yukukotaniyukukotaniCommunityยท1 Agent Skills Included

pi-voice

Push-to-talk voice control for coding sessions

Adds hands-free voice control to the Pi coding agent through a background daemon. Hold a hotkey, speak your request, and hear spoken responses without typing commands. Supports local Whisper, Gemini, OpenAI, and ElevenLabs speech providers with simple configuration. Eliminates constant keyboard switching during long coding and debugging sessions.
npx skills add yukukotani/pi-voice --all -g -y
Available:

Instructs the agent to use Bun commands and built-in Bun APIs instead of Node.js, npm, or external libraries when working in this project.

All Skills in This Repository (1)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install pi-voice?โ–ผ

Run `npx skills add yukukotani/pi-voice --all -g -y` in your terminal to install everything globally.

How to control a coding agent with voice?โ–ผ

Start the pi-voice daemon, hold the push-to-talk hotkey, speak your instruction, and release to send it to the Pi agent with spoken feedback.

Which speech providers does pi-voice support?โ–ผ

It supports a free local Whisper mode plus Gemini, OpenAI, and ElevenLabs, selectable in the pi-voice.json config file.

Does pi-voice work offline?โ–ผ

Yes. The default local provider uses a downloaded Whisper model for speech-to-text and the macOS say command for speech output, so no API key is needed.

Can I change the push-to-talk hotkey?โ–ผ

Yes. Set the key field in .pi/pi-voice.json, for example ctrl+t or alt+space, to override the default shortcut.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’