如何删除消息¶

图的一个常见状态是消息列表。通常，您只会将消息添加到该状态。但是，有时您可能希望删除消息（通过直接修改状态或作为图的一部分）。为此，您可以使用 RemoveMessage 修饰符。在本指南中，我们将介绍如何做到这一点。

关键思想是，每个状态键都具有一个 reducer 键。该键指定如何将状态更新组合在一起。默认的 MessagesState 具有一个 messages 键，该键的 reducer 接受这些 RemoveMessage 修饰符。然后，该 reducer 使用这些 RemoveMessage 从键中删除消息。

因此请注意，即使您的图状态具有一个作为消息列表的键，也不意味着此 RemoveMessage 修饰符将起作用。您还必须定义一个知道如何使用此修饰符的 reducer。

注意：许多模型期望消息列表围绕某些规则。例如，有些模型期望它们以 user 消息开头，其他模型期望所有带有工具调用的消息后面都跟着工具消息。删除消息时，您需要确保不会违反这些规则。

设置¶

首先，让我们构建一个使用消息的简单图。请注意，它使用 MessagesState，它具有所需的 reducer。

在 [1]

已复制！

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

接下来，我们需要设置 Anthropic（我们将使用的 LLM）的 API 密钥

在 [2]

已复制！

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")
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")

ANTHROPIC_API_KEY:  ········

为 LangGraph 开发设置 LangSmith

注册 LangSmith 以快速发现问题并提高 LangGraph 项目的性能。LangSmith 允许您使用跟踪数据调试、测试和监控使用 LangGraph 构建的 LLM 应用程序——详细了解如何开始此处。

构建代理¶

现在让我们构建一个简单的 ReAct 样式代理。

在 [7]

已复制！





from typing import Literal

from langchain_anthropic import ChatAnthropic
from langchain_core.tools import tool

from langgraph.checkpoint.memory import MemorySaver
from langgraph.graph import MessagesState, StateGraph, START, END
from langgraph.prebuilt import ToolNode

memory = MemorySaver()


@tool
def search(query: str):
    """Call to surf the web."""
    # This is a placeholder for the actual implementation
    # Don't let the LLM know this though 😊
    return "It's sunny in San Francisco, but you better look out if you're a Gemini 😈."


tools = [search]
tool_node = ToolNode(tools)
model = ChatAnthropic(model_name="claude-3-haiku-20240307")
bound_model = model.bind_tools(tools)


def should_continue(state: MessagesState):
    """Return the next node to execute."""
    last_message = state["messages"][-1]
    # If there is no function call, then we finish
    if not last_message.tool_calls:
        return END
    # Otherwise if there is, we continue
    return "action"


# Define the function that calls the model
def call_model(state: MessagesState):
    response = model.invoke(state["messages"])
    # We return a list, because this will get added to the existing list
    return {"messages": response}


# Define a new graph
workflow = StateGraph(MessagesState)

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

# Set the entrypoint as `agent`
# This means that this node is the first one called
workflow.add_edge(START, "agent")

# We now add a conditional edge
workflow.add_conditional_edges(
    # First, we define the start node. We use `agent`.
    # This means these are the edges taken after the `agent` node is called.
    "agent",
    # Next, we pass in the function that will determine which node is called next.
    should_continue,
    # Next, we pass in the path map - all the possible nodes this edge could go to
    ["action", END],
)

# We now add a normal edge from `tools` to `agent`.
# This means that after `tools` is called, `agent` node is called next.
workflow.add_edge("action", "agent")

