Selenium MCP Server

A Model Context Protocol (MCP) server implementation for Selenium WebDriver automation using Python. This server allows you to control web browsers programmatically through Claude's desktop application.

Features

• Browser automation with Chrome and Firefox support
• Comprehensive set of Selenium WebDriver operations
• Easy integration with Claude desktop application
• Support for various browser interactions:
  • Navigation
  • Element interaction (click, send keys, etc.)
  • Screenshot capture
  • Window and iframe handling
  • Local storage management
  • And much more!

Prerequisites

• Python 3.7+
• Chrome or Firefox browser installed
• Claude desktop application

Installation

1. Clone this repository:

git clone https://github.com/Jyothishkumarav/selenium-mcp-server-python.git
cd selenium-mcp-server-python

2. Install the required dependencies:

pip install -r requirements.txt

3. Install the MCP server in Claude:

mcp install server.py

Usage

1. Start the server:

python server.py

2. In Claude desktop application:
   • Go to Settings > Developer Options
   • You should see "MCP-Selenium" listed as an available server
   • Enable the server to start using Selenium automation features

Available Tools

The server provides the following automation capabilities:

• Browser Management:

  • Start browser sessions
  • Close sessions
  • Switch between windows

• Navigation:

  • Navigate to URLs
  • Refresh pages
  • Wait for page loads

• Element Interaction:

  • Find elements
  • Click elements
  • Send keys
  • Clear fields
  • Double click
  • Right click

• Element Information:

  • Get element text
  • Get element attributes
  • Check element presence
  • Check element visibility
  • Check element selection

• Page Operations:

  • Take screenshots
  • Get page content
  • Scroll pages
  • Manage local storage

• Frame Handling:

  • Switch to iframes
  • Return to default content

Example Usage

Here's a simple example of how to use the server with Claude:

1. Start a browser session
2. Navigate to a website
3. Find and interact with elements
4. Take screenshots
5. Close the session

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

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

Acknowledgments

• Selenium WebDriver
• Model Context Protocol
• Claude Desktop