如何使用ToolNode调用工具¶

本指南介绍了如何使用LangGraph的预构建ToolNode进行工具调用。

ToolNode是LangChain的一个可运行组件，它接受带有消息列表的图状态作为输入，并输出带有工具调用结果的状态更新。它设计为与LangGraph的预构建ReAct代理无缝配合，但也可以与任何具有适当reducer的StateGraph一起使用（参见MessagesState）。

设置¶

首先，让我们安装所需的包并设置API密钥

%%capture --no-stderr
%pip install --quiet -U langgraph langchain_anthropic

import getpass
import os


def _set_env(var: str):
    if not os.environ.get(var):
        os.environ[var] = getpass.getpass(f"{var}: ")


_set_env("ANTHROPIC_API_KEY")

为LangGraph开发设置LangSmith

注册LangSmith，可以快速发现并解决您的LangGraph项目中的问题，提高项目性能。LangSmith允许您使用跟踪数据来调试、测试和监控使用LangGraph构建的LLM应用程序——更多关于如何开始的内容，请参阅这里。

定义工具¶

from langchain_core.messages import AIMessage
from langchain_core.tools import tool

from langgraph.prebuilt import ToolNode

API Reference: AIMessage | tool | ToolNode

@tool
def get_weather(location: str):
    """Call to get the current weather."""
    if location.lower() in ["sf", "san francisco"]:
        return "It's 60 degrees and foggy."
    else:
        return "It's 90 degrees and sunny."


@tool
def get_coolest_cities():
    """Get a list of coolest cities"""
    return "nyc, sf"

tools = [get_weather, get_coolest_cities]
tool_node = ToolNode(tools)

手动调用`ToolNode`¶

ToolNode 在图状态上操作，该状态包含一系列消息。它期望列表中的最后一条消息是一个带有 tool_calls 参数的 AIMessage。

让我们首先看看如何手动调用工具节点：

message_with_single_tool_call = AIMessage(
    content="",
    tool_calls=[
        {
            "name": "get_weather",
            "args": {"location": "sf"},
            "id": "tool_call_id",
            "type": "tool_call",
        }
    ],
)

tool_node.invoke({"messages": [message_with_single_tool_call]})

{'messages': [ToolMessage(content="It's 60 degrees and foggy.", name='get_weather', tool_call_id='tool_call_id')]}

注意，通常你不需要手动创建AIMessage，它将由任何支持工具调用的LangChain聊天模型自动生成。

你也可以通过将多个工具调用传递给AIMessage的tool_calls参数来实现并行工具调用，使用ToolNode：

message_with_multiple_tool_calls = AIMessage(
    content="",
    tool_calls=[
        {
            "name": "get_coolest_cities",
            "args": {},
            "id": "tool_call_id_1",
            "type": "tool_call",
        },
        {
            "name": "get_weather",
            "args": {"location": "sf"},
            "id": "tool_call_id_2",
            "type": "tool_call",
        },
    ],
)

tool_node.invoke({"messages": [message_with_multiple_tool_calls]})

{'messages': [ToolMessage(content='nyc, sf', name='get_coolest_cities', tool_call_id='tool_call_id_1'),
  ToolMessage(content="It's 60 degrees and foggy.", name='get_weather', tool_call_id='tool_call_id_2')]}

与聊天模型结合使用¶

我们将使用Anthropic提供的一个小型对话模型作为示例。为了使用带有工具调用功能的对话模型，我们首先需要确保模型了解可用的工具。我们通过调用ChatAnthropic模型的.bind_tools方法来实现这一点。

from typing import Literal

from langchain_anthropic import ChatAnthropic
from langgraph.graph import StateGraph, MessagesState
from langgraph.prebuilt import ToolNode


model_with_tools = ChatAnthropic(
    model="claude-3-haiku-20240307", temperature=0
).bind_tools(tools)

API Reference: ChatAnthropic | StateGraph | ToolNode

model_with_tools.invoke("what's the weather in sf?").tool_calls

[{'name': 'get_weather',
  'args': {'location': 'San Francisco'},
  'id': 'toolu_01Fwm7dg1mcJU43Fkx2pqgm8',
  'type': 'tool_call'}]

如你所见，聊天模型生成的AI消息中已经填充了tool_calls，所以我们可以直接将其传递给ToolNode

tool_node.invoke({"messages": [model_with_tools.invoke("what's the weather in sf?")]})

{'messages': [ToolMessage(content="It's 60 degrees and foggy.", name='get_weather', tool_call_id='toolu_01LFvAVT3xJMeZS6kbWwBGZK')]}

