how do I access chat_ctx if I have agent? this is my code: agent = setup_voice_agent( ctx=ctx, stt_agent=deepgram_stt_agent, tts_agent=tts_agent, vad=ctx.proc.userdata["vad"] ) ctx.proc.userdata["agent"] = agent I set up the voice agent using the stt and tts pipeline but how do I access the llm here?

Thanks for reaching out, Sai. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.

how to access the content being played out in LLMStream

Unfortunately, I couldn't find specific information on how to access the content being played out in

LLMStream

. The available documentation only mentions speech playback functionality with different source types, including

LLMStream

, but does not detail how to retrieve the content. If you have further details or specific use cases, feel free to share, and I can assist further or escalate the query if needed.

how do I force an agent turn?

Currently, I couldn't locate specific information on how to force an agent turn in LiveKit. The documentation primarily discusses turn detection methods like Voice Activity Detection (VAD) and custom turn-detector models for determining when a user has finished speaking, but it does not detail how to manually trigger an agent turn. If you have additional context or requirements, please let me know, and I can assist further or escalate the query if necessary.

stream = self.agent.llm.chat(chat_ctx=self.agent.chat_ctx) this is how I get LLM Stream. how do I access the actual text of the stream before doing agent.say?

I'm sorry, but I don't have the information you need at the moment. If you would like a member of the LiveKit team to respond, please ask your question in one of the other channels.

what is the difference between await self.ctx.room.disconnect() and await self.api.room.remove_participant( api.RoomParticipantIdentity( room=self.room.name, identity=self.participant.identity, ) ) ? which actually ends the call>