What is Mcp Chrome Server

MCP Chrome Server is a browser automation server based on the Model-Controller-Prompt (MCP) architecture, designed to facilitate automated interactions with web browsers using Google Chrome.

Use cases

Use cases for MCP Chrome Server include automating data extraction from websites, testing web applications for functionality and performance, and creating automated workflows for repetitive browser tasks.

How to use

To use MCP Chrome Server, install it using the command ‘npx -y @smithery/cli install @dlwjdtn535/mcp-chrome-server --client claude’. Configure the server based on your operating system (Windows, macOS, or Linux) by setting up the appropriate command and environment variables in a JSON configuration file.

Key features

Key features of MCP Chrome Server include support for multiple operating systems, integration with Google Chrome, and the ability to automate browser tasks efficiently using the MCP architecture.

Where to use

MCP Chrome Server can be used in various fields including web scraping, automated testing of web applications, and any scenario requiring automated browser interactions.

MCP Chrome Server

A Chrome server based on MCP (Model-Controller-Prompt) for browser automation.

Installation

Prerequisites

• Python 3.12
• Google Chrome browser installed
• uv (Python package installer) or Docker

Installing via Smithery

npx -y @smithery/cli install @dlwjdtn535/mcp-chrome-server --client claude

Configuration Setup

Choose one of the following setup methods based on your environment:

1. Using uv (Recommended)

Windows Setup:

{
  "mcpServers": {
    "mcp-chrome-server": {
      "command": "uv",
      "args": [
        "run",
        "--directory",
        "%LOCALAPPDATA%\\Programs\\mcp-chrome-server\\src",
        "mcp-chrome-server"
      ],
      "env": {
        "CHROME_PROFILE_PATH": "%LOCALAPPDATA%\\Google\\Chrome\\User Data"
      }
    }
  }
}

macOS Setup:

{
  "mcpServers": {
    "mcp-chrome-server": {
      "command": "uv",
      "args": [
        "run",
        "--directory",
        "/usr/local/bin/mcp-chrome-server/src",
        "mcp-chrome-server"
      ],
      "env": {
        "CHROME_PROFILE_PATH": "$HOME/Library/Application Support/Google/Chrome"
      }
    }
  }
}

Linux Setup:

{
  "mcpServers": {
    "mcp-chrome-server": {
      "command": "uv",
      "args": [
        "run",
        "--directory",
        "/usr/local/bin/mcp-chrome-server/src",
        "mcp-chrome-server"
      ],
      "env": {
        "CHROME_PROFILE_PATH": "$HOME/.config/google-chrome"
      }
    }
  }
}

Configuration

Chrome Profile Paths

Default Chrome profile paths for each operating system:

Important Notes

• Close all running Chrome instances before starting the automation server
• Ensure proper permissions for the Chrome profile directory
• For Docker setup, make sure the volume mount path matches your system’s Chrome profile path

Credential Management

Securely store and manage login information using the system keychain:

# Save credentials
result = tool_save_credentials(
    site="example.com",
    username="your_username",
    password="your_password"
)

# Retrieve saved credentials
result = tool_get_credentials(
    site="example.com",
    username="your_username"
)

Key Features

Browser Control

# Open browser
result = tool_open_browser()

# Navigate to URL
result = tool_navigate(url="https://example.com")

# Close browser
result = tool_close_browser()

Web Login

result = tool_web_login(
    url="https://example.com/login",
    credentials={
        "username": "your_username",
        "password": "your_password"
    },
    selectors={
        "username": "#id",
        "password": "#pw",
        "submit": ".login-button"
    }
)

Special handling:

• Waits for user to solve CAPTCHA when detected
• Automatic detection of 2-factor authentication
• Detailed analysis of login failure scenarios

Element Manipulation

# Click element
result = tool_click(selector=".button")

# Type text
result = tool_type(
    selector="#input-field",
    text="Hello, World!"
)

# Get text
result = tool_get_text(selector=".content")

# Get multiple elements
result = tool_get_elements(selector=".items")

Important Considerations

1. Chrome Profile Usage

   • Verify correct profile path configuration
   • Close all other Chrome windows using the profile

2. Automation Detection Prevention

   • Simulation of natural user behavior
   • Maintain appropriate delays between login attempts

3. Security

   • Always use system keychain for important credentials
   • Never expose credentials directly in environment variables or configuration files