ReAct 代理¶

接下来，让我们看看如何在LangGraph图中使用ToolNode。我们将设置一个ReAct代理的图实现。该代理将某些查询作为输入，然后反复调用工具，直到它有足够的信息来解决查询。我们将使用ToolNode和我们刚刚定义的Anthropic模型及工具。

from typing import Literal

from langgraph.graph import StateGraph, MessagesState, START, END


def should_continue(state: MessagesState):
    messages = state["messages"]
    last_message = messages[-1]
    if last_message.tool_calls:
        return "tools"
    return END


def call_model(state: MessagesState):
    messages = state["messages"]
    response = model_with_tools.invoke(messages)
    return {"messages": [response]}


workflow = StateGraph(MessagesState)

# Define the two nodes we will cycle between
workflow.add_node("agent", call_model)
workflow.add_node("tools", tool_node)

workflow.add_edge(START, "agent")
workflow.add_conditional_edges("agent", should_continue, ["tools", END])
workflow.add_edge("tools", "agent")

app = workflow.compile()

API Reference: StateGraph | START | END

from IPython.display import Image, display

try:
    display(Image(app.get_graph().draw_mermaid_png()))
except Exception:
    # This requires some extra dependencies and is optional
    pass

让我们试一试！

# example with a single tool call
for chunk in app.stream(
    {"messages": [("human", "what's the weather in sf?")]}, stream_mode="values"
):
    chunk["messages"][-1].pretty_print()

================================[1m Human Message [0m=================================

what's the weather in sf?
==================================[1m Ai Message [0m==================================

[{'text': "Okay, let's check the weather in San Francisco:", 'type': 'text'}, {'id': 'toolu_01LdmBXYeccWKdPrhZSwFCDX', 'input': {'location': 'San Francisco'}, 'name': 'get_weather', 'type': 'tool_use'}]
Tool Calls:
  get_weather (toolu_01LdmBXYeccWKdPrhZSwFCDX)
 Call ID: toolu_01LdmBXYeccWKdPrhZSwFCDX
  Args:
    location: San Francisco
=================================[1m Tool Message [0m=================================
Name: get_weather

It's 60 degrees and foggy.
==================================[1m Ai Message [0m==================================

The weather in San Francisco is currently 60 degrees with foggy conditions.

# example with a multiple tool calls in succession

for chunk in app.stream(
    {"messages": [("human", "what's the weather in the coolest cities?")]},
    stream_mode="values",
):
    chunk["messages"][-1].pretty_print()

================================[1m Human Message [0m=================================

what's the weather in the coolest cities?
==================================[1m Ai Message [0m==================================

[{'text': "Okay, let's find out the weather in the coolest cities:", 'type': 'text'}, {'id': 'toolu_01LFZUWTccyveBdaSAisMi95', 'input': {}, 'name': 'get_coolest_cities', 'type': 'tool_use'}]
Tool Calls:
  get_coolest_cities (toolu_01LFZUWTccyveBdaSAisMi95)
 Call ID: toolu_01LFZUWTccyveBdaSAisMi95
  Args:
=================================[1m Tool Message [0m=================================
Name: get_coolest_cities

nyc, sf
==================================[1m Ai Message [0m==================================

[{'text': "Now let's get the weather for those cities:", 'type': 'text'}, {'id': 'toolu_01RHPQBhT1u6eDnPqqkGUpsV', 'input': {'location': 'nyc'}, 'name': 'get_weather', 'type': 'tool_use'}]
Tool Calls:
  get_weather (toolu_01RHPQBhT1u6eDnPqqkGUpsV)
 Call ID: toolu_01RHPQBhT1u6eDnPqqkGUpsV
  Args:
    location: nyc
=================================[1m Tool Message [0m=================================
Name: get_weather

It's 90 degrees and sunny.
==================================[1m Ai Message [0m==================================

[{'id': 'toolu_01W5sFGF8PfgYzdY4CqT5c6e', 'input': {'location': 'sf'}, 'name': 'get_weather', 'type': 'tool_use'}]
Tool Calls:
  get_weather (toolu_01W5sFGF8PfgYzdY4CqT5c6e)
 Call ID: toolu_01W5sFGF8PfgYzdY4CqT5c6e
  Args:
    location: sf
=================================[1m Tool Message [0m=================================
Name: get_weather

It's 60 degrees and foggy.
==================================[1m Ai Message [0m==================================

Based on the results, it looks like the weather in the coolest cities is:
- New York City: 90 degrees and sunny
- San Francisco: 60 degrees and foggy

So the weather in the coolest cities is a mix of warm and cool temperatures, with some sunny and some foggy conditions.

ToolNode 还可以处理工具执行过程中的错误。你可以通过设置 handle_tool_errors=True（默认启用）来启用或禁用此功能。有关如何在 ToolNode 中处理错误的指南，请参阅此处。

如何使用ToolNode调用工具¶

设置¶

定义工具¶

手动调用ToolNode¶

与聊天模型结合使用¶

ReAct 代理¶

Comments

手动调用`ToolNode`¶