Browser Automation · MODEL CONTEXT PROTOCOL

Puppeteer MCP Server (Python Implementation)

Puppeteer MCP Server (Python Implementation) is a community-listed Model Context Protocol server. Review the source repository before installation.

连接能力

Puppeteer MCP Server (Python Implementation) 为 AI Agent 提供什么

来源目录未提供结构化功能列表,请在使用前阅读源仓库文档。

安全

MCP 收录不等于安全审核

MCP 服务器可能获得模型上下文、凭据、本地文件或调用外部系统的权限。连接 Agent 前,请检查代码、环境变量、网络行为、软件包来源与维护状态。

README

仓库文档

Puppeteer MCP Server (Python Implementation)

A Model Context Protocol server that provides browser automation capabilities using Playwright (Python's equivalent to Puppeteer). This server enables LLMs to interact with web pages, take screenshots, and execute JavaScript in a real browser environment.

Overview

This Python implementation provides a stable alternative to the TypeScript version, offering the same capabilities with improved error handling and logging. It uses Playwright, which is the Python equivalent to Puppeteer, providing robust browser automation capabilities.

Key Features

  • Full browser automation
  • Page navigation
  • Screenshot capture (full page or elements)
  • Form interaction (clicking and filling)
  • JavaScript execution
  • Console log monitoring
  • Configurable timeouts
  • Detailed error handling
  • Comprehensive logging

Prerequisites

  • Python 3.8+
  • pip (Python package installer)

Installation

  1. Install the required packages:
pip install -r requirements.txt
  1. Install Playwright browsers:
playwright install

Usage

Starting the Server

Run the server directly:

python puppeteer_server.py

Claude Desktop Configuration

Add this to your Claude configuration file:

{
  "mcpServers": {
    "puppeteer": {
      "command": "python",
      "args": ["path/to/puppeteer.py"]
    }
  }
}

Available Tools

puppeteer_navigate

Navigate to any URL in the browser.

{
  "name": "puppeteer_navigate",
  "arguments": {
    "url": "https://example.com",
    "timeout": 60000  // optional, defaults to 60000ms
  }
}

puppeteer_screenshot

Capture screenshots of the entire page or specific elements.

{
  "name": "puppeteer_screenshot",
  "arguments": {
    "name": "my_screenshot",
    "selector": "#specific-element",  // optional
    "width": 1280,  // optional, default: 1280
    "height": 720,  // optional, default: 720
    "timeout": 30000  // optional, defaults to 30000ms
  }
}

puppeteer_click

Click elements on the page.

{
  "name": "puppeteer_click",
  "arguments": {
    "selector": ".button-class",
    "timeout": 30000  // optional, defaults to 30000ms
  }
}

puppeteer_fill

Fill out input fields.

{
  "name": "puppeteer_fill",
  "arguments": {
    "selector": "#input-id",
    "value": "text to fill",
    "timeout": 30000  // optional, defaults to 30000ms
  }
}

puppeteer_evaluate

Execute JavaScript in the browser console.

{
  "name": "puppeteer_evaluate",
  "arguments": {
    "script": "document.title",
    "timeout": 30000  // optional, defaults to 30000ms
  }
}

Error Handling

The server provides detailed error messages for common scenarios:

  • Navigation failures
  • Element not found
  • Timeout errors
  • JavaScript execution errors
  • Screenshot failures

Logging

Comprehensive logging is implemented with different levels:

  • INFO: Standard operations
  • ERROR: Operation failures
  • DEBUG: Detailed execution information

Notes

  • Browser launches in non-headless mode for better debugging
  • Default viewport size is 1280x720
  • All timeouts are configurable
  • Console logs are captured and stored
  • Screenshots are stored in memory with base64 encoding

Contributing

Contributions are welcome! Please read the repository's contributing guidelines before submitting pull requests.

Author

Todd Wolven - Lead AI Software Developer and open-source GenAI engineer

License

This project is licensed under the Apache 2.0 License - see the LICENSE file for details.

在 GitHub 查看 README

社区

收藏、评分与讨论

正在载入社区数据…