>
Skills(SKILL.md)は、AIエージェント(Claude Code、Cursor、Codexなど)に特定の能力を追加するための設定ファイルです。
詳しく見る →>
Delegate CJK/Chinese content tasks to Gemini CLI. Use this skill for writing Chinese reports, Threads posts (美洲更新), comments, translations, and any task requiring native-quality Traditional Chinese output. Gemini handles the content generation while Claude plans, reviews, and integrates.
内容生产和分发统一管线。素材收集→出稿→排版→封面→朋友圈文案→多平台转换→一键分发。涵盖公众号写作、小红书轮播图、即刻文案、播客音频、品牌视频、Chrome CDP 自动发布。
headless Claude Code (`claude -p`) 1 体と Codex sub-agent 1 体で軽量クロスレビューを行う。修正は行わない。普段使いのレビュー、実装直後の一次確認、review-cycle のデフォルトで使う。
Review Vue 3 code for Composition API, reactivity, components, state (Pinia), routing, and performance. Framework-only atomic skill; output is a findings list.
React/Next.jsのプロジェクトで、UI=計算モデル(コンポーネント/状態/レンダリング)を軸に、設計・実装・レビュー・性能改善の判断を整理する。doc/input/rdd.md に「技術スタック React」または「技術スタック Next.js」があるリポジトリ、あるいはReactの状態管理/レンダリング/Server Components/SSR/Streaming/バンドル/パフォーマンス相談で使う。
設計書に記載された要件を理解し、実装に必要な情報を抽出します。機能名やコンポーネント名を指定すると、関連する設計書を横断的に参照してMVP版の仕様、実装方針、チェックリストを提供します。
Parallel code review using both review-pr and codex-cli-review. Spawns two independent subagents, collects findings, outputs side-by-side reports with consolidated conclusion. Triggers: dual review, double review, full review, review with codex. Not for: single-tool review (use /review-pr or /codex-cli-review), pre-landing diff review (use /review).
Use when provisioning infrastructure, building containers, configuring CI/CD, or deploying services - ensures all infrastructure is codified, versioned, and reviewable with repeatable deployment strategies and proper secrets management | インフラのプロビジョニング、コンテナのビルド、CI/CDの構成、サービスのデプロイ時に使用 - すべてのインフラがコード化、バージョン管理、レビュー可能であることを保証し、再現可能なデプロイ戦略と適切なシークレット管理を実現
Build and review Taiwan stock day-trading plans, pre-market checklists, intraday entry/exit rules, day-risk controls, and `quantitative-trading-decision-system` workflow checks. Use when the user asks about 台股當沖 planning, 盤前準備, 盤中補充標的, VWAP/量能/大盤條件, 停損停利, force-exit timing, or debugging the `當沖` project. Do not use for long-term investing, pure財報研究, or classic波段 setups unless the user explicitly wants an intraday adaptation.
Three-mode translation skill: **quick** for direct translation, **normal** for analysis-informed translation, **refined** for full publication-quality workflow with review and polish.
大家好,我是二哥!
Agent-first stock and market analysis skill for A-shares, Hong Kong stocks, and US stocks. Use when an agent needs structured stock analysis, market review, strategy-based analysis, or reusable command-line execution for equity research. Supports unified request/response contracts, Tushare-backed market data, news-search-assisted workflows, and deterministic scripts.
完全無料ディープリサーチパイプライン。有料API一切不要。WebSearch+WebFetch+opencli-rs(認証不要)+Ollama で全自動リサーチ。トリガー: 「無料リサーチ」「フリーリサーチ」「コストゼロで調査」「APIキーなしでリサーチ」
GitHub Trending查看、项目搜索、Star/Fork/Watch操作、Issues管理、Pull Request操作、代码内容获取、评论管理等完整助手。当用户需要查看GitHub Trending排行榜、搜索GitHub项目、对仓库进行操作、管理Issues/PR、获取代码内容时使用此技能。
本文件為本庫的協作與工程化規範(適用於整個專案目錄)。目標:免費優先(Free-first)的 CI / Security / Review / 文件骨架,並避免動到既有業務邏輯。
这个仓库是中文主导、双语术语的 AI 开发运行协议。
位於 `~/.claude/agents/`:
- `docs/development-policy.md`: コーディング規約、ブランチ戦略、コミットメッセージ規約など
エージェントの使用ルール
位于 `~/.claude/agents/`:
> L2 | 父级: /Users/henry/PARA/[01] Projects/Vibe/ohmymkt/ohmymkt/.claude/AGENTS.md
各エージェントを Agent Tool で Spawn する際のプロンプトテンプレートと、Phase 6 の実装 & PR ワークフロー。
位于 `~/COMBINED/workspace-config/claude/agents/` 中:
- 复杂功能拆解 → planner
位于 `~/COMBINED/workspace-config/claude/agents/`:
ギャルデヴ
impl-orchestrator の Stage 4 で生成する3つの並列レビューエージェントの指示内容。
**A comprehensive guide for AI agents performing code reviews**, organized by priority and impact.
macOSアプリのセキュリティレビュー。Notarization、Hardened Runtime、Sandbox、コード署名をチェック。Use when: macOS、公証、Notarization、Sandbox、署名 を依頼された時。
Intelligent skill router that analyzes user requests and automatically dispatches to the most appropriate skill(s) or zen-mcp tools. Routes to zen-chat for Q&A, zen-thinkdeep for deep problem investigation, codex-code-reviewer for code quality, simple-gemini for standard docs/tests, deep-gemini for deep analysis, or plan-down for planning. Use this skill proactively to interpret all user requests and determine the optimal execution path.