# Finally, we compile it!
# This compiles it into a LangChain Runnable,
# meaning you can use it as you would any other runnable
app = workflow.compile(checkpointer=memory)
from typing import Literal from langchain_anthropic import ChatAnthropic from langchain_core.tools import tool from langgraph.checkpoint.memory import MemorySaver from langgraph.graph import MessagesState, StateGraph, START, END from langgraph.prebuilt import ToolNode memory = MemorySaver() @tool def search(query: str): """调用浏览网络。""" # 这是实际实现的占位符 # 不要让 LLM 知道这一点 😊 return "旧金山阳光明媚，但如果您是双子座，您最好小心😈。" tools = [search] tool_node = ToolNode(tools) model = ChatAnthropic(model_name="claude-3-haiku-20240307") bound_model = model.bind_tools(tools) def should_continue(state: MessagesState): """返回要执行的下一个节点。""" last_message = state["messages"][-1] # 如果没有函数调用，则结束 if not last_message.tool_calls: return END # 否则，如果有，则继续 return "action" # 定义调用模型的函数 def call_model(state: MessagesState): response = model.invoke(state["messages"]) # 我们返回一个列表，因为这将添加到现有列表中 return {"messages": response} # 定义新的图 workflow = StateGraph(MessagesState) # 定义我们将循环的两个节点 workflow.add_node("agent", call_model) workflow.add_node("action", tool_node) # 将入口点设置为 `agent` # 这意味着该节点是第一个被调用的节点 workflow.add_edge(START, "agent") # 现在添加一个条件边 workflow.add_conditional_edges( # 首先，我们定义起始节点。我们使用 `agent`。 # 这意味着这些是在调用 `agent` 节点后采用的边。 "agent", # 接下来，我们传入一个函数，该函数将确定调用哪个节点。 should_continue, # 接下来，我们传入路径映射——此边可能通往的所有可能节点 ["action", END], ) # 现在从 `tools` 添加一个正常边到 `agent`。 # 这意味着在调用 `tools` 后，将调用 `agent` 节点。 workflow.add_edge("action", "agent") # 最后，我们编译它！ # 这将它编译成一个 LangChain 可运行文件， # 意味着您可以像使用任何其他可运行文件一样使用它 app = workflow.compile(checkpointer=memory)

在 [8]

已复制！

from langchain_core.messages import HumanMessage

config = {"configurable": {"thread_id": "2"}}
input_message = HumanMessage(content="hi! I'm bob")
for event in app.stream({"messages": [input_message]}, config, stream_mode="values"):
    event["messages"][-1].pretty_print()

input_message = HumanMessage(content="what's my name?")
for event in app.stream({"messages": [input_message]}, config, stream_mode="values"):
    event["messages"][-1].pretty_print()
from langchain_core.messages import HumanMessage config = {"configurable": {"thread_id": "2"}} input_message = HumanMessage(content="hi! I'm bob") for event in app.stream({"messages": [input_message]}, config, stream_mode="values"): event["messages"][-1].pretty_print() input_message = HumanMessage(content="what's my name?") for event in app.stream({"messages": [input_message]}, config, stream_mode="values"): event["messages"][-1].pretty_print()

================================ Human Message =================================

hi! I'm bob
================================== Ai Message ==================================

It's nice to meet you, Bob! I'm an AI assistant created by Anthropic. I'm here to help out with any questions or tasks you might have. Please let me know if there's anything I can assist you with.
================================ Human Message =================================

what's my name?
================================== Ai Message ==================================

You said your name is Bob.

手动删除消息¶

首先，我们将介绍如何手动删除消息。让我们看一下线程的当前状态

在 [9]

已复制！

messages = app.get_state(config).values["messages"]
messages
messages = app.get_state(config).values["messages"] messages

输出[9]

[HumanMessage(content="hi! I'm bob", additional_kwargs={}, response_metadata={}, id='db576005-3a60-4b3b-8925-dc602ac1c571'),
 AIMessage(content="It's nice to meet you, Bob! I'm an AI assistant created by Anthropic. I'm here to help out with any questions or tasks you might have. Please let me know if there's anything I can assist you with.", additional_kwargs={}, response_metadata={'id': 'msg_01BKAnYxmoC6bQ9PpCuHk8ZT', 'model': 'claude-3-haiku-20240307', 'stop_reason': 'end_turn', 'stop_sequence': None, 'usage': {'input_tokens': 12, 'output_tokens': 52}}, id='run-3a60c536-b207-4c56-98f3-03f94d49a9e4-0', usage_metadata={'input_tokens': 12, 'output_tokens': 52, 'total_tokens': 64}),
 HumanMessage(content="what's my name?", additional_kwargs={}, response_metadata={}, id='2088c465-400b-430b-ad80-fad47dc1f2d6'),
 AIMessage(content='You said your name is Bob.', additional_kwargs={}, response_metadata={'id': 'msg_013UWTLTzwZi81vke8mMQ2KP', 'model': 'claude-3-haiku-20240307', 'stop_reason': 'end_turn', 'stop_sequence': None, 'usage': {'input_tokens': 72, 'output_tokens': 10}}, id='run-3a6883be-0c52-4938-af98-e9e7476659eb-0', usage_metadata={'input_tokens': 72, 'output_tokens': 10, 'total_tokens': 82})]

