livekit-agent-tools

Design and implement LiveKit agent tools with the @function_tool decorator.

2|4|Updated Nov 21, 2025
One-click install
npx skills add https://github.com/Okeysir198/P20251122-claude-skills --skill livekit-agent-tools
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: livekit-agent-tools
Source: https://github.com/Okeysir198/P20251122-claude-skills/tree/main/skills-reference/livekit-agent-tools
Command: npx skills add https://github.com/Okeysir198/P20251122-claude-skills --skill livekit-agent-tools

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve?

This Skill provides a practical, end-to-end guide for building robust LiveKit voice agent tools using the @function_tool decorator. It helps developers create well-scoped tools, document parameters clearly, and implement safe, production-ready patterns for real-time voice tasks.

Core Features & Use Cases

  • Clear tool design patterns: single-responsibility functions, precise tool descriptions, and meaningful return values that guide LLM usage.
  • RunContext and state management: access to userdata, RunContext, and session orchestration for multi-agent workflows.
  • Production readiness: error handling, logging, monitoring, and testing strategies for reliable agent behavior.
  • Multi-agent patterns: agent handoff, shared state, dynamic tool creation, and context-preserving transfers.

Quick Start

Read the companion assets (agent-template.py) and references (references/*) to implement a basic LiveKit tool, then experiment with a simple two-agent workflow by following the patterns shown in the repository.

Frequently Asked Questions about livekit-agent-tools

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build LiveKit agent tools using the function_tool decorator?

You design LiveKit agent tools by applying the @function_tool decorator to single-responsibility Python functions, ensuring precise descriptions and meaningful return values that guide LLM usage in real-time voice tasks.

How does RunContext handle state in multi-agent LiveKit workflows?

RunContext handles state in multi-agent LiveKit workflows by providing access to userdata and enabling session orchestration for agent handoffs, shared state, and context-preserving transfers.

Can I use LiveKit agent tools for multi-agent workflows with dynamic tooling?

Yes, you can use LiveKit agent tools for multi-agent workflows by implementing agent handoffs, sharing state via RunContext, and creating dynamic tools that preserve context during transfers.

What patterns should I follow for error handling and testing LiveKit voice agents?

For error handling and testing LiveKit voice agents, follow production readiness patterns that include robust error management, logging, monitoring, and testing strategies to ensure reliable agent behavior.

Do I need Python to implement interruption patterns for LiveKit agent tools?

Yes, you need a standard Python environment to implement interruption patterns for LiveKit agent tools, utilizing provided templates and references for hands-on learning and production-ready deployment.