Explorar skills

実装・commit 後、`acceptance-check` 通過後・PR 作成前に別の AI (Codex CLI または Claude CLI headless) に diff を渡してセカンドオピニオンのコードレビューを得る。backend は環境変数 `CROSS_REVIEW_BACKEND` で `codex` / `claude-self` から選択でき、未指定時は利用可能な CLI を自動検出する。

特定の GitHub issue への実装着手と PR 作成を依頼されたときに使う。issue 番号・URL・会話内で選んだ issue のいずれかを起点に、実装・commit・lint・受け入れ条件チェック・cross-review・PR 作成・CI 確認まで一気通貫で自動進行する。コードを書いてプルリクを出す作業全般が対象で、issue 選定相談・タイトル編集・クローズ操作・PR レビュー単体には使わない。

Analyze pending git changes and write a structured, conventional-commit plan to COMMITS.md so the user can review and stage commits in batches before committing. Use whenever the user wants to group working-tree or staged changes into sensible commits, draft commit messages for a dirty working tree, or prepare a commit plan from a diff.

Audit the latest GitHub Actions workflow runs on the current branch/PR for warnings, errors, failures, deprecation notices, and other log noise — then fix what's fixable in-repo (workflow YAML, configs, source, tests). Hands commit + push back to the developer. Use after CI finishes and before merge, or whenever the user asks about failing/noisy CI — phrasings like "audit the PR", "why did CI fail

Validates Claude Code skills against official best practices from Anthropic documentation. Fetches latest documentation dynamically to ensure current standards. Checks frontmatter, structure, line count, descriptions, references, workflows, and provides actionable recommendations. Use when asked to validate skill, check skill quality, review skill, or audit skill compliance.

Run multiple skills on the same task in parallel and compare outputs side by side. A creative pitch room -- give it a brief, and it finds skills (local + trusted community repos) to work on it independently, then presents all outputs for comparison. Use when asked to "skill battle", "compare skills", "run multiple skills", "pitch room", "get different takes", "split test this", "give me options",

Quick-reference card for shakespeare mode — all intensity levels, commands, and activation phrases. One-shot display, not a persistent mode. Use when the user types /shakespeare-help or asks "what shakespeare commands are there".

Use this skill when implementing features, writing code, creating functions or classes, refactoring existing code, or building any software component. Triggers on: "implement", "write the code for", "build this feature", "create a function", "refactor this", "add this to the codebase", "make this work", "write a class", "implement the logic". This skill enforces clean code principles, proper abstr

LinkedIn image types and generation methods for each post format. Quote cards, article screenshots, brand photos, and selfie content with exact dimensions and tools.

Dynamic Expert Council QA system. Selects domain-relevant expert personas to review deliverables before final delivery. 3 tiers with build/review mode toggle. Works with any AI assistant.

Set up a build/test/lint validation loop for the current project. Detects the tech stack and configures CLAUDE.md with verification commands.

Trace the root cause of a bug or unexpected behaviour. Use when the user says "why is this broken", "why does this fail", "whybroken", "something's wrong with X", or pastes an error. Never patches symptoms — always finds the cause first.