Voice, screen and desktop context
Use supported Windows controls, inspect visible context and work with connected applications.

How it fits into your day
A useful desktop assistant needs to know what is available before it acts. Jarvis exposes supported Windows controls alongside application adapters, screen capture and local text recognition, with capability checks for the machine it is actually running on.
Inspect installed capabilities
The desktop status command reports what this machine and its integrations support.
uv run jarvis desktop statusConnect supported applications
Browser and editor adapters add structured selection and command support. Arbitrary application internals are not available just because their windows are visible.
Choose the right vision path
Local OCR needs the Windows recognition language; interpreting a screenshot with the task model requires image support from that provider.
Scope and availability
Device support varies. Brightness, audio controls and window operations depend on the hardware, Windows environment and installed dependencies. Check status rather than assuming universal support.
Computer input is a separate tool. The high-level computer tools can click, type and press keys where available. Those actions still need the correct target and configured permissions; they do not provide deep integration with every app.
A practical first request
You plugged in a device but Windows is still sending sound to the wrong output.
Try: “Show my available audio outputs and switch the default to my headphones.”
A verified Windows default output change where supported. Existing playback streams may keep their previous output.