Computer Control

MCP server that provides computer control capabilities, like mouse, keyboard, OCR, etc. using PyAutoGUI, RapidOCR, ONNXRuntime. Similar to 'computer-use' by Anthropic. With Zero External Dependencies.

python

120

GitHub Stars

python

Language

6 months ago

First Indexed

2 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": {
    "ab498-computer-control-mcp": {
      "command": "computer-control-mcp",
      "args": [],
      "env": {
        "COMPUTER_CONTROL_MCP_WGC_PATTERNS": "obs, discord, game, steam",
        "COMPUTER_CONTROL_MCP_SCREENSHOT_DIR": "C:\\\\Users\\\\YourName\\\\Pictures\\\\Screenshots"
      }
    }
  }
}

You can automate computer interactions with this MCP server, which provides mouse and keyboard control, OCR, screenshots, and window management. It runs locally or via a remote MCP client, letting you script repetitive tasks and accessibility workflows across your computer.

How to use

To control your computer through the MCP client, connect to the Computer Control MCP server and call its actions from your automation scripts. You can move the mouse, click, type text at the current cursor, drag and drop, take screenshots with optional OCR, and manage windows. There are also options tailored for GPU-accelerated windows to improve screenshot reliability. If you want to tailor how screenshots are saved or how WGC is applied, you can configure environment variables on startup.

How to install

{
  "mcpServers": {
    "computer-control-mcp": {
      "command": "uvx",
      "args": ["computer-control-mcp@latest"]
    }
  }
}

Alternatively, install the package with Python’s package manager. After installation, start the server by calling the runtime command directly.

pip install computer-control-mcp

computer-control-mcp

Configuration and notes

Environment variables let you customize behavior without changing code. The following options are shown for screenshot handling and GPU-accelerated Windows use:

{
  "mcpServers": {
    "computer-control-mcp": {
      "command": "uvx",
      "args": ["computer-control-mcp@latest"],
      "env": {
        "COMPUTER_CONTROL_MCP_SCREENSHOT_DIR": "C:\\Users\\YourName\\Pictures\\Screenshots",
        "COMPUTER_CONTROL_MCP_WGC_PATTERNS": "obs, discord, game, steam"
      }
    }
  }
}

If you prefer to keep screenshots in your Downloads folder by default, you can omit the environment variable and the server will save to the standard downloads location. You can also enable Windows Graphics Capture (WGC) automatically for specific windows by listing title patterns in COMPUTER_CONTROL_MCP_WGC_PATTERNS. This helps avoid black screens with GPU-accelerated apps.

Available tools

click_screen

Click at specific screen coordinates (x, y) to simulate a mouse click.

move_mouse

Move the mouse cursor to specified coordinates (x, y).

drag_mouse

Drag the mouse from one position to another with an optional duration.

mouse_down

Hold down a mouse button (left, right, or middle).

mouse_up

Release a held mouse button.

type_text

Type the provided text at the current cursor position.

press_key

Press a single keyboard key.

key_down

Hold down a specific keyboard key until released.

key_up

Release a held keyboard key.

press_keys

Press one or more keys, including sequences and combinations.

take_screenshot

Capture the screen or a window, with optional OCR and WGC options.

take_screenshot_with_ocr

Capture and extract text from a screenshot using OCR, with optional coordinate data.

get_screen_size

Return the current screen resolution.

list_windows

List all open windows.

activate_window

Bring a window matching a title pattern to the foreground.

wait_milliseconds

Pause execution for a specified amount of time.

Built by

VeilStrat

AI signals for GTM teams