自定义状态¶
在本教程中,你将向状态添加其他字段,以定义复杂行为而不依赖消息列表。聊天机器人将使用其搜索工具查找特定信息并将其转发给人类进行审查。
Note
本教程基于添加人机协同控制。
1. 向状态添加键¶
通过向状态添加 name 和 birthday 键来更新聊天机器人以研究实体的生日:
from typing import Annotated
from typing_extensions import TypedDict
from langgraph.graph.message import add_messages
class State(TypedDict):
messages: Annotated[list, add_messages]
# highlight-next-line
name: str
# highlight-next-line
birthday: str
import { MessagesZodState } from "@langchain/langgraph";
import { z } from "zod";
const State = z.object({
messages: MessagesZodState.shape.messages,
// highlight-next-line
name: z.string(),
// highlight-next-line
birthday: z.string(),
});
将此信息添加到状态中，使其可以轻松被其他图节点（如存储或处理信息的下游节点）以及图的持久层访问。
2. Update the state inside the tool¶
现在，在 human_assistance 工具内部填充状态键。这允许人类在信息存储到状态之前进行审查。使用 Command 从工具内部发出状态更新。
from langchain_core.messages import ToolMessage
from langchain_core.tools import InjectedToolCallId, tool
from langgraph.types import Command, interrupt
@tool
# Note that because we are generating a ToolMessage for a state update, we
# generally require the ID of the corresponding tool call. We can use
# LangChain's InjectedToolCallId to signal that this argument should not
# be revealed to the model in the tool's schema.
def human_assistance(
name: str, birthday: str, tool_call_id: Annotated[str, InjectedToolCallId]
) -> str:
"""Request assistance from a human."""
human_response = interrupt(
{
"question": "Is this correct?",
"name": name,
"birthday": birthday,
},
)
# If the information is correct, update the state as-is.
if human_response.get("correct", "").lower().startswith("y"):
verified_name = name
verified_birthday = birthday
response = "Correct"
# Otherwise, receive information from the human reviewer.
else:
verified_name = human_response.get("name", name)
verified_birthday = human_response.get("birthday", birthday)
response = f"Made a correction: {human_response}"
# This time we explicitly update the state with a ToolMessage inside
# the tool.
state_update = {
"name": verified_name,
"birthday": verified_birthday,
"messages": [ToolMessage(response, tool_call_id=tool_call_id)],
}
# We return a Command object in the tool to update our state.
return Command(update=state_update)
现在，在 humanAssistance 工具内部填充状态键。这允许人类在信息存储到状态之前进行审查。使用 Command 从工具内部发出状态更新。
import { tool } from "@langchain/core/tools";
import { ToolMessage } from "@langchain/core/messages";
import { Command, interrupt } from "@langchain/langgraph";
const humanAssistance = tool(
async (input, config) => {
// Note that because we are generating a ToolMessage for a state update,
// we generally require the ID of the corresponding tool call.
// This is available in the tool's config.
const toolCallId = config?.toolCall?.id as string | undefined;
if (!toolCallId) throw new Error("Tool call ID is required");
const humanResponse = await interrupt({
question: "Is this correct?",
name: input.name,
birthday: input.birthday,
});
// We explicitly update the state with a ToolMessage inside the tool.
const stateUpdate = (() => {
// If the information is correct, update the state as-is.
if (humanResponse.correct?.toLowerCase().startsWith("y")) {
return {
name: input.name,
birthday: input.birthday,
messages: [
new ToolMessage({ content: "Correct", tool_call_id: toolCallId }),
],
};
}
// Otherwise, receive information from the human reviewer.
return {
name: humanResponse.name || input.name,
birthday: humanResponse.birthday || input.birthday,
messages: [
new ToolMessage({
content: `Made a correction: ${JSON.stringify(humanResponse)}`,
tool_call_id: toolCallId,
}),
],
};
})();
// We return a Command object in the tool to update our state.
return new Command({ update: stateUpdate });
},
{
name: "humanAssistance",
description: "Request assistance from a human.",
schema: z.object({
name: z.string().describe("The name of the entity"),
birthday: z.string().describe("The birthday/release date of the entity"),
}),
}
);
The rest of the graph stays the same.
