title: "Quick Start" description: "Get Claude Code running in 10 easy steps" section: "Getting Started" readTime: "11 min" badge: "NEW"

Quick Start: The Developer's Way

Most developers use Claude Code like ChatGPT. Here's the setup that actually works.

You need a Claude Pro, Max, Team, or Enterprise subscription, a Claude Console (API) account, or access through Amazon Bedrock, Google Cloud's Agent Platform, or Microsoft Foundry. The free claude.ai plan does not include Claude Code.

Installation

Choose your method:

Native (Recommended) - macOS, Linux, WSL

curl -fsSL https://claude.ai/install.sh | bash

Windows PowerShell

irm https://claude.ai/install.ps1 | iex

Windows CMD

curl -fsSL https://claude.ai/install.cmd -o install.cmd && install.cmd && del install.cmd

Homebrew / WinGet

brew install --cask claude-code      # stable channel; use claude-code@latest for the latest channel
winget install Anthropic.ClaudeCode

npm Alternative (requires Node.js 22+)

npm install -g @anthropic-ai/claude-code

Native installs auto-update in the background. Homebrew, WinGet, and npm installs don't, so upgrade them with your package manager (or run claude update). Never use sudo npm install -g.

Step 1: Verify Installation

claude --version

You should see a version number followed by (Claude Code), for example 2.1.211 (Claude Code).

For a fuller health check (install, settings validation, auto-update status), run:

claude doctor

Step 2: Authenticate

Start Claude Code in your project. It prompts you to log in on first use:

cd my-project
claude

Follow the browser prompt to authenticate. You can also sign in from the shell with claude auth login (add --console to bill against an Anthropic Console API account instead of a subscription), check status with claude auth status, or switch accounts inside a session with /login.

Step 3: Initialize Your Project

Inside the running session, run:

/init

Claude analyzes your codebase and generates a starting CLAUDE.md with build commands, test instructions, and conventions it discovers. If a CLAUDE.md already exists, /init suggests improvements instead of overwriting it.

Step 4: Create Your CLAUDE.md

This is your agent's constitution. Keep it under 200 lines - longer files consume more context and reduce adherence. Move instructions that only matter for part of the codebase into path-scoped rules in .claude/rules/.

## Tech Stack
- Node.js 20+
- TypeScript
- Next.js 15
- Tailwind CSS v4
 
## Conventions
- Use kebab-case for file names
- Use camelCase for variables/functions
- ESLint runs on pre-commit
 
## Architecture
- /src/app - Next.js App Router
- /src/components - React components
- /src/lib - Utilities and helpers
 
## Security
- Never commit .env files
- API keys in environment only

Put private, per-project preferences in CLAUDE.local.md (and add it to .gitignore). Claude also keeps its own notes across sessions with auto memory - manage both with /memory.

Step 5: Your First Session

claude "Add a login form to the homepage with validation"

This starts an interactive session with that prompt as the first message. Claude Code will:

1. Read your CLAUDE.md and the files it needs (you don't have to add context manually)
2. Ask for permission before actions your current permission mode doesn't allow
3. Write the code, and run tests if you ask it to

Claude doesn't commit on its own - ask it to ("commit my changes with a descriptive message") when you're happy with the result.

Step 6: Review the Changes

Check what was changed:

git diff

Or run /diff inside the session to review the working-tree changes, including Claude's edits.

Step 7: Use Plan Mode for Complex Work

Start a session in plan mode:

claude --permission-mode plan "Refactor the authentication module for OAuth2 support"

Or, inside a session, press Shift+Tab until the status bar shows plan mode on, or prefix a single prompt with /plan:

/plan refactor the authentication module for OAuth2 support

Claude Code researches and shows its plan first, lets you review, then executes once you approve.

Step 8: Parallel Work with /batch

Inside a session, describe one large change:

/batch migrate src/ from JavaScript to TypeScript

/batch is a bundled skill. It researches the codebase, splits the work into 5 to 30 independent units, and shows you a plan. Once approved, it spawns one background subagent per unit in an isolated git worktree; each implements its unit, runs tests, and publishes its change. It needs a git repository.

For long-lived collaborating sessions, see Agent Teams.

Step 9: Manage Checkpoints

Claude Code automatically creates a checkpoint before every prompt you send. To go back, press Esc twice (with an empty prompt) or run:

/rewind

Pick a prompt from the list, then choose Restore code and conversation, Restore conversation, Restore code, Summarize from here, or Summarize up to here. /checkpoint and /undo are aliases of /rewind.

Step 10: Advanced: Headless Mode

For CI/CD pipelines and scripts, use print mode (-p), which runs one query and exits:

claude -p "Run the tests and summarize failures" \
  --allowedTools "Bash(npm test *)" "Read" \
  --output-format json 2>&1 | tee log.txt

Useful flags for automation:

Only use --dangerously-skip-permissions (equivalent to --permission-mode bypassPermissions) in isolated containers or VMs.

Essential Shortcuts

Pre-Session Checklist

Next Steps

• Read Agent Development Kit to unlock advanced features
• Browse Cheatsheet for the commands that matter most
• Check Security for production setup