uMouse
Empower AI agents with physical USB HID mouse control and pluggable voice dictation on Windows.
uMouse provides AI agents with direct control over physical mouse hardware and speech dictation capabilities on Windows x64 systems. Moving beyond virtual software-emulated inputs, it interfaces through USB HID to perform accurate clicks, drags, scrolling, and ballistic-compensated pointer movements across multi-monitor workspaces. Alongside physical cursor control, it features an extensible speech-to-text pipeline supporting local engines like Windows Speech and Whisper as well as cloud ASR providers, giving agentic workflows complete desktop interaction capabilities.
主要功能
01Hardware-level mouse control via USB HID for physical clicks, movements, scrolls, and drags
02Adaptive pointer ballistics that dynamically calibrate for precise pixel-level targeting
03Pluggable voice dictation supporting local Windows Speech, Whisper, and cloud ASR providers
04Task exclusivity locks to guarantee dedicated control and prevent conflicting input tasks
05Read-only input sensors to monitor cursor coordinates, key states, and active window contexts
06114 GitHub stars
使用案例
01Autonomous desktop automation requiring genuine physical USB HID input to bypass synthetic event restrictions
02Automated end-to-end GUI and browser testing with verifiable hardware interactions
03Hands-free voice typing and command execution directly into active system windows