I would like to create and chat with a text only agent. I tried following the documentation and setting

audio_enabled

to

False

in

RoomInputOptions

and `RoomOutputOptions`:

await session.start(
    agent=client_agent,
    room=ctx.room,
    room_input_options=RoomInputOptions(
        text_enabled=True,
        audio_enabled=False,
    ),
    room_output_options=RoomOutputOptions(
        audio_enabled=False,
    ),
)

But when I connected to this agent in the Agent Playground, it would not create or respond to any messages in the chat. Am I doing something wrong here?

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

Thats already what I did but it isn't working, it works fine if audio_enabled=True for both, so it isn't a problem with the agent.

Hey there! I totally understand the transcription confusion you're experiencing. It sounds like there might be a tricky little issue with how transcriptions get published when audio_enabled=False. Some apps actually depend on the audio track being created to trigger those legacy transcription events, even if you're just using the text stream. Here's a couple of friendly ways we can tackle this: 1. Keep

audio_enabled=True

but just mute the audio on your client side - super simple fix! 2. Make sure your frontend is set up to grab text directly from the text stream (

lk.transcription

) instead of waiting on those older transcription events. Want to chat more about getting this sorted? I'm here to help! Just let me know if you need any extra guidance with these tweaks. Sources: Text and transcriptions | LiveKit Docs | GitHub::livekit::agents | examples/other/transcription/translator.py

The frontend I am using is the Agent Playground, is it configured like that?

Hey there! 👋 Let me break down how the Agent Playground works with text-only agents in a super friendly way. • Input: Your agent will receive text through the

lk.chat

text stream topic – think of it like getting a message from a friend. • Output: Your agent sends responses via the

lk.transcription

text stream topic, and it won't need that extra

lk.transcribed_track_id

attribute. If you're finding that the Agent Playground isn't responding to your text-only agent, don't worry! It might just be a small hiccup with how the lk.transcription text stream is being handled. Could be a configuration quirk or a current limitation. I'd be more than happy to help you troubleshoot and get things running smoothly. Just let me know what's going on, and we'll figure it out together! 😊 Sources: Text and transcriptions | LiveKit Docs | GitHub::livekit::agents | examples/other/chat-stream-receiver.py