josix/awesome-claude-md

Curated collection of exemplary claude.md files and onboarding patterns from public GitHub projects. Includes analyses, best practices, and templates to help developers create effective AI onboarding documentation for their codebases.

What it solves

When developers want to introduce an AI assistant like Claude into a complex codebase, they need clear, well-structured instructions so the assistant understands the project's architecture, conventions, and workflows. Many projects lack good onboarding documents, leading to misaligned or inefficient AI contributions. This repository addresses that by collecting high-quality CLAUDE.md files from real open-source projects, demonstrating how to write effective instructions for AI assistants.

How it works

The project is a curated collection, not a tool. It organizes examples in a browsable format: you can start with hand-picked Top Picks (categorized as Essential and Advanced Patterns), filter by technology (TypeScript/JavaScript, Python, Rust, Go, Swift, and others) or by use case (AI development, infrastructure, Web3, high-performance computing, getting started). Each entry links to a dedicated page with the example CLAUDE.md and an analysis explaining why it is exemplary. The README also provides quality standards and contribution guidelines, so the community can add more examples. Badges (stars, license, last commit) help you gauge the popularity and maintenance of the source projects.

Who it’s for

This is for developers, team leads, and technical writers who use AI coding assistants (especially Claude) and want to learn how to craft effective CLAUDE.md files for their own projects. It is also useful for anyone curious about how top open-source projects structure their AI onboarding documentation.

Highlights

  • Hand-curated top picks with explanations of what makes each example great, split into Essential and Advanced Patterns.
  • Quick filters to jump directly to examples by language (TypeScript/JS, Python, Rust, Go, Swift, others) or by purpose (AI development, infrastructure, Web3/Blockchain, high-performance computing, getting started).
  • Each example includes an analysis that points out specific techniques and patterns to learn from, not just a bare file.
  • Badges show repository activity (stars, license, last commit) to help you assess the quality and maintenance of the source projects.
  • Contributing guidelines and quality standards are included, encouraging the community to expand the collection with new, high-quality examples.

Related

  • Project
    agentty-org/agentty

    agentty is a native, single‑binary terminal coding assistant written in C++26. It talks to any LLM (Claude, GPT, Ollama, etc.), retrieves only the relevant parts of your repository via a hybrid BM25 + embedding RAG engine, runs shell/build commands inside OS sandboxes, and offers an optional Smart Mode that routes easy tasks to cheap models and complex planning to a flagship model. The tool starts in under 1 ms, has zero runtime dependencies, works on Linux/macOS/Windows/Android, and can operate in an air‑gapped SSH‑relay mode.

  • Project
    agentara/skills

    agentara/skills is a personal collection of installable prompt extensions ("skills") for Claude Code. Each skill is a folder with a `SKILL.md` file that grants Claude Code a new specialized ability — from generating image posters and planning videos, to triaging symptoms, reviewing code, or running scenario simulations. Skills are grouped into five categories (AIGC, Entertainment, Health, Engineering, Productivity). Install via `npx skills add <skill-URL>` or manually by symlinking a skill folder into `~/.agents/skills/` and `~/.claude/skills/`. Most content is MIT-licensed; engineering skills are adapted from Google Engineering Practices (CC-BY 3.0).

  • Project
    wangrongding/frontend-park

    frontend-park 是一个开源的前端趣味知识案例集合，包含基于 TensorFlow.js 的姿态识别、WebRTC 音视频协同、Three.js 示例、隐写术以及千图成像等功能。它为前端开发者提供了可直接运行和修改的趣味实验代码，帮助探索浏览器高级 API。

  • Project
    WoJiSama/skill-based-architecture

    A framework for organizing project rules and instructions into a verifiable "Skill" to ensure coding agents follow business logic and a strict definition of done.