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/uid13/skills/agents-mdgit clone --depth 1 https://github.com/uid13/skillsWhat 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 | $0.00908 | $0.00908 |
| Opus 5 | $0.00454 | $0.00454 |
| Sonnet 5 | $0.00182 | $0.00182 |
| Haiku 4.5 | $0.00091 | $0.00091 |
Grade A, and why
skills AGENTS.md scanned grade A with 1 finding 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 2d ago.
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.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
- 使用 cross-spawn 进行跨平台进程调用,不要直接 `child_process.spawn` How it starts
The opening of the file, as written. The whole thing — 81 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
本文档面向 AI 编码助手(Claude Code、Codex、OpenCode、Cursor、Copilot 等) 目的是让 AI 在贡献本项目代码时保持风格一致、遵循项目规范。
项目简介
这是一个 AI Agent Skills 的 Monorepo 工程,使用前端工程化方式(Vite 8 + Rolldown + TypeScript)开发多个 Agent 技能, 编译产物符合 Agent Skills 开放标准 和 skills.sh 规范。
技术栈:
- 构建工具:Vite 8(使用 Rolldown,Rust 实现的打包器)
- 语言:TypeScript 5.4+;Vue 3 + UnoCSS(pp 技能使用)
- 运行环境:Node.js 22+
技能类型:
- 代码型技能(music、hq):TypeScript 开发,Vite SSR 编译为
.mjs - 资源型技能(imagegen-magick、wc、wc-jp):纯文档 + 静态页面,Vite publicDir 机制复制
- 网页型技能(pp):Vue 3 + UnoCSS,vite-plugin-singlefile 编译为单文件 HTML
每个技能独立可用,跨平台(Windows / macOS / Linux),零 pnpm install 即可运行。
各技能专属的开发指南见 src/<skill>/AGENTS.md。
核心原则
1. 零依赖分发
- 代码型技能:所有运行时依赖通过 Vite 打包进 dist
- 资源型技能:无代码依赖,仅文档和资源文件
- 网页型技能:所有依赖通过 vite-plugin-singlefile 内联到 HTML
- 用户拿到技能后不需要
pnpm install即可使用 - 仅开发时才需要
pnpm install
2. 跨平台兼容
- 所有脚本必须同时工作于 Windows / macOS / Linux
- 使用 cross-spawn 进行跨平台进程调用,不要直接
child_process.spawn - 文件路径统一使用
node:path
3. 输出 ESM
- 所有编译产物统一为
.mjs(ESM 格式) - Node.js 版本要求 >= 22
- 不发布 source map(.mjs.map)——减小安装体积
编码风格
- 使用 2 空格缩进(不要 Tab)
- 使用 LF 换行(不要 CRLF)
- 单引号字符串
- 结尾无分号
- 使用 ESM(
import/export,不要require)
构建命令速查
# 全局命令(根目录执行,pnpm -r 递归到全部 workspace)
pnpm build # 编译所有技能
pnpm dev # 监听所有技能的改动
pnpm typecheck # 类型检查(含 TS 源码的技能:hq / music / pp)
# 单个技能命令(--filter 指定 workspace 包名)
pnpm --filter imagegen-magick-src build # 编译图像生成技能(资源型)
pnpm --filter wc-src build # 编译英语词汇学习技能(资源型)
pnpm --filter wc-jp-src build # 编译日语词汇学习技能(资源型)
pnpm --filter music-src build # 编译音乐播放技能
pnpm --filter hq-src build # 编译行情查询技能
pnpm --filter pp-src build # 编译图片画廊技能
workspace 定义见 pnpm-workspace.yaml,锁文件为 pnpm-lock.yaml(提交到 git)。
参考资源
- Agent Skills 规范:https://agentskills.io/
- skills.sh 目录:https://skills.sh/
- Vite 8 官方文档:https://vite.dev/
- Vite public 目录:https://vitejs.dev/guide/assets.html#the-public-directory
- Rolldown 文档:https://rolldown.rs/
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.
- 2d ago First seen · 81 lines · 908 tokens per session scan A ec1ad8951b0e
skills AGENTS.md is an instructions file published in the GitHub repository uid13/skills (5 stars, last pushed 6d ago), licensed MIT. It adds 908 tokens to every session, about $0.0045 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
librosa AGENTS.md
Instructions for librosa/librosa, covering agent interaction policy for librosa, 1. automated platform interactions, 2. exception for project-authorized automation and 3. machine-readable directive.
nuclear AGENTS.md
AGENTS.md instructions for nukeop/nuclear, covering agents.md, project overview, packages, commands and development.
Undefined CLAUDE.md
Instructions for 69gg/Undefined, covering claude.md, 项目概述, 开发命令, 依赖安装 / 初始化 and 启动.
lastfm-mcp CLAUDE.md
Instructions for rianvdm/lastfm-mcp, covering claude.md, build / test / lint commands, running tests, code style and formatting (prettier).
skills AGENTS.md
Instructions for scenario-labs/skills, covering agents.md, what this repo is, layout, public content only and authoring contract.
discogs-mcp CLAUDE.md
Instructions for rianvdm/discogs-mcp, covering claude.md, project, commands, architecture and key patterns.