Anti-Detect Browser MCP Agent

•
490
•

Enables AI agents to control anti-detect browsers with persistent profiles and real-device fingerprints via Model Context Protocol (MCP) tools.

Browser MCP Agent provides AI models with direct browser automation capabilities using the Model Context Protocol (MCP) powered by anti-detect Chromium builds. Unlike generic headless setups that trigger bot detection flags, this skill integrates kernel-level hardware and browser fingerprinting—including Canvas, WebGL, TLS ClientHello, audio, and font consistency—with persistent profile management. It allows Claude and other AI agents to autonomously navigate, authenticate, fill forms, extract dynamic web content, run JavaScript, and stream real-time sessions without requiring custom Playwright or Puppeteer automation scripts.

主要功能

01Kernel-level anti-detect fingerprinting across Canvas, WebGL, WebGPU, audio, and TLS stacks
02Persistent profiles that preserve authenticated login sessions across multiple runs
03Direct browser control via MCP tools for navigation, clicks, form filling, and DOM extraction
04490 GitHub stars
05Device emulation support including Android mobile client hints, touch events, and viewport sizing
06Live View screen streaming to monitor and debug autonomous agent browser operations in real time

使用场景

01Extracting dynamic web data and rendering JavaScript without triggering headless bot detection
02Enabling AI agents to conduct authenticated research across isolated browser profiles
03Automating multi-step workflows and form submissions on protected web applications