The unofficial field manual for Claude Code

What it does This repository is a sprawling community reference for Claude Code, Anthropic’s CLI AI coding assistant. It gathers setup recipes, keyboard shortcuts, slash commands, configuration tricks, and advanced workflows into a single navigable document. If the official docs are the brochure, this is the shop manual.

The interesting bit Rather than stopping at installation, the guide maps out Claude Code’s lesser-known control surfaces: Plan Mode, Auto Mode, sub-agents, MCP integrations, hooks, and even a “Dangerous Mode” security warning. The author also wires it into a broader ecosystem—linking out to hundreds of community agent skills, Discord MCP servers, and Mermaid templates—turning a README into a portal.

Key highlights

• Covers the full setup matrix: native installers, npm, Docker, and WSL, plus auth via console or CLI.
• Documents advanced features like worktree isolation, LSP tool integration, remote sessions, and an agent management CLI.
• Includes automation recipes for auto PR review, issue triage, and scripting workflows.
• Explicit security section with best practices and a dedicated “Dangerous Mode” explainer.
• Links to 954+ agent skills and third-party integrations such as DeepSeek.

Verdict Worth bookmarking if you live in Claude Code and want a single page to find that obscure slash command or hidden config flag. If you were hoping for a new open-source tool to install, this is strictly a reading reference—there is no codebase to run beyond the documentation itself.