DEMO / RECORDING
Overview & Result
A real-time voice-activated desktop control agent for macOS. Combines local Whisper speech recognition with Jev System 1 reflexes to translate natural spoken commands into immediate system actions (opening apps, managing windows, triggering automations) without sluggish pauses.
How Jev fits in the loop
- Continuously stream microphone audio through local whisper.cpp for zero-latency speech transcription.
- Parse recognized voice intents into candidate system action options via Jev Choice (<40ms).
- Verify safety boundaries to prevent destructive desktop shell execution.
- Invoke native macOS AppleScript or Accessibility APIs to execute desktop state changes immediately.
How to reproduce
- Inspect author post and reproduction notes.
- Deploy agent loop and bind Jev decision endpoints.
- Run test inputs and observe latency and accuracy.
Why this build matters
Delivers an instantaneous voice assistant user experience that feels as snappy as native keyboard shortcuts.
Limitations
- Community observed build; metrics are author-reported.
- Requires third-party dependencies as described in source.