Cleans up and organizes project files. Use when user mentions cleanup, archive, file bloat, or organizing Plans.md/session logs. Do NOT load for: implementation work, reviews, new feature development, or deployment.
Git操作(add, commit, switch, push)とGitHub CLI(PR作成・編集、Issue作成、コメント取得)を実行。コミット作成、ブランチ管理、プルリクエスト作成・編集、Issue管理が必要な場合に使用。「コミットして」「PRを作成」「Issueを作成」「ブランチを切って」などのリクエストで起動。
ローカルサーバーで起動している Marp スライドを視覚的に検証し、レイアウト問題を自動修正するスキル。「スライドをレビュー」「Marp スライドの視覚的チェック」「スライドのレイアウトを確認」「プレゼン資料の見た目を検証」などのリクエストでトリガーします。テキストの切れ目、重なり、配置問題、コントラスト不足、はみ出し、余白不足などの視覚的問題を検出・修正します。
デプロイ済みアプリの作成者レビューを実施する。「レビュー」「改善」「ブラッシュアップ」を依頼されたときに使用する。
Structure the invention into a formal disclosure based on: **$ARGUMENTS**
Assess patentability of: **$ARGUMENTS**
> Override for Codex users who want **Gemini**, not a second Codex agent, to act as the reviewer. Install this package **after** `skills/skills-codex/*`.
Draft a grant proposal based on: **$ARGUMENTS**
Write the patent specification based on: **$ARGUMENTS**
GitHub Copilot CLIを使って実装計画・設計ドキュメントを観点別に並列レビューするスキル。プランファイル、設計書、アーキテクチャドキュメントを複数のサブエージェントがcopilot -pで同時にレビューする。ユーザーが「計画をレビュー」「プランを確認」「設計を見て」「計画の問題点を指摘」などと言ったときに使用する。
セッション開始時にゴール・完了条件・タイムボックスを明確化する。作業セッションの冒頭で方向性を揃えるために使う。
性能・保守性・安全性・アクセシビリティなど非機能要件を体系的にチェックし、明示すべき非機能要件を一覧化する。Use when: WF-02 で非機能要件を確認したい時、性能/セキュリティ/保守性の要件を見落としなく整理したい時。
Commit, push, merge, and complete task with quality review (タスク完了)
pr-create
PR 管理:浏览、摘要、创建、更新、查看 review 评论、回复 review。触发词:'pr'、'pull request'、'创建 PR'、'回复 review'。
Quality assurance system that verifies deliverables match user requirements and execution plans. Features: (1) Auto-Memory - record conversations and tasks, (2) Double Check - verify deliverables in 3 scenarios: plan confirmation, post-delivery verification, user-triggered check, (3) Idle Reflection - periodic review and lesson extraction. Triggers on 'double check it', 'dc it', 'what's wrong', '再检查一下', '到底怎么回事'.
Operate Microsoft Visio through the DrawForge bridge with stable atomic actions. Use when an agent needs to create sessions, add/edit/connect/style shapes, save documents, export PNG previews, or execute a prepared drawing spec without taking over planning decisions.
当用户主动汇报自己完成的具体动作、求反馈或默认求认可时使用。明说型:'我把 leetcode hot100 刷完了'/'我做完了 XX 项目'/'我读完了《XX》'/'我考了 CET-6/雅思/某证书'/'我这周投了 50 家'/'我简历改了 5 版'/'我今天面了 X 家'/'X 我终于弄完了'/'我跑通了 XXX'/'我做了 X 感觉怎么样'/'这样算不算'。隐蔽型(压低期待的汇报,用户自己在克制但实际在求认可):'我好像把 X 搞懂了'/'我试了一下 X'/'我昨天随手做了个 X'/'我不知道这算不算什么,但我…'/'可能不算多,但我…'。边界:用户**已经做完具体动作**在汇报(区别于 coach-concern-direct 的'我够不够'担忧、coach-resume-review 的'帮我看简历'请求、coach-request-deliver 的'帮我梳理'请求)。应用本 skill:第 1 句具体认可他做了什么(不是'真棒'),第 2-3 句把动作从'作业视角'翻译到'求职/能力积累视角'给出可立刻自测的具体锚点,第 4 句根据定位结果动态生成 A/B 选择(深挖/横向/修漏洞/暂停诊断 四选二,不是固定二分)。禁止空洞鼓励、禁止直接打击、禁止无脑加码。
End-to-end source code repository deep dive and review. Use when you need to understand a repo’s architecture, core abstractions, entrypoints and execution flow, extension points, and operational/developer setup; and when you must deliver a structured report (Markdown + standalone HTML) with diagrams (Mermaid), actionable recommendations, and a scored evaluation.