home / mcp / web browser mcp server

Web Browser MCP Server

A Minimum Control Program (MCP) server implementation for web browsing capabilities using BeautifulSoup4

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "blazickjp-web-browser-mcp-server": {
      "command": "uv",
      "args": [
        "tool",
        "run",
        "web-browser-mcp-server"
      ],
      "env": {
        "REQUEST_TIMEOUT": "30"
      }
    }
  }
}

The Web Browser MCP Server lets AI models browse websites, extract content, and understand pages through a simple MCP interface. It enables precise content extraction with CSS selectors, fast async processing, robust error handling, and cross-platform compatibility so you can integrate web data into your AI workflows with confidence.

How to use

You will connect your MCP client to the Web Browser MCP Server and issue browse requests to fetch and extract data from web pages. Start with a basic fetch to retrieve a page, then refine what you pull using CSS selectors to target headlines, main content, or navigation links. The server returns structured results that include titles, links, and the extracted content.

How to install

Prerequisites you need before installation: a working Python environment for the server code, and the ability to run the UV tool for local MCP execution.

Step 1: Set up the runtime and virtual environment.

Step 2: Install the server in development mode to run tests and contribute.

Step 3: Run the server locally so you can connect an MCP client and start exploring web data.

# Step 1: Create a Python virtual environment
uv venv
source .venv/bin/activate

# Step 2: Install with test dependencies (development mode)
uv pip install -e ".[test]"

# Step 3: Install the server for local experimentation
uv tool install web-browser-mcp-server

# Step 4: Run the server (example startup via UV, do not modify paths)
uv run web-browser-mcp-server

Configuration and runtime environment

The server reads its runtime settings from environment variables. The primary variable shown is REQUEST_TIMEOUT, which defines the webpage request timeout in seconds. The default value is 30 seconds.

{
  "mcpServers": {
    "web-browser-mcp-server": {
      "command": "uv",
      "args": [
        "tool",
        "run",
        "web-browser-mcp-server"
      ],
      "env": {
        "REQUEST_TIMEOUT": "30"
      }
    }
  }
}

Available tools

browse_webpage

Browse and extract content from web pages with optional CSS selectors. Returns structured data including titles, content, and links.