Everything you need to supercharge your AI-powered development workflow.
CLI tools, MCP servers, IDE plugins, skills, hooks, and more.
Claude Code is built on a modular, extensible architecture. Every layer is customizable.
Built-in capabilities that make Claude Code a powerful agentic coding assistant.
Read, Edit, Write, Bash, Glob, Grep, Agent — a complete toolkit for autonomous coding tasks.
Fine-grained permission controls — allowlist tools, auto-approve patterns, or confirm each action.
Project-level instructions that persist across sessions. Define conventions, rules, and context for your codebase.
Persistent file-based memory across conversations — user preferences, project context, and feedback.
Spawn specialized agents for exploration, planning, code simplification, and parallel task execution.
Run Claude Code non-interactively in CI/CD, scripts, and automation pipelines with --print flag.
Extend Claude Code's capabilities with MCP servers, custom skills, and hooks.
Model Context Protocol servers connect Claude Code to external systems — databases, APIs, browsers, design tools, and more.
Slash commands that teach Claude Code domain-specific workflows. Skills are markdown prompts with tool access.
Skills live in ~/.claude/skills/ or project-level .claude/skills/
Shell commands triggered by Claude Code events — pre/post tool execution, notification, and validation hooks.
Build custom multi-agent systems using Claude Code as the runtime. Orchestrate agents programmatically.
Use Claude Code where you already work — terminal, editor, or IDE.
Native CLI experience. The primary interface.
Inline chat, diff view, and terminal integration.
IntelliJ, WebStorm, PyCharm — all supported.
Terminal-first integration with split panes.
Real-world patterns for leveraging the Claude Code ecosystem.
/commit with conventional message--print headless modeEverything you need to get started and go deeper.
Complete documentation for Claude Code features, configuration, and best practices.
Source code, issue tracker, and community discussions.
Model Context Protocol — the open standard for tool integration.
All commands, flags, and keyboard shortcuts.
Configure event-driven shell commands for validation and automation.
Build custom multi-agent systems with the Claude Agent SDK.
# Install Claude Code
$ npm install -g @anthropic-ai/claude-code
# Start coding
$ cd your-project
$ claude
# Or run headless
$ claude --print "explain this codebase"