Enhance Gemini CLI with voice interaction capabilities.

[yashaspancham]

What would you like to be added?

Suggestion: Allow for the integration of Speech-to-Text (STT) and Text-to-Speech (TTS) CLI tools to enable voice-controlled interaction with the Gemini CLI agent.
 
 Mechanism: Users could build a wrapper script that:
1. Captures spoken input via an STT CLI tool, converting it to text.
 2. Pipes this text as a prompt to the Gemini CLI.
 3. Captures the Gemini CLI's text response.
 4. Pipes the text response to a TTS CLI tool for spoken output.

Why is this needed?

This would create a hands-free, more natural, and voice-controlled user experience, similar to interacting with an AI assistant like JARVIS.

Additional context

No response

[fayerman-source]

I have implemented a native, local-first solution for this using sox/arecord and Whisper. It integrates directly into the InputPrompt with a visual recording indicator (🎤) and Alt+V/Ctrl+Q shortcuts.

Key features:

• Native Integration: Lives inside the CLI process (no external servers).
• Privacy: All audio stays local; transcription happens via a local Whisper binary.
• Optimized UX: Low-latency transcription insertion at the cursor position.
• Unit Tested: Includes comprehensive tests for the hook and shortcut bindings.

I'm ready to open a PR for review. Would the maintainers like me to proceed?

[fayerman-source]

My PR #18499 introduces an experimental version of voice input. It handles recording and transcription within the CLI application.