This repository is a bilingual teaching guide for studying the Claude Code source code from zero to one.
It is intended for research and learning purposes. The goal is to help readers build a clear mental model of Claude Code's architecture, core modules, execution flow, and source-reading methodology.
The original upstream source snapshot is intentionally excluded from version control in this repository. If you want to follow the guide alongside the real codebase, use your own lawful local copy of the upstream project as the reference source.
books/: bilingual teaching chapters organized by learning pathdocs/: analysis notes, planning documents, and progress trackingconversations/: saved Claude Code working transcripts that show how this guide was planned and written with the spec_driven_develop skill workflow.vitepress/: VitePress site configuration, theme overrides, and navigation logic for the reading website.github/workflows/: GitHub Pages deployment workflow.src/: optional local ignored reference copy for personal study only
The conversations/ directory preserves the original Claude Code transcripts behind this repository. It records how the guide was bootstrapped with the spec_driven_develop skill and then continued through chapter-writing sessions, so readers can inspect not only the final books/ output but also the actual planning and execution process used to produce it.
npm install
npm run docs:dev
npm run docs:buildThe repository now includes a VitePress site and a GitHub Pages workflow, so the Markdown guide can be read as a navigable website instead of only through raw files.
- Claude Code source study
- AI coding agent architecture learning
- Step-by-step teaching material for self-study and discussion
This project is released under the MIT License. See LICENSE.