PyMCP AutoGUI

Example Use Cases

• 1

  Perform desktop automation via mouse and keyboard actions (move, click, drag, write, hotkeys).

• 2

  Perceive the screen and locate images or capture screenshots for vision-based workflows or testing.

• 3

  Manage application windows (activate, move, resize, minimize/maximize) and interact with users via dialogs.

Description

PyMCPAutoGUI exposes pyautogui and pygetwindow functionality over an MCP-compatible server so agents can see the screen, click/type, take screenshots, locate images, and manage windows. It integrates easily with MCP clients like Cursor and runs as a simple Python module (python -m pymcpautogui.server). Useful for automating repetitive GUI tasks, end-to-end GUI testing, and enabling agent-driven desktop workflows across Windows, macOS, and Linux. The project is open-source (MIT) and documents available tools and usage in the README and server module.