One search for your own team's files and the public ecosystem. See how real projects write their agent instructions, and bring the good ideas home.
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
ZhangHanDongSkillZhangHanDong/rust-skills/skills/unsafe-checker/SKILL.md
CRITICAL: Use for unsafe Rust code review and FFI. Triggers on: unsafe, raw pointer, FFI, extern, transmute, *mut, *const, union, #[repr(C)], libc, std::ffi, MaybeUninit, NonNull, SAFETY comment, soundness, undefined behavior, UB, safe wrapper, memory layout, bindgen, cbindgen, CString, CStr, 安全抽象, 裸指针, 外部函数接口, 内存布局, 不安全代码, FFI 绑定, 未定义行为
ZhangHanDongCLAUDE.mdZhangHanDong/harness-engineering-from-cc-to-ai-coding/CLAUDE.md
本项目基于 Claude Code v2.1.88 的逆向工程源码，编写一本关于 AI Agent 驾驭工程（Harness Engineering）的技术书籍（29 章 + 4 附录）。 任务契约和实施计划在 specs/ 目录： 编写任何章节前，先读对应的 spec 文件。大纲在 docs/book-outline.md。 章节文件写在 book/src/partN/chNN.md，mdbook 通过 SUMMARY.md 自动组织目录。 编写本书时，AI agent 必须加载以下 skill： 每章开始写作前，必须声明预算（tech-writer skill 的 Step 0）： 如果某个论点需要第 4 层深度（production 约束讨论），在预算中显式标注。 源码在 restored-src/src/ 下，关键文件索引见 docs/book-outline.md 附录 A。 绝不修改 restored-src/ 下的任何文件。 每章写完后，按 tech-writer skill 的 Phase 3 执行 3-agent 并行 review： 三个 agent 必须在单条消息中并行启动。
ZhangHanDongSkillZhangHanDong/harness-engineering-from-cc-to-ai-coding/.claude/skills/skillify/SKILL.md
Capture this session's repeatable process into a reusable skill. Use when: user says "make this a skill", "capture this workflow", "skillify", "save this as a skill", "turn this into a skill", "create skill from session", "把这个流程变成技能", "提炼成skill", "保存为技能"
Often used withios-cleanplan-tuneios-design-review
Plain text files in a repository that tell a coding agent how the project works: commands to run, conventions to follow and things to avoid. CLAUDE.md, AGENTS.md, cursor rules and skills are the common kinds.
CLAUDE.md is read by Claude Code. AGENTS.md is an open format that Codex, Cursor and other agents read. Many projects keep one and point the other at it.
A folder with a SKILL.md that describes one capability, such as filling PDFs or reviewing code. The agent loads it only when the task calls for it.
Your agents already can, over MCP, limited to the files you're allowed to read. Searching them from this page is coming.
