blackwhite084/playwright-plus-python-mcp

Source on GitHub

A Python-based MCP (Model Context Protocol) server using Playwright for browser automation, suitable for LLM integration.

Category: Code Execution & Automation / MCP Servers
Tags: playwright, python, browser-automation, llm

Features

Resources

• Implements a simple note storage system:
  • Uses a custom note:// URI scheme for accessing individual notes.
  • Each note resource has a name, description, and text/plain mimetype.

Prompts

• Provides a "summarize-notes" prompt:
  • Summarizes all stored notes.
  • Optional style argument to control summary detail level (brief/detailed).
  • Generates a prompt combining all current notes with style preference.

Tools

Implements multiple Playwright-based browser automation tools:

• playwright_navigate: Navigates to a specified URL (creates a new session if none exists).
  • Requires url (string).
• playwright_screenshot: Takes a screenshot of the current page or a specific element.
  • Requires name (string) for screenshot file name.
  • Optional selector (string) for CSS selector; if omitted, takes full-page screenshot.
• playwright_click: Clicks an element using a CSS selector.
  • Requires selector (string).
• playwright_fill: Fills an input field.
  • Requires selector (string) and value (string).
• playwright_evaluate: Executes JavaScript in the browser console.
  • Requires script (string).
• playwright_click_text: Clicks an element by its text content.
  • Requires text (string).
• playwright_get_text_content: Gets the text content of all visible elements.
• playwright_get_html_content: Gets HTML content of a page or element.
  • Requires selector (string).

Configuration & Deployment

• Can be configured for development or published server modes.
• Supports integration with Claude Desktop (macOS/Windows).
• Built and distributed via uv toolchain; can be published to PyPI.
• Debugging recommended with MCP Inspector.

License

• Apache-2.0

Pricing

No pricing information provided (open source project).

Quickstart

• Install using the provided configuration for Claude Desktop or run via command line using uv or uvx.
• Build and publish using uv commands (uv sync, uv build, uv publish).
• Set PyPI credentials via environment variables or command-line flags when publishing.

Development

• 100% Python implementation.
• Contributions and issues managed via GitHub.