LogoClawIndex
CasesSkillsAbout
LogoClawIndex

livekit-agents - Orchestrate an offline LiveKit voice agent

Orchestrates a self-hosted offline voice pipeline connecting local STT, LLM, TTS, push-to-talk controls, and operator UI state.

Tags

Updated: 2026-10-16

Capabilities

Typical Inputs

Typical Outputs

What this skill does

  • Orchestrate realtime voice pipelines
  • Connect local STT, LLM, TTS
  • Configure push-to-talk turns
  • Stream responses by sentence
  • Inject retrieved SOP context
  • Publish transcript and safety state
  • Self-host LiveKit offline

Inputs

  • Audio input
  • Local STT endpoint or plugin
  • Local LLM endpoint or plugin
  • Local TTS endpoint or plugin
  • Retrieved SOP context
  • Frontend turn-control RPCs
  • LiveKit URL and credentials

Outputs

  • Spoken agent responses
  • Live transcript
  • SOP card and citations
  • Safety banner and escalation state
  • Realtime audio stream

Requirements

  • Python 3.10 to 3.14
  • livekit-agents 1.5.x
  • Self-hosted livekit-server
  • Local STT, LLM, and TTS support
  • Local Silero VAD support
  • Offline-capable Apple-Silicon MacBook

Source

  • Spec: SKILL.md

ClawIndex

OpenClaw Skills & Use Case Index

ClawIndex is an ecosystem-driven index of OpenClaw skills and real-world use cases.

Index

Skills·
Cases

Meta

About·
Disclaimer·
Email·
GitHub
© 2026 ClawIndex All Rights Reserved.
LiveKit
voice agent
offline inference
speech-to-text
text-to-speech
push-to-talk
self-hosting
Orchestrate realtime voice pipelines
Connect local STT, LLM, TTS
Configure push-to-talk turns
Stream responses by sentence
Audio input
Local STT endpoint or plugin
Local LLM endpoint or plugin
Spoken agent responses
Live transcript
SOP card and citations