WSLSnapit

WSLSnapit MCP Server lets you capture screenshots and read from the Windows clipboard from within WSL, exposing these capabilities to MCP clients. It is optimized for multi-monitor setups and provides efficient image processing to keep data transfer light while enabling direct responses to AI assistants or saved outputs on disk.

How to use

You interact with the WSLSnapit MCP Server through an MCP client. Start the server as a local, stdio-based MCP endpoint and invoke its tools from your client to capture screenshots or read the Windows clipboard. The available tools include take_screenshot for image capture across monitors or specific windows and read_clipboard for retrieving text or images from the Windows clipboard. Use the client to request the desired capture or clipboard read, and the server will return the results (direct image data or file paths) suitable for your AI workflow.

Common usage patterns include capturing all monitors or targeting a single window by title or process name, and choosing whether to return the image directly to your AI assistant or save it to disk for later use.

How to install

Prerequisites you need before installing and running the server:

• Windows 10/11 with WSL2 installed
• Node.js 18.0.0 or higher
• npm, yarn, or bun as a package manager
• PowerShell (comes with Windows)

Then follow these concrete steps to install and prepare the MCP endpoint locally:

# Clone the repository
git clone https://github.com/yourusername/WSLSnapit-MCP.git
cd WSLSnapit-MCP

# Install dependencies
npm install
# or
yarn install
# or
bun install

Configuration and running the MCP server

To run the WSLSnapit MCP Server locally under WSL, you use a stdio-based endpoint that executes the Windows-side operations from within the Linux environment. The following command configuration starts the server in WSL and points it to the project directory where index.js resides.

{
  "mcpServers": {
    "wslsnapit": {
      "command": "wsl",
      "args": [
        "--cd",
        "/mnt/h/dev/mcp/wslsnapit-mcp",
        "node",
        "index.js"
      ]
    }
  },
  "envVars": []
}

Security considerations

The WSLSnapit MCP Server is designed to minimize permissions and keep processing local to protect your data. It uses Windows APIs for screen capture and clipboard access, runs Windows-side components in a separate process, and validates inputs to avoid injection. This ensures that clipboard data and captured images are handled securely within your local environment.

Troubleshooting

If you encounter issues, try these common fixes:

• No windows found: ensure the target window is visible and not minimized. Consider using a different or shorter window title string or switch to capturing by process name.
• Large screenshots failing to return: the server compresses images to stay under 1MB. If needed, lower the quality setting.
• Clipboard reading returns empty: verify that content is actually in the clipboard and consider forcing the format to text or image explicitly.
• Path not found errors: confirm the folder path exists and remember that WSL paths use /mnt/<drive-letter>/… when crossing between Windows and WSL.

Notes and tips

You can combine take_screenshot and read_clipboard tools in your MCP client workflows to create seamless automation that captures current screen state and relevant clipboard data for AI processing.

Tools exposed by the server

This server provides the following tools to MCP clients:

• take_screenshot — Capture screenshots from Windows, with options for monitor targeting, window targeting, and output control.
• read_clipboard — Read text or images from the Windows clipboard with optional format specification.

Available tools

take_screenshot

Capture screenshots across monitors or specific windows with options for window title, process name, and output handling (direct return or saved to disk).

read_clipboard

Read text or image content from the Windows clipboard with auto-detection or explicit format control.