Web Speech API trigger with listening, processing, error, and unsupported states.
npx shadcn@latest add https://ui.edbn.me/r/audio-speech-input.json