile, api, integration, load — as configured in docs/test_plan.md).git revert -m 1 <merge_commit>.CHECKPOINT — MANDATORY — NEVER SKIP Run:
bash scripts/checkpoint.sh --skill ship --phase smoke --issue $ARGUMENTSIf exit code ≠ 0: STOP immediately and report the failure. Do NOT proceed.
Post-ship test gap advisory (optional, non-blocking):
git diff --name-only HEAD~1 HEADsrc/module.py → tests/test_module.pysrc/component.ts → src/component.test.ts or src/component.spec.tsTest coverage gaps detected in shipped code:
- src/auth.py → no corresponding test file
- src/webhook.py → test exists but may lack coverage for new functions
Suggestion: run `/testgen <path>` to generate missing tests.
Review-quality eval advisory (optional, non-blocking — ISSUE-002):
docs/review_notes/$ARGUMENTS.md (or legacy docs/review_notes.md). If neither exists, skip silently.claude -p process (no separate billing; uses the session auth):
python3 scripts/eval_review.py --pr <pr_number> --notes docs/review_notes/$ARGUMENTS.md
docs/review_eval_<pr>.md and prints eval: pass|concerns → <path>. Surface the verdict + any missed Critical/High findings to the user.eval_review.py always exits 0 — on concerns, or if the claude CLI is unavailable (eval skipped: …), report it and continue. It never fails the ship.IMPORTANT: Never commit issues.md, STATUS.md, or CHANGELOG.md to the feature branch.
These are registry files managed only on main. Always use bash scripts/registry_edit.sh <file> -- bash -c '<update command>' — the wrapper resolves the main repo root internally.
gh pr checks <pr> — all CI checks must be green.gh pr merge fails: report the error (e.g., merge conflicts, branch protection rules).git revert -m 1 <merge_commit> to revert the merge.git revert -m 1 <merge_commit_sha> on main.