3. Prompt the chatbot¶
提示聊天机器人查找 LangGraph 库的"生日"，并在获得所需信息后指示聊天机器人访问 human_assistance 工具。通过在工具的参数中设置 name 和 birthday，你可以强制聊天机器人为这些字段生成建议。
user_input = (
"Can you look up when LangGraph was released? "
"When you have the answer, use the human_assistance tool for review."
)
config = {"configurable": {"thread_id": "1"}}
events = graph.stream(
{"messages": [{"role": "user", "content": user_input}]},
config,
stream_mode="values",
)
for event in events:
if "messages" in event:
event["messages"][-1].pretty_print()
提示聊天机器人查找 LangGraph 库的"生日"，并在获得所需信息后指示聊天机器人访问 humanAssistance 工具。通过在工具的参数中设置 name 和 birthday，你可以强制聊天机器人为这些字段生成建议。
import { isAIMessage } from "@langchain/core/messages";
const userInput =
"Can you look up when LangGraph was released? " +
"When you have the answer, use the humanAssistance tool for review.";
const events = await graph.stream(
{ messages: [{ role: "user", content: userInput }] },
{ configurable: { thread_id: "1" }, streamMode: "values" }
);
for await (const event of events) {
if ("messages" in event) {
const lastMessage = event.messages.at(-1);
console.log(
"=".repeat(32),
`${lastMessage?.getType()} Message`,
"=".repeat(32)
);
console.log(lastMessage?.text);
if (
lastMessage &&
isAIMessage(lastMessage) &&
lastMessage.tool_calls?.length
) {
console.log("Tool Calls:");
for (const call of lastMessage.tool_calls) {
console.log(` ${call.name} (${call.id})`);
console.log(` Args: ${JSON.stringify(call.args)}`);
}
}
}
}
================================ Human Message =================================
Can you look up when LangGraph was released? When you have the answer, use the human_assistance tool for review.
================================== Ai Message ==================================
[{'text': "Certainly! I'll start by searching for information about LangGraph's release date using the Tavily search function. Then, I'll use the human_assistance tool for review.", 'type': 'text'}, {'id': 'toolu_01JoXQPgTVJXiuma8xMVwqAi', 'input': {'query': 'LangGraph release date'}, 'name': 'tavily_search_results_json', 'type': 'tool_use'}]
Tool Calls:
tavily_search_results_json (toolu_01JoXQPgTVJXiuma8xMVwqAi)
Call ID: toolu_01JoXQPgTVJXiuma8xMVwqAi
Args:
query: LangGraph release date
================================= Tool Message =================================
Name: tavily_search_results_json
[{"url": "https://blog.langchain.dev/langgraph-cloud/", "content": "We also have a new stable release of LangGraph. By LangChain 6 min read Jun 27, 2024 (Oct '24) Edit: Since the launch of LangGraph Platform, we now have multiple deployment options alongside LangGraph Studio - which now fall under LangGraph Platform. LangGraph Platform is synonymous with our Cloud SaaS deployment option."}, {"url": "https://changelog.langchain.com/announcements/langgraph-cloud-deploy-at-scale-monitor-carefully-iterate-boldly", "content": "LangChain - Changelog | ☁ 🚀 LangGraph Platform: Deploy at scale, monitor LangChain LangSmith LangGraph LangChain LangSmith LangGraph LangChain LangSmith LangGraph LangChain Changelog Sign up for our newsletter to stay up to date DATE: The LangChain Team LangGraph LangGraph Platform ☁ 🚀 LangGraph Platform: Deploy at scale, monitor carefully, iterate boldly DATE: June 27, 2024 AUTHOR: The LangChain Team LangGraph Platform is now in closed beta, offering scalable, fault-tolerant deployment for LangGraph agents. LangGraph Platform also includes a new playground-like studio for debugging agent failure modes and quick iteration: Join the waitlist today for LangGraph Platform. And to learn more, read our blog post announcement or check out our docs. Subscribe By clicking subscribe, you accept our privacy policy and terms and conditions."}]
