Agent Browser
Enables AI agents to directly control your real Chrome browser with full context including login sessions, cookies, and open tabs. It provides tools for page scanning, JavaScript execution, CDP control, screenshots, and physical mouse/keyboard input
What is Agent Browser?
Agent Browser is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to enables ai agents to directly control your real chrome browser with full context including login sessions, cookies, and open tabs. it provides tools for page scanning, javascript execution, cdp contro...
Enables AI agents to directly control your real Chrome browser with full context including login sessions, cookies, and open tabs. It provides tools for page scanning, JavaScript execution, CDP control, screenshots, and physical mouse/keyboard input
This server falls under the Browser Automation category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Enables AI agents to directly control your real Chrome brows
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx agent-browserConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Agent Browser
Agent Browser MCP connects AI agents to your real, already-running Chrome browser — complete with your existing login sessions, cookies, and open tabs — rather than spinning up an isolated browser instance. It exposes tools for page scanning, JavaScript execution, Chrome DevTools Protocol (CDP) commands, screenshot capture, and physical mouse and keyboard input, enabling agents to interact with web applications exactly as a human would. It is particularly useful for automating workflows that require authenticated sessions or complex UI interactions.
Prerequisites
- Python 3.10 or later on macOS or Windows
- Google Chrome installed
- pip for installing the package
- An MCP-compatible AI client such as Claude Desktop, Cursor, or Hermes Agent
- Developer Mode enabled in Chrome (for loading the companion extension)
Install the agent-browser-mcp package
Clone the repository and install it in editable mode, or build and install the wheel.
git clone https://github.com/335234131/agent-browser-mcp.git
cd agent-browser-mcp
pip install -e .Find the Chrome extension path
After installation, run the helper command to get the path to the bundled Chrome extension directory.
agent-browser-mcp extension-pathLoad the extension into Chrome
Open Chrome, navigate to chrome://extensions, enable Developer Mode, click 'Load unpacked', and select the directory printed by the previous command. Then open any regular webpage (not about:blank) to establish a valid browsing session.
Run the diagnostics check
Verify that the extension is connected and the MCP server can communicate with Chrome.
agent-browser-mcp doctorAdd to your MCP client configuration
Register the agent-browser-mcp server in your Claude Desktop or Cursor configuration file.
{
"mcpServers": {
"agent-browser": {
"command": "agent-browser-mcp",
"args": []
}
}
}Verify available tools
Once connected, the following tools will be available to your AI assistant: get_setup_status, list_tabs, switch_tab, open_url, open_new_tab, scan_page, execute_js, cdp_command, cdp_batch, get_cookies, capture_page_screenshot, capture_desktop_screenshot, mouse_move, mouse_click, mouse_drag, type_text, hotkey, and pointer_info.
Agent Browser Examples
Client configuration
Add agent-browser-mcp to your Claude Desktop configuration to give Claude control of your real Chrome browser.
{
"mcpServers": {
"agent-browser": {
"command": "agent-browser-mcp",
"args": []
}
}
}Prompts to try
Once connected, you can instruct the AI to interact with your browser using your existing sessions.
- "List all my open Chrome tabs and tell me what I have open."
- "Go to https://github.com and take a screenshot of the current page."
- "Click the Login button on the current page and type my username."
- "Execute JavaScript on the current page to get the page title."
- "Get all cookies for the current site."
- "Switch to the tab that has Gmail open and scan the page content."Troubleshooting Agent Browser
agent-browser-mcp doctor reports the extension is not connected
Ensure you loaded the extension from the correct path (use agent-browser-mcp extension-path), Developer Mode is enabled in Chrome, and you have a real webpage open (not about:blank or a new tab page).
Physical mouse/keyboard tools do not work as expected
The tools require Chrome to be the focused application. Ensure no other window is capturing input focus when the agent sends mouse_click or type_text commands.
The server fails to start with a Python version error
agent-browser-mcp requires Python 3.10 or later. Check your version with python --version and upgrade if needed. On macOS, use pyenv or Homebrew to install a newer Python.
Frequently Asked Questions about Agent Browser
What is Agent Browser?
Agent Browser is a Model Context Protocol (MCP) server that enables ai agents to directly control your real chrome browser with full context including login sessions, cookies, and open tabs. it provides tools for page scanning, javascript execution, cdp control, screenshots, and physical mouse/keyboard input It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Agent Browser?
Follow the installation instructions on the Agent Browser GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Agent Browser?
Agent Browser works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Agent Browser free to use?
Yes, Agent Browser is open source and available under the MIT License license. You can use it freely in both personal and commercial projects.
Agent Browser Alternatives — Similar Browser Automation Servers
Looking for alternatives to Agent Browser? Here are other popular browser automation servers you can use with Claude, Cursor, and VS Code.
Chrome DevTools MCP
★ 40.6kAI-powered Chrome automation server with natural language element detection. Control Chrome browser through MCP protocol for testing, debugging, and performance analysis. Features 91% accuracy in element location, works with free AI models, and suppo
UI TARS Desktop
★ 34.9k📇 🏠 - Browser automation capabilities using Puppeteer, both support local and remote browser connection.
Playwright
★ 32.8kA production-ready browser automation server that enables AI assistants to interact with web pages using tools for navigation, element interaction, and data extraction. It features a built-in Inspector UI and robust crash recovery for reliable automa
Page Agent
★ 18.0kJavaScript in-page GUI agent. Control web interfaces with natural language.
Chrome
★ 11.7kAn extension-based MCP server that enables AI assistants to control your browser, leveraging existing sessions and login states for automation and content analysis. It provides over 20 tools for semantic tab search, interactive element manipulation,
LAMDA
★ 7.8kThe most powerful Android RPA agent framework, next generation mobile automation.
Browse More Browser Automation MCP Servers
Explore all browser automation servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Agent Browser in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Agent Browser?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.