37downloads
Debug LiveKit agents through live conversations
Try a LiveKit agent in a live multi turn conversation with lk agent debugger. Inspect what it says and does after changing its prompts or tools.
When it runs
Your agent reads this description from SKILL.md to decide when to load the skill:
Drives a live multi-turn conversation with a LiveKit agent running locally, using `lk agent debugger`. Use when the user says "test my agent", "try my agent", "does this work", or "why did it call that tool", after editing an agent to check how it behaves, or when an agent on a speech-to-speech model or one that mishears users needs checking. The default for a bare "test"; for regression tests use testing-livekit-agents, for graded runs running-livekit-simulations.
What SKILL.md covers
- The loop
- Debugging with it
- Audio mode
- When to use something else
- Related skills
Files in the ZIP
Related skills
Build LiveKit voice and chat AI agents
Design LiveKit voice and chat agents with clear tasks, handoffs and tool behavior. Keep context focused and plan for conversational latency.
Deploy and operate LiveKit agents in production
Ship and roll back LiveKit agents, manage worker processes and investigate production failures. Review secrets, shutdown behavior, upgrades and observability.
Look up LiveKit APIs, models and CLI documentation
Read current LiveKit documentation before using an API, CLI flag or provider feature. Find SDK signatures, examples and changelog details from source.