Windows-MCP
A lightweight, open-source MCP server that bridges AI agents and the Windows operating system, for file navigation, application control, UI interaction and QA testing.
Windows-MCP is a lightweight, open-source MCP server that lets AI agents perform tasks such as file navigation, application control, UI interaction and QA testing on Windows 7 through Windows 11, without relying on traditional computer-vision techniques or fine-tuned models -- it works with any LLM.
It provides tools for basic keyboard and mouse operations and for capturing window or UI state, is minimally dependent and fully open source under the MIT license, and can be extended or adapted for other automation or AI integration needs. A DOM mode focuses exclusively on web page content, filtering out browser UI elements for cleaner web automation in Chrome, Edge, and, with a fallback for its accessibility API, Firefox.
Typical latency between actions, such as one mouse click to the next, ranges from 0.2 to 0.5 seconds depending on system load and the inferencing speed of the connected LLM. The project has reported more than two million users through Claude Desktop Extensions, and installs from PyPI or the MCP Registry.