MCP-Server-Playwright

Purpose

MCP Server Playwright is a Model Context Protocol (MCP) server designed to provide browser automation capabilities to large language models (LLMs). Built on Playwright, it enables LLMs to interact with web pages, capture screenshots, and execute JavaScript in a real browser environment.

Key Capabilities

According to official information, the server provides the following tools and resources:

• Full browser automation capabilities: navigate to any URL, click page elements (via CSS selectors or text content), hover over elements (via CSS selectors or text content), fill in input fields, select options in SELECT elements (via CSS selectors or text content).
• Screenshot capture: can capture the entire page or specific elements, and access PNG images via the screenshot://<n> resource.
• Console log monitoring: browser console output can be accessed in text format via the console://logs resource, including all console messages.
• Execute JavaScript in the browser context.

Usage

Official installation via Smithery is provided, supporting Claude Desktop and the Cursor editor.

Install to Claude Desktop

Can be installed automatically via Smithery. The installation command uses npx or mcp-get (the specific command content is not provided in the materials). This command will: check operating system compatibility (Windows/macOS), create or update the Claude configuration file, and configure the Playwright server integration. The configuration file is automatically created/updated at:

• Windows: %APPDATA%\Claude\claude_desktop_config.json
• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

Use in Cursor

1. Install the Playwright browser (if not already installed): npx playwright install
2. Use Smithery to install MCP Server Playwright for Cursor: npx -y @smithery/cli install @automatalabs/mcp-server-playwright --client cursor
3. Configuration file setup: If you are not using Claude, the configuration file (claude_desktop_config.json) may not be created automatically. On Windows, you need to create a folder named Claude in %APPDATA% (usually C:\Users<YourName>\AppData\Roaming), and create a claude_desktop_config.json file inside that folder with the content: { "serverPort" : 3456 }. Then follow the remaining steps in the installation section to complete the setup.

After completing the setup, you can use all the browser automation tools provided by MCP Server Playwright in Cursor's AI features, such as web navigation, screenshot capture, and JavaScript execution.

Note: Make sure Node.js is installed and npx is available in the system PATH.

Applicable Scenarios

Suitable for scenarios that require an LLM or AI assistant to directly operate a browser, such as automated web interaction, information extraction, page testing, screenshot recording, etc. It can be integrated into MCP-supporting clients such as Claude Desktop or Cursor.

Limitations

The materials do not provide specific information regarding pricing, platform restrictions, performance, or effectiveness. The project uses the MIT license; for details, please refer to the LICENSE file.