# Quick Start

> ⚡ Not set up yet? One command does it all: `curl -fsSL https://qcode.cc/install/claude-code.sh | bash` (Windows: `irm https://qcode.cc/install/claude-code.ps1 | iex`). See [One-Click Setup Script](/docs/getting-started/one-click-install).

After completing installation and environment configuration, let's master Claude Code's core features in 5 minutes.

## Launch Claude Code

Navigate to your project directory in the terminal and run:

```bash
cd /path/to/your/project
claude
```

You'll see Claude Code's interactive interface, ready to start a conversation.

## Core Interactions

### 1. Understand Code

```
> What's the architecture of this project?
> Explain how src/auth/login.ts works
> Find all places that call the database
```

### 2. Generate Code

```
> Write a user registration API endpoint
> Add unit tests for the UserService class
> Create a React login form component
```

### 3. Edit Existing Code

```
> Refactor utils.js to improve readability
> Fix the memory leak in handleSubmit function
> Add TypeScript types to all public functions
```

### 4. Debug Issues

```
> Why is the user redirected to 404 after login?
> Analyze this error: TypeError: Cannot read property 'id' of undefined
> Investigate why the API response is slow
```

### 5. Git Operations

```
> Show me what files I've modified
> Commit these changes with a good commit message
> Create a PR description
```

## Reference Files with @

Use the `@` symbol to precisely reference specific files, letting Claude read their contents directly:

```
> Review @src/components/Header.tsx and optimize performance
> Compare @src/v1/api.js with @src/v2/api.js
> Analyze project dependencies based on @package.json
```

**Tip**: Press `Tab` after typing `@` to autocomplete file paths.

## Pipe Input

Pass output from other commands to Claude Code for analysis:

```bash
# Explain git changes
git diff | claude "explain these code changes"

# Analyze log errors
cat error.log | claude "analyze these errors and provide solutions"

# Review PR changes
gh pr diff 123 | claude "review this PR for code quality"
```

## Common Slash Commands

Use `/` commands during a session:

| Command | Description |
|---------|-------------|
| `/help` | Show help info and all available commands |
| `/clear` | Clear current conversation history |
| `/compact` | Compress conversation history to save tokens |
| `/model` | View or switch models (e.g., opus, sonnet) |
| `/memory` | Edit project memory file CLAUDE.md |
| `/init` | Auto-generate project configuration |
| `/commit` | Smart commit with generated message |
| `/review` | Code review |
| `/cost` | View current session cost |
| `/doctor` | Diagnose installation and configuration issues |
| `/config` | View or edit configuration |
| `/mcp` | View MCP server status |
| `/loop` | Run Claude Code in autonomous loop mode |
| `/bug` | Report an issue |

## Keyboard Shortcuts

| Shortcut | Function |
|----------|----------|
| `Tab` | Autocomplete file paths |
| `Esc` | Interrupt current output |
| `Ctrl+C` | Cancel current operation |
| `↑` / `↓` | Browse command history |

## Initialize Project Configuration

On first use, run `/init` to let Claude analyze your project and generate configuration:

```
> /init
```

This creates a `CLAUDE.md` file containing:

- Project tech stack and architecture overview

- Common build and test commands

- Code style guidelines

- Project-specific conventions

**Purpose of CLAUDE.md**: Claude Code automatically reads this file on startup, helping the AI better understand your project context.

## Practical Examples

### Example 1: Add a New Feature

```
> I need to add a user avatar upload feature with:
  1. Support for JPG/PNG formats
  2. 2MB file size limit
  3. Auto-crop to 200x200 after upload
  4. Store in S3
```

### Example 2: Fix a Bug

```
> User feedback: Page freezes after clicking "Save" button
> Please investigate the issue in @src/pages/EditProfile.tsx
```

### Example 3: Code Review

```
> /review
> Focus on security issues and performance bottlenecks
```

### Example 4: Smart Commit

```
> /commit
```

Claude will automatically:

1. Analyze all changes

2. Generate a proper commit message

3. Complete the commit

## Efficiency Tips

1. **Be specific**: Clearly describe what you want, including constraints

2. **Break it down**: Split complex tasks into smaller steps

3. **Interrupt early**: Press `Esc` immediately if output isn't what you want

4. **Use references**: Use `@` to specify files instead of letting AI guess

5. **Maintain context**: Claude remembers previous conversation in the same session

## Next Steps

Congratulations on mastering Claude Code essentials! Continue with:

1. Explore [CLI Tips](/docs/usage/cli-tips) for 34 practical tips

2. Learn [Workflow Tips](/docs/usage/workflow-tips) to boost productivity

3. Check out [Codex Integration](/docs/ide/codex) to use another AI tool