livekit-agents

What problem does it solve?

This Skill provides best practices for developing high-performance, reliable Python voice agents using LiveKit, integrating Speech-to-Text (STT) and Text-to-Speech (TTS) services. It streamlines the creation of interactive AI voice experiences.

Core Features & Use Cases

• Basic Voice Agent Structure: Outlines the fundamental setup for a LiveKit agent, integrating STT (Deepgram), LLM (OpenAI), and TTS (Cartesia) components for end-to-end voice interaction.
• Async Patterns: Emphasizes asyncio.gather for parallel operations and asyncio.create_task for background work to maximize agent responsiveness and efficiency.
• UV Environment Setup: Guides on using uv for efficient dependency management and local execution of Python agents, ensuring a consistent development environment.
• Use Case: A developer needs to create a new voice assistant that transcribes user input, generates an LLM response, and speaks it back. They can use this skill to quickly set up the agent structure, integrate STT/TTS, and ensure efficient asynchronous processing.

Quick Start

Create a basic LiveKit voice agent in 'apps/agent/src/main.py' that uses Deepgram for STT, OpenAI for LLM, and Cartesia for TTS.