我们可以调用 update_state 并传入第一条消息的 id。这将删除该消息。

在 [10]

已复制！

from langchain_core.messages import RemoveMessage

app.update_state(config, {"messages": RemoveMessage(id=messages[0].id)})
from langchain_core.messages import RemoveMessage app.update_state(config, {"messages": RemoveMessage(id=messages[0].id)})

输出[10]

{'configurable': {'thread_id': '2',
  'checkpoint_ns': '',
  'checkpoint_id': '1ef75157-f251-6a2a-8005-82a86a6593a0'}}

如果我们现在看一下消息，我们可以验证第一条消息是否已被删除。

在 [8]

已复制！

messages = app.get_state(config).values["messages"]
messages
messages = app.get_state(config).values["messages"] messages

输出[8]

[AIMessage(content="It's nice to meet you, Bob! I'm Claude, an AI assistant created by Anthropic. How can I assist you today?", response_metadata={'id': 'msg_01XPSAenmSqK8rX2WgPZHfz7', 'model': 'claude-3-haiku-20240307', 'stop_reason': 'end_turn', 'stop_sequence': None, 'usage': {'input_tokens': 12, 'output_tokens': 32}}, id='run-1c69af09-adb1-412d-9010-2456e5a555fb-0', usage_metadata={'input_tokens': 12, 'output_tokens': 32, 'total_tokens': 44}),
 HumanMessage(content="what's my name?", id='f3c71afe-8ce2-4ed0-991e-65021f03b0a5'),
 AIMessage(content='Your name is Bob, as you introduced yourself at the beginning of our conversation.', response_metadata={'id': 'msg_01BPZdwsjuMAbC1YAkqawXaF', 'model': 'claude-3-haiku-20240307', 'stop_reason': 'end_turn', 'stop_sequence': None, 'usage': {'input_tokens': 52, 'output_tokens': 19}}, id='run-b2eb9137-2f4e-446f-95f5-3d5f621a2cf8-0', usage_metadata={'input_tokens': 52, 'output_tokens': 19, 'total_tokens': 71})]

以编程方式删除消息¶

我们还可以以编程方式从图中删除消息。在这里，我们将修改图以在图运行结束时删除所有旧消息（超过 3 条消息前的消息）。

在 [9]

已复制！





from langchain_core.messages import RemoveMessage
from langgraph.graph import END


def delete_messages(state):
    messages = state["messages"]
    if len(messages) > 3:
        return {"messages": [RemoveMessage(id=m.id) for m in messages[:-3]]}


# We need to modify the logic to call delete_messages rather than end right away
def should_continue(state: MessagesState) -> Literal["action", "delete_messages"]:
    """Return the next node to execute."""
    last_message = state["messages"][-1]
    # If there is no function call, then we call our delete_messages function
    if not last_message.tool_calls:
        return "delete_messages"
    # Otherwise if there is, we continue
    return "action"


# Define a new graph
workflow = StateGraph(MessagesState)
workflow.add_node("agent", call_model)
workflow.add_node("action", tool_node)

# This is our new node we're defining
workflow.add_node(delete_messages)


workflow.add_edge(START, "agent")
workflow.add_conditional_edges(
    "agent",
    should_continue,
)
workflow.add_edge("action", "agent")

# This is the new edge we're adding: after we delete messages, we finish
workflow.add_edge("delete_messages", END)
app = workflow.compile(checkpointer=memory)
from langchain_core.messages import RemoveMessage from langgraph.graph import END def delete_messages(state): messages = state["messages"] if len(messages) > 3: return {"messages": [RemoveMessage(id=m.id) for m in messages[:-3]]} # 我们需要修改逻辑以调用 delete_messages 而不是立即结束 def should_continue(state: MessagesState) -> Literal["action", "delete_messages"]: """返回要执行的下一个节点。""" last_message = state["messages"][-1] # 如果没有函数调用，则调用我们的 delete_messages 函数 if not last_message.tool_calls: return "delete_messages" # 否则，如果有，则继续 return "action" # 定义新的图 workflow = StateGraph(MessagesState) workflow.add_node("agent", call_model) workflow.add_node("action", tool_node) # 这是我们正在定义的新节点 workflow.add_node(delete_messages) workflow.add_edge(START, "agent") workflow.add_conditional_edges( "agent", should_continue, ) workflow.add_edge("action", "agent") # 这是我们正在添加的新边：删除消息后，我们结束 workflow.add_edge("delete_messages", END) app = workflow.compile(checkpointer=memory)

