Sponsored by Deepsite.site

CodingBaby-Browser-MCP

Created By
buyitsydney8 months ago
MCP Server allowing AI agents to control Google Chrome via the CodingBaby Extension
Content

CodingBaby-Browser-MCP

Overview

CodingBaby-Browser-MCP is a powerful tool that enables AI assistants like Claude 3.7 Sonnet in Cursor to control Chrome browser for automated tasks. This tool bridges the gap between AI and web browser interaction through a WebSocket-based communication protocol.

Features

  • Browser Automation: Control Chrome browser programmatically to navigate websites, fill forms, and perform clicks
  • Screenshot Capture: Take screenshots of entire pages or specific areas
  • Multi-tab Support: Create, list, select, and close browser tabs
  • Form Interaction: Type text, press keys, and select form elements
  • Batch Commands: Execute multiple browser operations in sequence
  • Viewport Control: Adjust browser window size for responsive testing

Architecture

The project consists of two main components:

  1. MCP Tool Server: A Node.js server that implements the Model Context Protocol (MCP) to communicate with AI assistants in Cursor
  2. Chrome Extension: A browser extension that receives commands from the MCP server and controls the browser

The system uses WebSocket (port 9876 by default) to establish a bidirectional communication channel between the MCP server and the Chrome extension.

Installation

Prerequisites

  • Node.js (v14 or higher)
  • Chrome browser
  • Cursor editor with Claude 3.7 Sonnet

MCP Tool Setup

  1. In Cursor, go to Settings → MCP
  2. Add new global MCP server with the following configuration:
{
  "mcpServers": {
    "CodingBaby-Browser-MCP": {
      "command": "npx",
      "args": ["@sydneyassistent/codingbaby-browser-mcp"]
    }
  }
}

Chrome Extension Setup

  1. Install the CodingBaby Extension from the Chrome Web Store
  2. Enable the extension and ensure it has the necessary permissions

Usage

Once installed, you can ask Claude 3.7 in Cursor to control your browser:

Use the browser to navigate to https://example.com

Available Commands

  • navigate: Open a URL in the browser
  • click: Click on elements
  • type: Enter text in form fields
  • pressKey: Simulate keyboard actions
  • scroll: Scroll in any direction
  • takeScreenshot: Capture browser content
  • wait: Pause execution for specified time
  • setViewport: Change browser window dimensions
  • tabNew, tabList, tabSelect, tabClose: Tab management
  • batch: Execute multiple commands in sequence
  • close: Close the browser session

Development and Debugging

If you've downloaded the source code, you can set up the project for development and debugging purposes.

Debugging Chrome Extension

To load and debug the Chrome extension from source code:

  1. Open Chrome and navigate to chrome://extensions/
  2. Enable "Developer mode" by toggling the switch in the top-right corner
  3. Click "Load unpacked" button
  4. Navigate to the chrome-extension directory in the project and select it
  5. The extension will now be loaded in developer mode
  6. You can view console logs by right-clicking the extension icon, selecting "Inspect" and opening the Console tab
  7. Make changes to the extension code and click the refresh icon on the extension card to apply changes

Debugging MCP Server in Cursor

To use the local MCP server code for debugging:

  1. Clone or download the repository to your local machine
  2. Navigate to the project directory and install dependencies:
    cd Browser-MCP
    npm install
    
  3. In Cursor, go to Settings → MCP
  4. Add a new MCP server with the following configuration:
    {
      "mcpServers": {
        "CodingBaby-Browser-MCP-Dev": {
          "command": "node",
          "args": ["/absolute/path/to/your/Browser-MCP/index.js"]
        }
      }
    }
    
    Replace /absolute/path/to/your/ with the actual path to the downloaded project
  5. Click "Refresh" to load the MCP server
  6. You can now make changes to the MCP server code and restart the server by clicking "Refresh" in Cursor's MCP settings

For debugging, you can:

  • Check the Cursor MCP logs by clicking on the MCP status icon
  • Add console.error() statements to the code for more detailed logging
  • Run the MCP server manually from the terminal for full console output

Troubleshooting

  • Port Conflict: If port 9876 is already in use, the tool will attempt to release it automatically
  • Connection Issues: Ensure the Chrome extension is properly installed and enabled

License

MIT

Recommend Clients
TraeBuild with Free GPT-4.1 & Claude 3.7. Fully MCP-Ready.
Y GuiA web-based graphical interface for AI chat interactions with support for multiple AI models and MCP (Model Context Protocol) servers.
WindsurfThe new purpose-built IDE to harness magic
Visual Studio Code - Open Source ("Code - OSS")Visual Studio Code
Continue⏩ Create, share, and use custom AI code assistants with our open-source IDE extensions and hub of models, rules, prompts, docs, and other building blocks
ChatWiseThe second fastest AI chatbot™
HyperChatHyperChat is a Chat client that strives for openness, utilizing APIs from various LLMs to achieve the best Chat experience, as well as implementing productivity tools through the MCP protocol.
A Sleek AI Assistant & MCP Client5ire is a cross-platform desktop AI assistant, MCP client. It compatible with major service providers, supports local knowledge base and tools via model context protocol servers .
ZedCode at the speed of thought – Zed is a high-performance, multiplayer code editor from the creators of Atom and Tree-sitter.
Refact.aiOpen-source AI Agent for VS Code and JetBrains that autonomously solves coding tasks end-to-end.
CursorThe AI Code Editor
Cline – #1 on OpenRouterAutonomous coding agent right in your IDE, capable of creating/editing files, executing commands, using the browser, and more with your permission every step of the way.
chatmcpChatMCP is an AI chat client implementing the Model Context Protocol (MCP).
DeepChatYour AI Partner on Desktop
Cherry Studio🍒 Cherry Studio is a desktop client that supports for multiple LLM providers.
y-cli 🚀A Tiny Terminal Chat App for AI Models with MCP Client Support
LutraLutra is the first MCP compatible client built for everyone
MCP PlaygroundCall MCP Server Tools Online
VISBOOM
MCP ConnectEnables cloud-based AI services to access local Stdio based MCP servers via HTTP requests
Roo Code (prev. Roo Cline)Roo Code (prev. Roo Cline) gives you a whole dev team of AI agents in your code editor.