Claude Code#

Anthropic’s official agentic CLI — runs Claude directly in your terminal, reads and edits files, runs commands, manages git, and coordinates multi-step software engineering tasks through natural-language instructions.

What’s in this section#

Page	Topics
Overview	Install, interactive mode, headless scripting, CLAUDE.md project memory
CLI Reference	Every flag, subcommand, env var, output format, and exit code
Slash Commands	Complete built-in command reference, custom commands with arguments
Settings & Permissions	settings.json schema, permission rules, environment variables
Hooks	PreToolUse, PostToolUse, UserPromptSubmit, Notification, Stop, SubagentStop
MCP Servers	Adding and configuring Model Context Protocol servers
Subagents	Task tool, Plan/Explore/general-purpose, worktree isolation, background runs
Skills	SKILL.md authoring, available-skills list, project/user/plugin scope
Statusline	settings.json, JSON context, dynamic statusline scripts
Power-User Tips	CLAUDE.md authoring, custom skills, multi-agent workflows, cost control

Sections#

• Overview — Install, auth, interactive and headless modes, CLAUDE.md memory system.
• CLI Reference — Complete command-line surface in one page.
• Slash Commands — Every built-in slash command with syntax, flags, and examples.
• Settings & Permissions — Configuring settings.json, permission policies, and env vars.
• Hooks — Automating actions at tool lifecycle events.
• MCP Servers — Integrating external tools via the Model Context Protocol.
• Subagents — Spawning isolated agents via the Task tool.
• Skills — Packaging reusable capabilities as loadable skills.
• Statusline — Customising the bottom-of-REPL HUD.
• Power-User Tips — Advanced workflows, custom skills, sub-agents, and cost optimisation.

[!TIP] Start with Overview if you’re new, or jump straight to Settings and Hooks if you’re already up and running.