Using with a custom agent

from typing_extensions import Literal
from langchain_core.messages import ToolMessage
from langchain_core.tools import tool
from langgraph.graph import MessagesState, StateGraph, START
from langgraph.types import Command


def make_agent(model, tools, system_prompt=None):
    model_with_tools = model.bind_tools(tools)
    tools_by_name = {tool.name: tool for tool in tools}

    def call_model(state: MessagesState) -> Command[Literal["call_tools", "__end__"]]:
        messages = state["messages"]
        if system_prompt:
            messages = [{"role": "system", "content": system_prompt}] + messages

        response = model_with_tools.invoke(messages)
        if len(response.tool_calls) > 0:
            return Command(goto="call_tools", update={"messages": [response]})

        return {"messages": [response]}

    # NOTE: this is a simplified version of the prebuilt ToolNode
    # If you want to have a tool node that has full feature parity, please refer to the source code
    def call_tools(state: MessagesState) -> Command[Literal["call_model"]]:
        tool_calls = state["messages"][-1].tool_calls
        results = []
        for tool_call in tool_calls:
            tool_ = tools_by_name[tool_call["name"]]
            tool_input_fields = tool_.get_input_schema().model_json_schema()[
                "properties"
            ]

            # this is simplified for demonstration purposes and
            # is different from the ToolNode implementation
            if "state" in tool_input_fields:
                # inject state
                tool_call = {**tool_call, "args": {**tool_call["args"], "state": state}}

            tool_response = tool_.invoke(tool_call)
            if isinstance(tool_response, ToolMessage):
                results.append(Command(update={"messages": [tool_response]}))

            # handle tools that return Command directly
            elif isinstance(tool_response, Command):
                results.append(tool_response)

        # NOTE: nodes in LangGraph allow you to return list of updates, including Command objects
        return results

    graph = StateGraph(MessagesState)
    graph.add_node(call_model)
    graph.add_node(call_tools)
    graph.add_edge(START, "call_model")
    graph.add_edge("call_tools", "call_model")

    return graph.compile()

Define tools

@tool
def add(a: int, b: int) -> int:
    """Adds two numbers."""
    return a + b


@tool
def multiply(a: int, b: int) -> int:
    """Multiplies two numbers."""
    return a * b

Create an agent

from langchain_openai import ChatOpenAI
model = ChatOpenAI(
    temperature=0,
    model="GLM-4-PLUS",
    openai_api_key="your api key",
    openai_api_base="https://open.bigmodel.cn/api/paas/v4/"
)

agent = make_agent(model, [add, multiply])

from langchain_core.messages import convert_to_messages


def pretty_print_messages(update):
    if isinstance(update, tuple):
        ns, update = update
        # skip parent graph updates in the printouts
        if len(ns) == 0:
            return

        graph_id = ns[-1].split(":")[0]
        print(f"Update from subgraph {graph_id}:")
        print("\n")

    for node_name, node_update in update.items():
        print(f"Update from node {node_name}:")
        print("\n")

        for m in convert_to_messages(node_update["messages"]):
            m.pretty_print()
        print("\n")

for chunk in agent.stream({"messages": [("user", "what's (3 + 5) * 12")]}):
    pretty_print_messages(chunk)

Update from node call_model:


==================================[1m Ai Message [0m==================================
Tool Calls:
  add (call_-9024510033517604232)
 Call ID: call_-9024510033517604232
  Args:
    a: 3
    b: 5


Update from node call_tools:


=================================[1m Tool Message [0m=================================
Name: add

8


Update from node call_model:


==================================[1m Ai Message [0m==================================
Tool Calls:
  multiply (call_-9024516080833702606)
 Call ID: call_-9024516080833702606
  Args:
    a: 8
    b: 12


Update from node call_tools:


=================================[1m Tool Message [0m=================================
Name: multiply

96


Update from node call_model:


==================================[1m Ai Message [0m==================================

The result of (3 + 5) * 12 is 96.

Define handoff tools

from typing import Annotated

from langchain_core.tools import tool
from langchain_core.tools.base import InjectedToolCallId
from langgraph.prebuilt import InjectedState


def make_handoff_tool(*, agent_name: str):
    """Create a tool that can return handoff via a Command"""
    tool_name = f"transfer_to_{agent_name}"

    @tool(tool_name)
    def handoff_to_agent(
        # # optionally pass current graph state to the tool (will be ignored by the LLM)
        state: Annotated[dict, InjectedState],
        # optionally pass the current tool call ID (will be ignored by the LLM)
        tool_call_id: Annotated[str, InjectedToolCallId],
    ):
        """Ask another agent for help."""
        tool_message = {
            "role": "tool",
            "content": f"Successfully transferred to {agent_name}",
            "name": tool_name,
            "tool_call_id": tool_call_id,
        }
        return Command(
            # navigate to another agent node in the PARENT graph
            goto=agent_name,
            graph=Command.PARENT,
            # This is the state update that the agent `agent_name` will see when it is invoked.
            # We're passing agent's FULL internal message history AND adding a tool message to make sure
            # the resulting chat history is valid. See the paragraph above for more information.
            update={"messages": state["messages"] + [tool_message]},
        )

    return handoff_to_agent

Define agents and graph

addition_expert = make_agent(
    model,
    [add, make_handoff_tool(agent_name="multiplication_expert")],
    system_prompt="You are an addition expert, you can ask the multiplication expert for help with multiplication.",
)
multiplication_expert = make_agent(
    model,
    [multiply, make_handoff_tool(agent_name="addition_expert")],
    system_prompt="You are a multiplication expert, you can ask an addition expert for help with addition.",
)

builder = StateGraph(MessagesState)
builder.add_node("addition_expert", addition_expert)
builder.add_node("multiplication_expert", multiplication_expert)
builder.add_edge(START, "addition_expert")
graph = builder.compile()

