Sign inSign up

catinbeard/browser-mcp

By catinbeard

•Updated 2 months ago

Playwright-based MCP server for LLM web navigation, content extraction, and browser automation.

Image
Machine learning & AI
0

407

catinbeard/browser-mcp repository overview

⁠Browser MCP Server

An MCP (Model Context Protocol) server for controlling a headless browser via Playwright. It enables LLMs to navigate the web, extract structured content, and perform automated tasks.

⁠Features
  • Headless Automation: Driven by Playwright and Chromium.
  • Content Extraction: Clean text via Readability.js, ARIA-based accessibility snapshots, Pug structural layouts, and link discovery.
  • Interactive: Support for clicks, input filling, and web search (DuckDuckGo).
  • Robust: Pre-configured with uBlock to reduce clutter.
⁠Usage

You can run the server using the pre-built image from Docker Hub:

docker run -p 8000:8000 --rm catinbeard/browser-mcp:latest
⁠License

Licensed under the MIT License⁠.

⁠Source Code

GitHub Repository⁠

Tag summary

Content type

Image

Digest

sha256:e593386c0…

Size

571.5 MB

Last updated

2 months ago

docker pull catinbeard/browser-mcp