Continuous voice guided development #21372

What would you like to be added?

Added a continuous listening mode with voice lock so that i can speak to the CLI, and it will ignore all the other people and noise in the vicinity. it is a prototype in python and Ubuntu. but i am willing to implement in typescript

By using continuous listening, I can provide real-time 'Steering Hints' to the model while it is executing long
running shell commands or file searches, making the development loop truly hands-free

The Voice Lock (Speaker Verification) ensures that the CLI doesn't accidentally execute a destructive shell command if someone else in the room (or a video/podcast) speaks a prompt. It uses MFCC-based spectral
fingerprinting to verify the operator's identity locally before processing audio

Why is this needed?

For everyday tasks for regular users. voice commands are more convenient and preferred as opposed to typing.

i do a lot of linux, and adding removing configurations have gotten easier with this cli. If it could be voice activated. it would be far more convenient.

Additional context

I propose extending the existing packages/cli/src/utils/voice.ts to support a continuous mode and adding a SpeakerVerificationService to the core package to handle the local audio signatures.