This covers the full stack of modern voice AI, from OpenAI's Realtime API for native voice-to-voice with GPT-4o to Vapi for quick phone agent deployment. You get patterns for Deepgram transcription, ElevenLabs synthesis, and LiveKit infrastructure, plus the WebRTC and WebSocket foundations you need. The focus is clearly on production concerns: latency budgets, provider selection, audio quality tuning. If you're building anything with real-time voice, whether it's a customer service bot or a voice-enabled app, this gives you the architecture patterns and integration code to get started. The honest take is that voice apps are unforgiving, they feel magical when fast and broken when slow, and this skill knows it.
npx skills add https://github.com/sickn33/antigravity-awesome-skills --skill voice-ai-development