Get the FREE Ultimate OpenClaw Setup Guide →

Windows .Net

A .NET-based Windows desktop automation MCP (Model Context Protocol) server that provides AI assistants with the ability to interact with the Windows desktop environment.

Installation
Run this command in your terminal to add the MCP server to Claude Code.
Run in terminal:
Command
claude mcp add --transport stdio shuyu-labs-windows-mcp.net dnx WindowsMCP.Net@ --yes

How to use

This MCP server provides Windows desktop automation capabilities for AI assistants using the MCP (Model Context Protocol). It exposes a suite of desktop operation tools that let an agent launch applications, interact with PowerShell, capture desktop state, manipulate the clipboard, control the mouse and keyboard, manage windows, perform scrolling and dragging, scrape web content, open URLs in a browser, take screenshots, and perform file system operations, among many others. The server is designed to integrate with MCP clients so you can issue structured commands to automate routine desktop tasks, test applications, or perform complex workflows across the Windows environment. Typical workflows include starting an application by name, running PowerShell commands, reading or writing files, capturing the screen for reports, and performing UI interactions such as typing text, clicking buttons, resizing windows, and navigating between apps. The repository lists a broad set of tools organized under Desktop, FileSystem, OCR, UIElement, and SystemControl categories to support end-to-end automation scenarios.

To use the server, configure your MCP client with the WindowsMCP.Net server entry, then start the server (either via a global tool or by running from the source). Once running, your MCP client can send commands that map to these tools, enabling automated desktop automation tasks in a robust and extensible way.

How to install

Prerequisites:

  • Windows operating system
  • .NET 10.0 Runtime or higher (the project requires .NET 10)

Option A: Global tool installation (recommended)

  1. Install the Windows MCP.NET global tool:
dotnet tool install --global WindowsMCP.Net
  1. Run the server (the tool exposes the MCP server for use by clients):
WindowsMCP.Net

Option B: Run from source (development mode)

  1. Clone the repository:
git clone https://github.com/AIDotNet/Windows-MCP.Net.git
cd Windows-MCP.Net
  1. Build the project:
dotnet build
  1. Run the project:
dotnet run --project src/Windows-MCP.Net.csproj

Notes:

  • Ensure .NET 10 is installed before running. If not, download it from the official .NET 10 download page and install.
  • For development workflows, you can also configure MCP within VS Code using the provided example mcp.json configurations in the repository.

Additional notes

Tips and considerations:

  • The MCP server configuration is designed to work with stdio-based communication. The provided example uses a legacy 'dnx' command for global usage; in a modern environment you may switch to running the project with dotnet run or configure a local development setup.
  • If you plan to develop or debug, using the development mode by cloning the repo and running the project directly is recommended as changes take effect without reinstalling.
  • The repository lists a comprehensive set of tools across Desktop, FileSystem, OCR, UIElement, and SystemControl. When integrating with an MCP client, map your intents to these tools (e.g., LaunchTool, PowershellTool, TypeTool, ClickTool, ScreenshotTool, ReadFileTool, etc.).
  • Ensure that any required permissions for automation (e.g., accessibility permissions on Windows) are granted to the running process to allow UI automation and screen capture capabilities.

Related MCP Servers

Sponsor this space

Reach thousands of developers