🪟 Windows-MCP

Windows MCP is a lightweight, open-source project that enables seamless integration between AI agents and the Windows operating system. Acting as an MCP server bridges the gap between LLMs and the Windows operating system, allowing agents to perform tasks such as file navigation, application control, UI interaction, QA testing, and more.

🎥 Demos

https://github.com/user-attachments/assets/d0e7ed1d-6189-4de6-838a-5ef8e1cad54e

https://github.com/user-attachments/assets/d2b372dc-8d00-4d71-9677-4c64f5987485

✨ Key Features

• Seamless Windows Integration
  Interacts natively with Windows UI elements, opens apps, controls windows, simulates user input, and more.

• Use Any LLM (Vision Optional) Unlike many automation tools, Windows MCP doesn't rely on any traditional computer vision techniques or specific fine-tuned models; it works with any LLMs, reducing complexity and setup time.

• Rich Toolset for UI Automation 25 tools for keyboard, mouse, clipboard, shell, window management, element search, UI state capture, notifications, process management, OCR, and audio control.

• Command Security Built-in protection against dangerous commands (format, shutdown, rm, etc.), injection attacks (blocked operators and encoded command arguments), working directory restrictions, and command length limits.

• Lightweight & Open-Source Minimal dependencies and easy setup with full source code available under MIT license.

• Customizable & Extendable
  Easily adapt or extend tools to suit your unique automation or AI integration needs.

• Real-Time Interaction
  Typical latency between actions (e.g., from one mouse click to the next) ranges from 1.5 to 2.3 secs, and may slightly vary based on the number of active applications and system load, also the inferencing speed of the llm.

Supported Operating Systems

• Windows 7
• Windows 8, 8.1
• Windows 10
• Windows 11

Installation

Prerequisites

• Python 3.13+
• Anthropic Claude Desktop app or other MCP Clients
• UV (Package Manager) from Astra, install with pip install uv
• DXT (Desktop Extension) from Antropic, install with npm install -g @anthropic-ai/dxt

🏁 Getting Started

PyPI (Any MCP Client)

Install from PyPI:

pip install windows-mcp-server

Then configure your MCP client to run windows-mcp as the command.

Claude Code

claude mcp add windows-mcp -- windows-mcp

Gemini CLI

1. Navigate to %USERPROFILE%/.gemini in File Explorer and open settings.json.

2. Add the windows-mcp config in the settings.json and save it.

{
  "theme": "Default",
  "selectedAuthType": "oauth-personal",
//MCP Server Config
  "mcpServers": {
    "windows-mcp": {
      "command": "uv",
      "args": [
        "--directory",
        "<path to the windows-mcp directory>",
        "run",
        "main.py"
      ]
    }
  }
}

3. Rerun Gemini CLI in terminal. Enjoy 🥳

Claude Desktop

1. Clone the repository.

git clone https://github.com/CursorTouch/Windows-MCP.git
cd Windows-MCP

2. Build Desktop Extension DXT:

npx @anthropic-ai/dxt pack

3. Open Claude Desktop:

Go to Claude Desktop: Settings->Extensions->Install Extension (locate the .dxt file)-> Install

Finally Enjoy 🥳.

For additional Claude Desktop integration troubleshooting, see the MCP documentation. The documentation includes helpful tips for checking logs and resolving common issues.

🛠️MCP Tools

Claude can access the following tools to interact with Windows:

📡 MCP Resources

🔒 Command Security

The Powershell-Tool includes multiple layers of protection:

• Blocked commands: format, shutdown, rm, del, regedit, net, and more
• Blocked arguments: -enc, -encodedcommand, --exec, and other injection-risk flags
• Blocked operators: ; and ` (backtick) to prevent command chaining and injection
• Path restriction: workingDir limited to user home directory and server working directory
• Length limit: Commands capped at 2000 characters

Star History

⚠️Caution

This MCP interacts directly with your Windows operating system to perform actions. Use with caution and avoid deploying it in environments where such risks cannot be tolerated.

📝 Limitations

• Selecting specific sections of the text in a paragraph, as the MCP is relying on a11y tree. (⌛ Working on it.)
• Type-Tool is meant for typing text, not programming in IDE because of it types program as a whole in a file. (⌛ Working on it.)

🪪License

This project is licensed under the MIT License - see the LICENSE file for details.

🤝Contributing

Contributions are welcome! Please see CONTRIBUTING for setup instructions and development guidelines.

Made with ❤️ by Jeomon George

Citation

@software{
  author       = {George, Jeomon},
  title        = {Windows-MCP: Lightweight open-source project for integrating LLM agents with Windows},
  year         = {2024},
  publisher    = {GitHub},
  url={https://github.com/CursorTouch/Windows-MCP}
}