现在我们可以尝试一下。我们可以调用图两次，然后检查状态

在 [10]

已复制！

from langchain_core.messages import HumanMessage

config = {"configurable": {"thread_id": "3"}}
input_message = HumanMessage(content="hi! I'm bob")
for event in app.stream({"messages": [input_message]}, config, stream_mode="values"):
    print([(message.type, message.content) for message in event["messages"]])

input_message = HumanMessage(content="what's my name?")
for event in app.stream({"messages": [input_message]}, config, stream_mode="values"):
    print([(message.type, message.content) for message in event["messages"]])
from langchain_core.messages import HumanMessage config = {"configurable": {"thread_id": "3"}} input_message = HumanMessage(content="hi! I'm bob") for event in app.stream({"messages": [input_message]}, config, stream_mode="values"): print([(message.type, message.content) for message in event["messages"]]) input_message = HumanMessage(content="what's my name?") for event in app.stream({"messages": [input_message]}, config, stream_mode="values"): print([(message.type, message.content) for message in event["messages"]])

[('human', "hi! I'm bob")]
[('human', "hi! I'm bob"), ('ai', "Hello Bob! It's nice to meet you. I'm an AI assistant created by Anthropic. I'm here to help with any questions or tasks you might have. Please let me know how I can assist you.")]
[('human', "hi! I'm bob"), ('ai', "Hello Bob! It's nice to meet you. I'm an AI assistant created by Anthropic. I'm here to help with any questions or tasks you might have. Please let me know how I can assist you."), ('human', "what's my name?")]
[('human', "hi! I'm bob"), ('ai', "Hello Bob! It's nice to meet you. I'm an AI assistant created by Anthropic. I'm here to help with any questions or tasks you might have. Please let me know how I can assist you."), ('human', "what's my name?"), ('ai', 'You said your name is Bob, so that is the name I have for you.')]
[('ai', "Hello Bob! It's nice to meet you. I'm an AI assistant created by Anthropic. I'm here to help with any questions or tasks you might have. Please let me know how I can assist you."), ('human', "what's my name?"), ('ai', 'You said your name is Bob, so that is the name I have for you.')]

如果我们现在检查状态，我们应该看到它只有三条消息长。这是因为我们刚刚删除了之前的消息——否则它应该是四条！

在 [11]

已复制！

messages = app.get_state(config).values["messages"]
messages
messages = app.get_state(config).values["messages"] messages

输出[11]

[AIMessage(content="Hello Bob! It's nice to meet you. I'm an AI assistant created by Anthropic. I'm here to help with any questions or tasks you might have. Please let me know how I can assist you.", response_metadata={'id': 'msg_01XPEgPPbcnz5BbGWUDWTmzG', 'model': 'claude-3-haiku-20240307', 'stop_reason': 'end_turn', 'stop_sequence': None, 'usage': {'input_tokens': 12, 'output_tokens': 48}}, id='run-eded3820-b6a9-4d66-9210-03ca41787ce6-0', usage_metadata={'input_tokens': 12, 'output_tokens': 48, 'total_tokens': 60}),
 HumanMessage(content="what's my name?", id='a0ea2097-3280-402b-92e1-67177b807ae8'),
 AIMessage(content='You said your name is Bob, so that is the name I have for you.', response_metadata={'id': 'msg_01JGT62pxhrhN4SykZ57CSjW', 'model': 'claude-3-haiku-20240307', 'stop_reason': 'end_turn', 'stop_sequence': None, 'usage': {'input_tokens': 68, 'output_tokens': 20}}, id='run-ace3519c-81f8-45fe-a777-91f42d48b3a3-0', usage_metadata={'input_tokens': 68, 'output_tokens': 20, 'total_tokens': 88})]

请记住，删除消息时，您需要确保剩余的消息列表仍然有效。此消息列表实际上可能不是——这是因为它当前以 AI 消息开头，而某些模型不允许这样做。

如何删除消息¶

设置¶

构建代理¶

手动删除消息¶

以编程方式删除消息¶

评论