playwright-mcp

## playwright-mcp 🌐 A specialized, Dockerized **Playwright MCP** server designed to act as a robust browser automation backend for AI agents. This implementation is optimized for high-performance containerized environments and supports the full suite of **Model Context Protocol (MCP)** features. It provides agents with the ability to navigate the web, interact with UI elements, and perform deep network traffic analysis. ### 🚀 Key Features * **Pre-configured Docker:** Built on Playwright v1.50.0 Noble, ensuring all Chromium dependencies are ready for production. * **Network Protocol Analysis:** Full support for core, core-navigation, and network capabilities, enabling agents to sniff API traffic, headers, and payloads. * **Cloud-Ready:** Configured to bind to 0.0.0.0 and utilize the PORT environment variable, making it compatible with Cloud Run, Railway, and other container platforms. * **Optimized for Security:** Includes necessary launch arguments (--no-sandbox) to run smoothly in restricted container environments. ### 📂 Configuration: mcp-config.json The server utilizes a specific configuration to target the Playwright binaries within the container and expose the necessary capabilities: ```json { "browser": { "browserName": "chromium", "launchOptions": { "executablePath": "/ms-playwright/chromium-1155/chrome-linux/chrome", "args": ["--no-sandbox", "--disable-setuid-sandbox"] } }, "server": { "host": "0.0.0.0", "allowedHosts": ["*"] }, "capabilities": ["core", "core-navigation", "network"] } ``` ### 🛠 Deployment To build and run the server using Docker: 1. **Build the image:** ```bash docker build -t playwright-mcp . ``` 2. **Run the container:** ```bash docker run -p 3000:3000 -e PORT=3000 playwright-mcp ``` The server will be available at http://0.0.0.0:3000. ### 📄 License This project is intended for research and development. Please ensure compliance with the target websites' Terms of Service when utilizing automated browser tools.