Visualize the graph

from IPython.display import Image, display

display(Image(graph.get_graph().draw_mermaid_png()))

Run the graph

for chunk in graph.stream(
    {"messages": [("user", "what's (3 + 5) * 12")]}, subgraphs=True
):
    pretty_print_messages(chunk)

Update from subgraph addition_expert:


Update from node call_model:


==================================[1m Ai Message [0m==================================
Tool Calls:
  transfer_to_multiplication_expert (call_-9024520341442308872)
 Call ID: call_-9024520341442308872
  Args:


Update from subgraph multiplication_expert:


Update from node call_model:


==================================[1m Ai Message [0m==================================
Tool Calls:
  transfer_to_addition_expert (call_-9024514672083873280)
 Call ID: call_-9024514672083873280
  Args:


Update from subgraph addition_expert:


Update from node call_model:


==================================[1m Ai Message [0m==================================
Tool Calls:
  add (call_-9024512129462455947)
 Call ID: call_-9024512129462455947
  Args:
    a: 3
    b: 5


Update from subgraph addition_expert:


Update from node call_tools:


=================================[1m Tool Message [0m=================================
Name: add

8


Update from subgraph addition_expert:


Update from node call_model:


==================================[1m Ai Message [0m==================================

The result of (3 + 5) * 12 is 96.

Using prebuilt agents

from langgraph.prebuilt import create_react_agent

addition_expert = create_react_agent(
    model,
    [add, make_handoff_tool(agent_name="multiplication_expert")],
    prompt="You are an addition expert, you can ask the multiplication expert for help with multiplication.",
)

multiplication_expert = create_react_agent(
    model,
    [multiply, make_handoff_tool(agent_name="addition_expert")],
    prompt="You are a multiplication expert, you can ask an addition expert for help with addition.",
)

builder = StateGraph(MessagesState)
builder.add_node("addition_expert", addition_expert)
builder.add_node("multiplication_expert", multiplication_expert)
builder.add_edge(START, "addition_expert")
graph = builder.compile()

run the graph

for chunk in graph.stream(
    {"messages": [("user", "what's (3 + 5) * 12")]}, subgraphs=True
):
    pretty_print_messages(chunk)

Update from subgraph addition_expert:


Update from node agent:


==================================[1m Ai Message [0m==================================
Tool Calls:
  transfer_to_multiplication_expert (call_-9024512232541694769)
 Call ID: call_-9024512232541694769
  Args:


Update from subgraph multiplication_expert:


Update from node agent:


==================================[1m Ai Message [0m==================================
Tool Calls:
  transfer_to_addition_expert (call_-9024517867540713141)
 Call ID: call_-9024517867540713141
  Args:


Update from subgraph addition_expert:


Update from node agent:


==================================[1m Ai Message [0m==================================
Tool Calls:
  add (call_-9024509037084815237)
 Call ID: call_-9024509037084815237
  Args:
    a: 3
    b: 5


Update from subgraph addition_expert:


Update from node tools:


=================================[1m Tool Message [0m=================================
Name: add

8


Update from subgraph addition_expert:


Update from node agent:


==================================[1m Ai Message [0m==================================

The result of (3 + 5) * 12 is 96.

another example

from langgraph.prebuilt import create_react_agent

addition_expert = create_react_agent(
    model,
    [add, make_handoff_tool(agent_name="English_expert")],
    prompt="You only speak in English, you can ask the Chinese expert for help with Chinese.",
)

multiplication_expert = create_react_agent(
    model,
    [multiply, make_handoff_tool(agent_name="中文专家")],
    prompt="你只讲中文，遇到英文问题，向别的agent请求帮助",
)

builder = StateGraph(MessagesState)
builder.add_node("addition_expert", addition_expert)
builder.add_node("multiplication_expert", multiplication_expert)
builder.add_edge(START, "addition_expert")
graph = builder.compile()

for chunk in graph.stream(
    {"messages": [("user", "what's deepseek")]}, subgraphs=True
):
    pretty_print_messages(chunk)

Update from subgraph addition_expert:


Update from node agent:


==================================[1m Ai Message [0m==================================

"Deepseek" isn't a widely recognized term in popular culture or technology as of my last update in early 2023. It could be a misspelling, a specific term used in a niche field, or a new concept that has emerged after my last update. Here are a few possibilities:

1. **Typo or Misspelling**: It might be a misspelling of a more common term like "deepfake," which refers to sophisticated AI-generated audio and video content that can mimic real people, or "deep learning," a subset of machine learning involving neural networks.

2. **Niche Term**: It could be a specific term used in a particular industry or academic field. For example, it might relate to a specific algorithm, software, or methodology in data science, AI, or another technical field.

3. **New Concept or Product**: If "deepseek" is a new term that has emerged recently, it might refer to a new technology, product, or service that wasn't widely known before 2023.

4. **Brand or Company Name**: It could also be the name of a company, brand, or product that is not widely recognized yet.

To get a more accurate answer, additional context would be helpful. If you have more details or a specific context in which you encountered the term "deepseek," please provide them, and I can give a more targeted response. Alternatively, searching online or checking recent publications might yield more current information.

参考链接：https://langchain-ai.github.io/langgraph/how-tos/agent-handoffs/#implement-a-handoff-tool