browser_ctl
Browser CTL MCP Server is an SSE Anthropic based MCP server built on Python Playwright, enabling AI agents to control and interact with web browsers using Chrome.
Top Comments
The Browser CTL MCP Server is designed to facilitate the interaction of AI agents with web browsers, specifically using the Chrome browser. Built on Python Playwright, this server provides a suite of tools that allow for real-time browser control, including executing JavaScript, navigating to URLs, and managing browser states such as opening, closing, and taking screenshots. The server operates with a default info log level and is accessible via a specified endpoint. It is designed to be easily set up and run on both Linux and Windows platforms, with detailed installation instructions provided.
Features
- Real-time JavaScript execution in the browser using Playwright's evaluate method.
- Ability to navigate to specified URLs in real-time.
- Control over browser state with open and close functionalities.
- Capability to capture and return screenshots of the current browser page.
- Randomized sleep function to pause browser activity for a specified duration.
MCP Tools
- execute_javascript: Execute JavaScript code in real-time browser using Playwright's evaluate method and return the script's result.
- go_to_url: Visit a target URL in real-time browser.
- sleep_for: Browser sleep for a random seconds between begin and end.
- close_browser: Close the opened browser.
- open_browser: Open browser in real-time.
- take_browser_screenshot: Capture a screenshot of the current browser page and return a base64-encoded image.
Usage with Different Platforms
Linux
bash
curl -LsSf https://astral.sh/uv/install.sh | sh
Windows
powershell
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"
Clone the repo
bash
git clone https://github.com/biteval/browser_ctl.git
Go to the project dir
bash
cd browser_ctl
Create virtual environment and activate it
bash
uv init
uv venv
Activate virtual environment on Linux
bash
source .venv/bin/activate
Activate virtual environment on Windows
bash
.venv\Scripts\activate
Install dependencies on Linux
bash
uv add "mcp[cli]" httpx playwright
Install dependencies on Windows
bash
uv add mcp[cli] httpx playwright
Download playwright browser binaries
bash
python -m playwright install
Start the Browser CTL MCP Server
bash
python browser_ctl.py
Related MCP Servers
View all browser_automation servers →Fetch
by modelcontextprotocol
Fetch MCP Server is designed to help language models retrieve web content by converting HTML to markdown for easier consumption. It includes features like content truncation, chunk reading, and customizable user-agent settings, making it highly adaptable for various web scraping tasks.