Multi-agent¶

A single agent might struggle if it needs to specialize in multiple domains or manage many tools. To tackle this, you can break your agent into smaller, independent agents and composing them into a multi-agent system.

In multi-agent systems, agents need to communicate between each other. They do so via handoffs — a primitive that describes which agent to hand control to and the payload to send to that agent.

Two of the most popular multi-agent architectures are:

supervisor — individual agents are coordinated by a central supervisor agent. The supervisor controls all communication flow and task delegation, making decisions about which agent to invoke based on the current context and task requirements.
swarm — agents dynamically hand off control to one another based on their specializations. The system remembers which agent was last active, ensuring that on subsequent interactions, the conversation resumes with that agent.

Supervisor¶

Supervisor

Use langgraph-supervisor library to create a supervisor multi-agent system:

npm install @langchain/langgraph-supervisor

import { ChatOpenAI } from "@langchain/openai";
import { createSupervisor } from "@langchain/langgraph-supervisor";
import { createReactAgent } from "@langchain/langgraph/prebuilt";
import { tool } from "@langchain/core/tools";
import { z } from "zod";

const bookHotel = tool(
  async (input: { hotel_name: string }) => {
    return `Successfully booked a stay at ${input.hotel_name}.`;
  },
  {
    name: "book_hotel",
    description: "Book a hotel",
    schema: z.object({
      hotel_name: z.string().describe("The name of the hotel to book"),
    }),
  }
);

const bookFlight = tool(
  async (input: { from_airport: string; to_airport: string }) => {
    return `Successfully booked a flight from ${input.from_airport} to ${input.to_airport}.`;
  },
  {
    name: "book_flight",
    description: "Book a flight",
    schema: z.object({
      from_airport: z.string().describe("The departure airport code"),
      to_airport: z.string().describe("The arrival airport code"),
    }),
  }
);

const llm = new ChatOpenAI({ modelName: "gpt-4o" });

// Create specialized agents
const flightAssistant = createReactAgent({
  llm,
  tools: [bookFlight],
  prompt: "You are a flight booking assistant",
  name: "flight_assistant",
});

const hotelAssistant = createReactAgent({
  llm,
  tools: [bookHotel],
  prompt: "You are a hotel booking assistant",
  name: "hotel_assistant",
});

const supervisor = createSupervisor({
  agents: [flightAssistant, hotelAssistant],
  llm,
  prompt: "You manage a hotel booking assistant and a flight booking assistant. Assign work to them, one at a time.",
}).compile();

const stream = await supervisor.stream({
  messages: [{
    role: "user",
    content: "first book a flight from BOS to JFK and then book a stay at McKittrick Hotel"
  }]
});

for await (const chunk of stream) {
  console.log(chunk);
  console.log("\n");
}

Swarm¶

Swarm

Use langgraph-swarm library to create a swarm multi-agent system:

npm install @langchain/langgraph-swarm

import { createReactAgent } from "@langchain/langgraph/prebuilt";
import { ChatAnthropic } from "@langchain/anthropic";
import { createSwarm, createHandoffTool } from "@langchain/langgraph-swarm";

const transferToHotelAssistant = createHandoffTool({
  agentName: "hotel_assistant",
  description: "Transfer user to the hotel-booking assistant.",
});

const transferToFlightAssistant = createHandoffTool({
  agentName: "flight_assistant",
  description: "Transfer user to the flight-booking assistant.",
});


const llm = new ChatAnthropic({ modelName: "claude-3-5-sonnet-latest" });

const flightAssistant = createReactAgent({
  llm,
  tools: [bookFlight, transferToHotelAssistant],
  prompt: "You are a flight booking assistant",
  name: "flight_assistant",
});

const hotelAssistant = createReactAgent({
  llm,
  tools: [bookHotel, transferToFlightAssistant],
  prompt: "You are a hotel booking assistant",
  name: "hotel_assistant",
});

const swarm = createSwarm({
  agents: [flightAssistant, hotelAssistant],
  defaultActiveAgent: "flight_assistant",
}).compile();

const stream = await swarm.stream({
  messages: [{
    role: "user",
    content: "first book a flight from BOS to JFK and then book a stay at McKittrick Hotel"
  }]
});

for await (const chunk of stream) {
  console.log(chunk);
  console.log("\n");
}

Handoffs¶

A common pattern in multi-agent interactions is handoffs, where one agent hands off control to another. Handoffs allow you to specify:

destination: target agent to navigate to
payload: information to pass to that agent

This is used both by langgraph-supervisor (supervisor hands off to individual agents) and langgraph-swarm (an individual agent can hand off to other agents).

