Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/devcxl/mattpocock-skills-zh/claude-mdgit clone --depth 1 https://github.com/devcxl/mattpocock-skills-zhWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00969 | $0.00969 |
| Opus 5 | $0.00485 | $0.00485 |
| Sonnet 5 | $0.00194 | $0.00194 |
| Haiku 4.5 | $0.00097 | $0.00097 |
Grade A, and why
mattpocock-skills-zh CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
技能按 bucket 文件夹组织在 skills/ 下:
engineering/:日常编码工作productivity/:日常非编码工作流工具misc/:保留但很少使用,不对外推广in-progress/:Beta:刻意公开、欢迎反馈、不随插件发布deprecated/:已不再使用
engineering/ 或 productivity/(已推广的 bucket)中的每个技能必须在顶层 README.md 中有一处引用,并在 .claude-plugin/plugin.json 的 skills 数组中有一项(Claude Code 插件只随附已推广的集合)。misc/、in-progress/ 和 deprecated/ 中的技能不得出现在两者之中。
安装命令从 .agents/install-block.md 逐字复制。.claude-plugin/marketplace.json 让本仓库成为自己的单插件市场:这是安装块解释的回退方案,不是文档化的正式途径。触碰任一清单文件后运行 claude plugin validate . --strict。为什么做 Claude 插件而(暂时)不做 Codex 插件,见 .agents/adr/0002-ship-as-a-claude-code-plugin.md。
顶层 README.md 中的每个技能条目必须把技能名链接到它的 SKILL.md。
每个 bucket 文件夹有一份 README.md,用一句话描述列出该 bucket 中的每个技能,技能名链接到它的 SKILL.md。已推广 bucket 的 README.md 与顶层 README.md 把条目分成 User-invoked(用户调用)与 Model-invoked(模型调用);非推广 bucket 的 README.md(misc/、in-progress/)使用扁平的列表。
engineering/ 与 productivity/ 中的技能还在 docs/<bucket>/<skill-name>.md 有一份面向人类的文档页(文档树镜像 skills/ 下的这两个 bucket 文件夹)。发布的 URL 是 https://aihero.dev/skills-<skill-name>,无论 bucket 如何:文档路径只是仓库组织形式。当你在 engineering/ 或 productivity/ 中新增、重命名或修改某个技能的行为时,按 .agents/writing-docs.md 创建或重新同步其文档页。一份完成的页面包含四个部分:它的作用、何时使用、常见问题、怎样算成功:writing-docs.md 给出模板、小节顺序以及去哪里找问题。非推广 bucket(misc/、in-progress/、deprecated/)中的技能没有文档页。
每份 SKILL.md 要么是用户调用型(disable-model-invocation: true 加上 agents/openai.yaml 中的 policy.allow_implicit_invocation: false,仅人类可达),要么是模型调用型(模型或用户均可达)。见 .agents/invocation.md。
ask-matt 是路由器,映射每一个用户可达的技能以及它们之间的关系。与重新同步文档页相同的触发条件适用于它:当你新增、重命名、移除一个用户可达技能,或修改它在流程中的位置时,重新读一遍 ask-matt 的 SKILL.md 并更新它,以保持地图准确:一个它从未提到的新技能,或一个它仍路由到的过时技能,都是在说谎的路由器。
要把所有技能(重新)链到本地 harness 技能目录(~/.claude/skills、~/.agents/skills),运行 scripts/link-skills.sh。每条都是一个指向本仓库的符号链接,所以 git pull 会保持已安装技能为最新;新增、移除或重命名技能后再次运行该脚本。
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 24 lines · 969 tokens per session scan A 6eac21e78625
mattpocock-skills-zh CLAUDE.md is an instructions file published in the GitHub repository devcxl/mattpocock-skills-zh (289 stars, last pushed 5d ago), licensed MIT. It adds 969 tokens to every session, about $0.0048 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
superpowers-zh AGENTS.md
Instructions for jnMetaCode/superpowers-zh, a project described as: 🦸 AI 编程超能力 · 中文增强版 — superpowers(116k+ ⭐)完整汉化 + 6 个中国原创 skills,让 Claude Code / Copilot CLI / Hermes Agent / Cursor / Windsurf / Kiro / Gemini CLI 等 16 款 AI 编程工具真正会干活.
superpowers-zh GEMINI.md
Instructions for jnMetaCode/superpowers-zh, covering superpowers-zh 中文增强版, 核心规则, 可用 skills and 如何使用.
ai-engineering-from-scratch-zh AGENTS.md
Instructions for fancyboi999/ai-engineering-from-scratch-zh, covering agents.md, 定位, 仓库结构, 强制规则 and 依赖约束.
claude-code-zh-cn AGENTS.md
Instructions for taekchef/claude-code-zh-cn, covering claude-code-zh-cn, 项目结构, 数据流, 技术要点 and 维护流程规则.
mazelingo CLAUDE.md
Instructions for Yeq6X/mazelingo, covering mazelingo, ビルド・開発, 構造 and ルール.
ai-agent-playbook AGENTS.md
Instructions for jukrap/ai-agent-playbook, covering agents.md, working rules, skill structure and verification.