Browser MCP

---
name: mcp-browser
description: |
  MCP server for browser automation. Trigger when the user needs to interact
  with web pages, take screenshots, fill forms, or scrape content via a browser.
  Phrases: "browse", "screenshot", "scrape website", "fill form", "open page".
mcpServers:
  browser:
    command: npx
    args:
      - "@anthropic/mcp-server-puppeteer"
---

# Browser MCP

Playwright-powered browser automation through the Model Context Protocol.
Navigate pages, take screenshots, click elements, fill forms, and extract
content from any website.

## Configuration

Add to your Claude Code MCP config (`~/.claude/mcp.json`):

```json
{
  "mcpServers": {
    "browser": {
      "command": "npx",
      "args": ["@anthropic/mcp-server-puppeteer"]
    }
  }
}
```

## Example usage

- "Take a screenshot of example.com"
- "Fill out the contact form on my staging site"
- "Scrape the pricing table from competitor.com"
- "Click the sign-in button and enter my credentials"

## Available tools

- **navigate** - Go to a URL
- **screenshot** - Capture the current page or a specific element
- **click** - Click an element by selector or text
- **fill** - Type text into an input field
- **evaluate** - Run JavaScript in the page context
- **get_text** - Extract all visible text from the page

## Requirements

- Node.js 18+
- npx available in PATH