To implement handoffs with createReactAgent, you need to:

Create a special tool that can transfer control to a different agent

const transferToBob = tool(
  async (_) => {
    return new Command({
      // name of the agent (node) to go to
      goto: "bob",
      // data to send to the agent
      update: { messages: ... },
      // indicate to LangGraph that we need to navigate to
      // agent node in a parent graph
      graph: Command.PARENT,
    });
  },
  {
    name: ...,
    schema: ...,
    description: ...
  }
);

Create individual agents that have access to handoff tools:

const flightAssistant = createReactAgent(
  ..., tools: [bookFlight, transferToHotelAssistant]
)
const hotelAssistant = createReactAgent(
  ..., tools=[bookHotel, transferToFlightAssistant]
)

Define a parent graph that contains individual agents as nodes:

import { StateGraph, MessagesAnnotation } from "@langchain/langgraph";

const multiAgentGraph = new StateGraph(MessagesAnnotation)
  .addNode("flight_assistant", flightAssistant)
  .addNode("hotel_assistant", hotelAssistant)
  ...

Putting this together, here is how you can implement a simple multi-agent system with two agents — a flight booking assistant and a hotel booking assistant:

import { ChatAnthropic } from "@langchain/anthropic";
import { createReactAgent } from "@langchain/langgraph/prebuilt";
import { StateGraph, MessagesAnnotation, Command, START, getCurrentTaskInput, END } from "@langchain/langgraph";
import { tool } from "@langchain/core/tools";
import { z } from "zod";
import { ToolMessage } from "@langchain/core/messages";

interface CreateHandoffToolParams {
  agentName: string;
  description?: string;
}

const createHandoffTool = ({
  agentName,
  description,
}: CreateHandoffToolParams) => {
  const toolName = `transfer_to_${agentName}`;
  const toolDescription = description || `Ask agent '${agentName}' for help`;

  const handoffTool = tool(
    async (_, config) => {
      const toolMessage = new ToolMessage({
        content: `Successfully transferred to ${agentName}`,
        name: toolName,
        tool_call_id: config.toolCall.id,
      });

      // inject the current agent state
      const state =
        getCurrentTaskInput() as (typeof MessagesAnnotation)["State"];  // (1)!
      return new Command({  // (2)!
        goto: agentName,  // (3)!
        update: { messages: state.messages.concat(toolMessage) },  // (4)!
        graph: Command.PARENT,  // (5)!
      });
    },
    {
      name: toolName,
      schema: z.object({}),
      description: toolDescription,
    }
  );

  return handoffTool;
};

const bookHotel = tool(
  async (input: { hotel_name: string }) => {
    return `Successfully booked a stay at ${input.hotel_name}.`;
  },
  {
    name: "book_hotel",
    description: "Book a hotel",
    schema: z.object({
      hotel_name: z.string().describe("The name of the hotel to book"),
    }),
  }
);

const bookFlight = tool(
  async (input: { from_airport: string; to_airport: string }) => {
    return `Successfully booked a flight from ${input.from_airport} to ${input.to_airport}.`;
  },
  {
    name: "book_flight",
    description: "Book a flight",
    schema: z.object({
      from_airport: z.string().describe("The departure airport code"),
      to_airport: z.string().describe("The arrival airport code"),
    }),
  }
);

const transferToHotelAssistant = createHandoffTool({
  agentName: "hotel_assistant",
  description: "Transfer user to the hotel-booking assistant.",
});

const transferToFlightAssistant = createHandoffTool({
  agentName: "flight_assistant",
  description: "Transfer user to the flight-booking assistant.",
});

const llm = new ChatAnthropic({ modelName: "claude-3-5-sonnet-latest" });

const flightAssistant = createReactAgent({
  llm,
  tools: [bookFlight, transferToHotelAssistant],
  prompt: "You are a flight booking assistant",
  name: "flight_assistant",
});

const hotelAssistant = createReactAgent({
  llm,
  tools: [bookHotel, transferToFlightAssistant],
  prompt: "You are a hotel booking assistant",
  name: "hotel_assistant",
});

const multiAgentGraph = new StateGraph(MessagesAnnotation)
  .addNode("flight_assistant", flightAssistant, { ends: ["hotel_assistant", END] })
  .addNode("hotel_assistant", hotelAssistant, { ends: ["flight_assistant", END] })
  .addEdge(START, "flight_assistant")
  .compile();

const stream = await multiAgentGraph.stream({
  messages: [{
    role: "user",
    content: "book a flight from BOS to JFK and a stay at McKittrick Hotel"
  }]
});

for await (const chunk of stream) {
  console.log(chunk);
  console.log("\n");
}

Access agent's state
The Command primitive allows specifying a state update and a node transition as a single operation, making it useful for implementing handoffs.
Name of the agent or node to hand off to.
Take the agent's messages and add them to the parent's state as part of the handoff. The next agent will see the parent state.
Indicate to LangGraph that we need to navigate to agent node in a parent multi-agent graph.

Note

This handoff implementation assumes that:

each agent receives overall message history (across all agents) in the multi-agent system as its input
each agent outputs its internal messages history to the overall message history of the multi-agent system

Check out LangGraph supervisor and swarm documentation to learn how to customize handoffs.