$ show luongnv89/claude-howto
This repository is a comprehensive educational guide for Claude Code, an AI coding assistant by Anthropic. It provides copy-paste templates, Mermaid architecture diagrams, and progressive learning modules covering all Claude Code features: slash commands, memory, skills, subagents, MCP, hooks, plugins, checkpoints, advanced features, and CLI usage. The repo also includes production-ready build/validation scripts (EPUB builder, static site generator, link checker, Mermaid syntax checker), multi-language translations (Vietnamese, Japanese, Chinese, Ukrainian), CI/CD workflows, and a design system. It is not a software application itself but a knowledge base with supporting tooling. The architecture is a static documentation site with Python helper scripts and pre-commit hooks. Key entry points are the CLAUDE.md project overview and the numbered module directories (01-10). The repository is actively maintained with frequent updates tracking Claude Code releases.
The LLM identified 4 files worth highlighting for reuse:
Standalone Python script that validates markdown cross-references across a documentation repository. Uses regex to find internal links, resolves relative paths, and reports broken references. Reusable in any project that needs to ensure markdown link integrity.
Parameterized EPUB builder that converts markdown documentation into an ebook. Handles chapters, Mermaid diagram rendering (via mmdc CLI), SVG embedding, and internal link conversion. Uses dataclasses for configuration and state, demonstrating a clean builder pattern.
Production-ready Claude Code hook script that auto-formats code on file write using language-specific formatters. Correctly implements the hook exit-code protocol, handles multiple file types, and supports dry-run mode. Designed to be dropped into any Claude Code project with minimal modification.
Configurable code smell detector for Python, JavaScript, and TypeScript. Detects 14 smell types (Long Method, Large Class, Magic Numbers, etc.) with configurable thresholds and severity levels. Uses AST parsing and regex, produces structured output with code snippets. Demonstrates a well-architected analysis pipeline.