A Windows MCP tool for AI agents to control a real browser
AI Browser MCP Server lets an AI agent control a real browser window on Windows. It works with tools such as Cursor, , Cline, OpenCode, custom LLM apps, and scripts that connect through HTTP or WebSocket. It provides 217 ready-made s for opening pages, reading page content, filling forms, running JavaScript, watching , taking screenshots, and using CDP debugging.
A person can give a plain instruction such as collecting product names and prices into JSON instead of writing a Playwright or Puppeteer script by hand. Setup is built around downloading the Windows package, running the browser app, checking the at 127.0.0.1:9222, and connecting an MCP client. Cursor and use an mcp_bridge.js bridge, while custom apps can call the local directly.
The tool includes batch calls and sync-wait, so an agent can chain several browser actions in fewer back-and-forth steps, which can lower token use. Repeated tasks can be saved as and run again, and the project is MIT open source.
Key points
- It connects AI agents to a real Windows browser through MCP.
- It offers 217 s for navigation, form filling, page reading, network inspection, screenshots, and CDP debugging.
- Batch calls and sync-wait can reduce token use by cutting down repeated agent turns.
- files let repeated browser jobs be saved and run again.
- It supports Cursor, , Cline, custom apps, HTTP, and WebSocket connections.