I’m happy to guide you through generating an acknowledgment to the caller right after a response but before tool calls are executed in Node.js Agent v1. You can achieve this by overriding the

llmNode()

method in a custom

Agent

subclass. This approach lets you insert your acknowledgment logic immediately after the LLM generates a response and just before any tool calls happen. Here’s a friendly example to help you get started:

import { voice, llm } from '@livekit/agents';

class MyAgent extends voice.Agent {
  async llmNode(
    chatCtx: llm.ChatContext,
    tools: llm.FunctionTool[],
    modelSettings: voice.ModelSettings,
  ): Promise<AsyncIterable<llm.ChatChunk | string>> {
    // First, proceed with the default LLM processing
    const chunks = await voice.Agent.default.llmNode(this, chatCtx, tools, modelSettings);

    // Then, create a warm acknowledgment message after the LLM response
    const acknowledgment: llm.ChatMessage = {
      id: `ack_${Date.now()}`,
      role: 'assistant',
      content: ['Just a moment, I’m processing your request...'],
      createdAt: new Date(),
    };

    // Add this acknowledgment into the chat context
    this.session.chatCtx.addMessage(acknowledgment);

    // Make sure the acknowledgment is visible to the caller
    this.session._conversationItemAdded(acknowledgment);

    // Finally, return the original LLM response chunks
    return chunks;
  }
}

This way, your caller will see a friendly acknowledgment before the tool execution starts. If you’d like me to walk you through any part of this or if you have more questions, just let me know—I’m here to help! Sources: livekit-agents/livekit/agents/voice/agent_activity.py | agents/src/llm/provider_format/openai.test.ts | Pipeline nodes and hooks | LiveKit Docs | Tool definition and use | LiveKit Docs