View Control

Provides local desktop control via MCP for mouse, keyboard, and screen capture using JSON-RPC over stdio.

typescript

0

GitHub Stars

typescript

Language

7 months ago

First Indexed

3 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "nakashima-takeo-view-control-mcp-server": {
      "command": "npx",
      "args": [
        "view-control-mcp-server"
      ]
    }
  }
}

View Control MCP Server lets you control your local PC from an MCP client using JSON-RPC 2.0 over stdio. It enables automated mouse and keyboard actions, screen capture, and other local tasks from an AI assistant. It is useful for integrating AI workflows with direct desktop automation.

How to use

You connect an MCP client to this server using stdio transport. After you initialize the session, you can issue commands to move the mouse, click or type, and capture screenshots. Use the server to perform automated desktop actions driven by your MCP client.

Typical workflows include guiding the cursor to a specific screen position, performing a click sequence, typing text into an active window, and requesting a full-screen capture encoded as image data for verification or processing by your AI client.

How to install

Prerequisites: Node.js and npm must be installed on your machine.

# Install the MCP server package
npm install view-control-mcp-server

Start the server in stdio mode to connect with MCP clients like Cursor or Claude Desktop. The recommended startup command is shown in the usage example.

npx view-control-mcp-server

Additional notes

This server implements the MCP protocol version 2024-11-05 and exposes a range of methods for mouse, keyboard, and screen operations. You can start by initializing the session and then calling available actions as needed.

Available tools

initialize

Initialize the MCP session with protocol version and client info.

ping

Check server liveness.

listOfferings

Retrieve the capabilities offered by the server.

tools/list

List available utility tools exposed by the server.

tools/call

Invoke a specific tool with provided arguments.

mouse.getPosition

Return current mouse cursor coordinates.

mouse.move

Move the mouse to a specified (x, y) position.

mouse.click

Perform a mouse click with a specified button.

keyboard.type

Type a string of text into the focused input.

keyboard.press

Press a single key with optional modifiers.

screen.capture

Capture the current screen and return image data.

server.info

Query server metadata and status.