Hello LiveKit AI, Help me with 1 command which gives me active participants ongoing calls on a specific server?

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

what If I dont even know the room name?? I just want to get the list of participants OR say list of rooms with participants in it..

If you’re not sure about the room name, no worries! You can easily see all the active rooms along with how many participants are in each by running this command:

lk room list

This will show you all the active rooms with details like: • Room ID • Room Name • Number of Participants • Number of Publishers If you’d prefer to work with the data programmatically, you can get everything in JSON format by using:

lk room list --json

And if you want to dig a bit deeper to see who’s in each room, just list the rooms first, then check participants for any room with:

lk room participants list <room_name>

Feel free to reach out if you have any questions or need a hand with anything else! Sources: autocomplete/fish_autocomplete

Okay, Suppose there's an existing room, where already a human participant and an AI agent is there, I want another user to be added in an existing room who can also see the chat_ctx and control VoicePipelineAgent conversation with human user

To add another user to an existing room and let them access the chat context and control the VoicePipelineAgent conversation, here’s a friendly step-by-step guide for you: 1. Generate an Access Token for the User: You’ll want to create a token that gives them permission to join the room, publish audio, publish data, and subscribe:

// Example in JavaScript
const grant = {
  room: roomName,
  roomJoin: true,
  canPublish: true,
  canPublishData: true,
  canSubscribe: true,
};
const at = new AccessToken('LIVEKIT_API_KEY', 'LIVEKIT_API_SECRET', { identity: 'user_identity' });
at.addGrant(grant);
const token = at.toJwt();

2. Initialize the VoicePipelineAgent: Set up the agent with components like VAD, STT, LLM, and TTS, and configure the chat context to make it super helpful:

const agent = new pipeline.VoicePipelineAgent(vad, stt, llm, tts, {
  chatCtx: new llm.ChatContext().append({
    role: llm.ChatRole.SYSTEM,
    text: 'You are a helpful assistant.',
  }),
  allowInterruptions: true,
});

3. Start the Agent: Connect the agent to the room and participant like this:

const participant = await ctx.waitForParticipant('participant_identity');
agent.start(ctx.room, participant);

4. Control the Conversation: Use the

say

method to have the agent speak and smoothly manage the chat:

await agent.say('Hello, how can I help you today?', true);

If you want me to walk you through any of these steps or if something’s unclear, just let me know—I’m here to help! Sources: complex-agents/nova-sonic/nova-sonic-form-agent/app/api/connection-details/... | Agents 0.x migration guide | LiveKit Docs | examples/src/outbound.ts | VoicePipelineAgent | LiveKit Agents