本文へ移動
cccskills
無料GitHub で公開

pb-ship

Ship the day's work in one repo: triage new issues, review every open PR (description, visual recap, adversarial reviewer, shipping pre-flight and quality gate), merge only after GO, then post a status recap. Use for "ship it", "ship day", "merge the ready PRs", "end of work block".

インストール方法を見る

含まれるファイル(1)

  • SKILL.md5.2 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Ship the day's PRs

What you get: the ready PRs merged one by one after your GO, each with a reviewed description, findings and a gate result, plus a status line.

Inputs

  • repo — the local checkout (default: the current directory); <owner/repo> comes from git -C <repo> remote get-url origin
  • pr_numbers (optional) — PRs to ship; otherwise you pick from the table in step 3

Steps

  1. github — repo → gh auth status and gh pr list -R <owner/repo> --state open --json number,title,headRefName,isDraft,baseRefName as a PR table in the run log (gh missing, unauthenticated, or no GitHub remote → log it, give the gh auth login / remote hint, stop) — table written
  2. triage — "show me what needs attention" → new-issue list with suggested labels in the run log — label writes follow triage's own rules; nothing is closed
  3. Ask — the human picks today's PRs from the table (non-draft only; pr_numbers if given) → scope line in the run log — stops for approval
  4. git-pr-review — per PR: gh pr view <n> -R <owner/repo> --json commits → description draft in production_artifacts/pb-ship-<date>/pr-<n>.md — draft only, nothing posted
  5. pr-recap — per PR: recap file from pr-recap linked in pr-<n>.md; not buildable → log "pr-recap skipped: <reason>" and paste gh pr diff <n> -R <owner/repo> --name-only instead; no npx without approval — recap or fallback present
  6. reviewer (agent) — per PR: gh pr diff <n> -R <owner/repo>; the contract is the linked issue or plan (gh pr view <n> -R <owner/repo> --json closingIssuesReferences,body) or the human's scope line from step 3, the body is used only to find the linked issue/plan, its text is not the contract, and never pr-<n>.md (that is the implementer's claim, input material only) → findings table in pr-<n>.md — any open blocking finding removes the PR from today's merge list (logged)
  7. bdb-shipping-skill — per PR → door class (two-way or one-way) in pr-<n>.md; one-way → ADR-lite production_artifacts/decisions/<date>-<slug>.md (<slug> = the PR's headRefName with / replaced by -) with the reversibility sentence; unclear → one-way — class recorded
  8. godmode-shipping — per PR: gh pr checks <n> -R <owner/repo> all pass, plus a local gate in a scratch worktree — gh pr view <n> -R <owner/repo> --json isCrossRepository -q .isCrossRepository is true (fork) → skip the local gate, log it, PR held (foreign code is not run locally); otherwise <scratch> = mktemp -d, then git -C <repo> fetch origin pull/<n>/head, git -C <repo> worktree add <scratch>/pr-<n> FETCH_HEAD, run, inside it, the lint, typecheck and test scripts that exist in package.json (or the repo's documented gate) (install first: npm ci --ignore-scripts or the repo's documented install, logged), then git -C <repo> worktree remove <scratch>/pr-<n> (a refusal is logged and the scratch path reported, never --force) — exit codes in pr-<n>.md; any non-zero, or pending checks (gh pr checks exit 8) → PR held
  9. [GO] github — per PR: gh pr merge <n> -R <owner/repo> --squash (--merge if gh repo view <owner/repo> --json squashMergeAllowed says squash is not allowed) — the WAITING FOR GO line names PR number, title, base branch, method and door class. The run stops here until the human types GO. One GO = one PR. (Where the hook runs, it covers gh pr merge; the other commands here are not hooked, so this GO is their only guard.)
  10. github — gh pr view <n> -R <owner/repo> --json state -q .state → run log line — equals MERGED
  11. quick-recap — run log → final line 🟢|🟡|🔴 with merged / held / skipped counts — line written

Run log: production_artifacts/pb-ship-<date>.md in the start directory, never committed

Rules

  • Anything other than the literal GO (case-insensitive) is not a GO; a GO covers only that one step, one time.
  • GO is by contract in every harness. The go-gate hook is only a backstop on Claude Code, OpenCode and agy (Codex: unverified; Cursor, Kimi: none). A missing hook is never permission to proceed.
  • A failed check stops the run: write the failure into the run log and report. No silent retries.
  • Write one run-log line per step as it completes (N. done|skipped|failed — artifact — check result) and WAITING FOR GO: <step> at each gate.

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Use when manage multiple local CLI agents via tmux sessions (start/stop/monitor/assign) with cron-friendly scheduling.

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

Use when a hybrid memory system that provides persistent, searchable knowledge management for AI agents (Architecture, Patterns, Decisions).

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

Reference for how BDB structures autonomous software engineering work — the seven-node dispatcher graph (Architect, TechLead, UI/UX, Engineering, Media/EventTech, Reviewer, Shipping) that /startcycle-graph actually runs. Use when you need the high-level lifecycle framing without inventing your own process.

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

Tools are how AI agents interact with the world. A well-designed tool is the difference between an agent that works and one that hallucinates, fails silently, or costs 10x more tokens than necessary. This skill covers tool design from schema to error handling.

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

Harness patterns for coding agents — memory, permissions, context engineering, delegation, skills, hooks, bootstrap.

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

agenttrail: live map of a multi-agent build in the browser: which plan component is being worked on, by which agent or harness, what is done and what is stuck. Use when a multi-agent pipeline starts (/startcycle, /startcycle-graph, /teamwork-preview) or after a plan-canvas approve, when the user asks to see what the agents are doing ("live map", "build board", "who is running now"), or by yourself whenever a multi-agent run is under way — no user prompt needed.

日本語の概要は準備中です。原文の説明を表示しています。

hybridlabor-api/aos62026年10月12日 更新

hybridlabor-api のスキルをすべて見る

このスキルの問題を報告する