MCPMonkey

Provides MCP server integration to run and connect MCP tools with a browser extension.

typescript

12

GitHub Stars

typescript

Language

4 months ago

First Indexed

3 weeks ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "kstrikis-mcpmonkey": {
      "command": "npx",
      "args": [
        "mcpmonkey-server"
      ]
    }
  }
}

MCPMonkey extends Violentmonkey to bring Model Context Protocol (MCP) support to your browser automation. It enables AI language models to interact with browser resources through a dedicated MCP server and a browser extension, creating powerful, context-aware automation and scripting workflows.

How to use

You connect the MCPMonkey server to your AI tools and use the browser extension to access powerful MCP tools from your browsing session. The server runs locally and communicates with your MCP client over stdio and WebSocket transports. Once connected, your AI tools can manage tabs, extract page styles, and interact with page content through the extension.

How to install

Prerequisites you need before installing MCPMonkey are Node.js and Yarn. Verify you have a compatible Node.js version and Yarn installed.

Step 1: Install and run the MCPMonkey server using npm/npx or by building from source.

# Option A: Run via npx (quick start)
# This uses a published MCPMonkey server package
npx mcpmonkey-server

# Option B: Build from source and run the built index
# Navigate to the server directory
cd mcpmonkey-server

# Install dependencies
yarn

# Build the server
yarn build

# Run the server from the built bundle
node build/index.js

Additional setup and configuration

You can configure Cursor to connect to the MCPMonkey server. If you run the server in one of the standard ways, Cursor will use the stdio transport to communicate with it. In the Cursor settings, you can add a new MCP server entry with the appropriate command to start the server.

Browser extension installation requires building the project and loading the extension into your browser. Use the development instructions to build and load the extension, then enable it for MCP communications.

Notes on debugging and testing

For development and testing, you can run the server with the MCP inspector for debugging. The internal WebSocket communications operate on port 3025, and you can observe WebSocket activity during development.

Security and usage considerations

Limit access to the MCP server to trusted clients. Use the appropriate permissions controls to restrict what resources each MCP server can access and review browser extension permissions to minimize exposure of sensitive data.

Example workflows

Let your AI tool open a specific set of tabs and extract their styles to inform a design analysis task.

Available tools

browserAction Tool

Get information about all open tabs; create, close, activate, and duplicate tabs to control the browsing session from your MCP-enabled AI.

getPageStyles Tool

Extract page styling information including global styles, computed styles, color schemes, and typography, returning structured JSON for AI analysis.