@ctx.room.on("track_subscribed") def on_track_subscribed(track: rtc.Track, publication: rtc.RemoteTrackPublication, participant: rtc.RemoteParticipant): if track.kind == rtc.TrackKind.KIND_AUDIO: logger.warning("====================Track log=====================") logger.info("track subscribed: %s from participant %s (%s)", publication.sid, participant.sid, participant.identity) stream = rtc.AudioStream(track) does this track_subcribed give only local_participant audio I can't get agent audio

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

best way to get agent and remote participant tracks

I'm happy to help you get both agent and remote participant tracks in LiveKit, and the best approach really depends on what you're aiming to do: • For AI Agents: You can use the

useVoiceAssistant

hook. This handy hook automatically grabs the agent's audio and video tracks, including those from avatar workers. Here's a quick example:

const { agent, audioTrack, videoTrack, state } = useVoiceAssistant();

• For Remote Participants: You might want to use

useTracks

or

useParticipantTracks

to access tracks from all participants or just specific ones. Check out these examples:

// All camera and mic tracks in the room
const allTracks = useTracks([Track.Source.Camera, Track.Source.Microphone]);

// Specific participant's tracks
const participantTracks = useParticipantTracks(
  [Track.Source.Camera, Track.Source.Microphone],
  participant.identity
);

If you'd like, I’m here to walk you through these or answer any questions you have about implementing them! Sources: Virtual avatar integrations | LiveKit Docs | packages/core/src/observables/track.ts

I'm using python SDK

Hi there! To help you get both agent (local) and remote participant tracks using the Python SDK in LiveKit, here’s a friendly step-by-step guide: • Access Remote Participant Tracks: You can listen for the

track_subscribed

event to handle tracks as they come in:

@room.on("track_subscribed")
def on_track_subscribed(track: rtc.Track, publication: rtc.RemoteTrackPublication, participant: rtc.RemoteParticipant):
    if track.kind == rtc.TrackKind.KIND_AUDIO:
        audio_stream = rtc.AudioStream(track)
        asyncio.create_task(process_audio(audio_stream))

• Access Agent (Local) Tracks: For managing the agent’s own tracks, use

room.local_participant

. Here’s how you can publish a local audio track:

source = rtc.AudioSource(SAMPLE_RATE, NUM_CHANNELS)
track = rtc.LocalAudioTrack.create_audio_track("example-track", source)
await room.local_participant.publish_track(track)

• Iterate Over Participants: If you want to go through all participants and their tracks, this snippet will help:

for participant in room.remote_participants.values():
    for publication in participant.track_publications.values():
        if publication.track:
            track = publication.track
            # Process the track

If you have any questions or need a hand with implementation, please don’t hesitate to ask—I’m here to help! Sources: Inside a session | LiveKit Docs | livekit-rtc/livekit/rtc/__init__.py | Receiving and publishing tracks | LiveKit Docs | Receiving and publishing tracks | LiveKit Docs | Virtual avatar integrations | LiveKit Docs