Mouseless

What it does\nProvides a powerful bridge for AI agents to control a macOS desktop. It allows agents to see the screen, move the cursor, type text, and manage applications, effectively turning an LLM into a desktop operator.\n\n## Tools\n- screenshot: Capture the primary display.\n- zoom: High-resolution capture of a specific screen region.\n- left_click / right_click / middle_click: Precise mouse clicks with modifiers.\n- mouse_move: Move the cursor to specific coordinates.\n- type: Type text into the currently focused element.\n- key / hold_key: Execute keyboard combinations or hold keys.\n- read_clipboard / write_clipboard: Interface with the macOS clipboard.\n- open_application: Launch or focus apps by name or bundle ID.\n- computer_batch: Execute a sequence of actions in a single call for efficiency.\n\n## Installation\nAdd the following to your claude_desktop_config.json:\n\njson\n{\n "mcpServers": {\n "mouseless": {\n "command": "mouseless"\n }\n }\n}\n\n\n## Supported hosts\n- Claude Desktop