livekit-agents
Build voice AI agents on LiveKit Cloud with structured workflows, minimal latency, and mandatory test coverage.
- Use LiveKit Cloud and LiveKit Inference for managed infrastructure and AI models without separate API keys
- Design agents around handoffs (agent-to-agent transitions) and tasks (scoped operations) to isolate context and reduce latency
- Every agent implementation requires tests covering basic conversation flow, tool invocation, error handling, and edge cases before deployment
- Always verify APIs against live documentation via MCP or web search; never rely on model memory for LiveKit SDK specifics
- Minimize system prompts and tool lists to keep voice interactions responsive; context bloat directly increases latency
LiveKit Agents Development for LiveKit Cloud
This skill provides opinionated guidance for building voice AI agents with LiveKit Cloud. It assumes you are using LiveKit Cloud (the recommended path) and encodes how to approach agent development, not API specifics. All factual information about APIs, methods, and configurations must come from live documentation.
This skill is for LiveKit Cloud developers. If you're self-hosting LiveKit, some recommendations (particularly around LiveKit Inference) won't apply directly.
MANDATORY: Read This Checklist Before Starting
Before writing ANY code, complete this checklist:
- Read this entire skill document - Do not skip sections even if MCP is available
- Ensure LiveKit Cloud project is connected - You need
LIVEKIT_URL,LIVEKIT_API_KEY, andLIVEKIT_API_SECRETfrom your Cloud project - Set up documentation access - Use MCP if available, otherwise use web search
- Plan to write tests - Every agent implementation MUST include tests (see testing section below)
- Verify all APIs against live docs - Never rely on model memory for LiveKit APIs
This checklist applies regardless of whether MCP is available. MCP provides documentation access but does NOT replace the guidance in this skill.