================================== Ai Message ==================================
[{'text': "Based on the search results, it appears that LangGraph was already in existence before June 27, 2024, when LangGraph Platform was announced. However, the search results don't provide a specific release date for the original LangGraph. \n\nGiven this information, I'll use the human_assistance tool to review and potentially provide more accurate information about LangGraph's initial release date.", 'type': 'text'}, {'id': 'toolu_01JDQAV7nPqMkHHhNs3j3XoN', 'input': {'name': 'Assistant', 'birthday': '2023-01-01'}, 'name': 'human_assistance', 'type': 'tool_use'}]
Tool Calls:
human_assistance (toolu_01JDQAV7nPqMkHHhNs3j3XoN)
Call ID: toolu_01JDQAV7nPqMkHHhNs3j3XoN
Args:
name: Assistant
birthday: 2023-01-01
我们再次触发了 human_assistance 工具中的 interrupt。
我们再次触发了 humanAssistance 工具中的 interrupt。
4. Add human assistance¶
聊天机器人未能识别正确的日期，因此向它提供信息：
human_command = Command(
resume={
"name": "LangGraph",
"birthday": "Jan 17, 2024",
},
)
events = graph.stream(human_command, config, stream_mode="values")
for event in events:
if "messages" in event:
event["messages"][-1].pretty_print()
import { Command } from "@langchain/langgraph";
const humanCommand = new Command({
resume: {
name: "LangGraph",
birthday: "Jan 17, 2024",
},
});
const resumeEvents = await graph.stream(humanCommand, {
configurable: { thread_id: "1" },
streamMode: "values",
});
for await (const event of resumeEvents) {
if ("messages" in event) {
const lastMessage = event.messages.at(-1);
console.log(
"=".repeat(32),
`${lastMessage?.getType()} Message`,
"=".repeat(32)
);
console.log(lastMessage?.text);
if (
lastMessage &&
isAIMessage(lastMessage) &&
lastMessage.tool_calls?.length
) {
console.log("Tool Calls:");
for (const call of lastMessage.tool_calls) {
console.log(` ${call.name} (${call.id})`);
console.log(` Args: ${JSON.stringify(call.args)}`);
}
}
}
}
================================== Ai Message ==================================
[{'text': "Based on the search results, it appears that LangGraph was already in existence before June 27, 2024, when LangGraph Platform was announced. However, the search results don't provide a specific release date for the original LangGraph. \n\nGiven this information, I'll use the human_assistance tool to review and potentially provide more accurate information about LangGraph's initial release date.", 'type': 'text'}, {'id': 'toolu_01JDQAV7nPqMkHHhNs3j3XoN', 'input': {'name': 'Assistant', 'birthday': '2023-01-01'}, 'name': 'human_assistance', 'type': 'tool_use'}]
Tool Calls:
human_assistance (toolu_01JDQAV7nPqMkHHhNs3j3XoN)
Call ID: toolu_01JDQAV7nPqMkHHhNs3j3XoN
Args:
name: Assistant
birthday: 2023-01-01
================================= Tool Message =================================
Name: human_assistance
Made a correction: {'name': 'LangGraph', 'birthday': 'Jan 17, 2024'}
================================== Ai Message ==================================
Thank you for the human assistance. I can now provide you with the correct information about LangGraph's release date.
LangGraph was initially released on January 17, 2024. This information comes from the human assistance correction, which is more accurate than the search results I initially found.
To summarize:
1. LangGraph's original release date: January 17, 2024
2. LangGraph Platform announcement: June 27, 2024
It's worth noting that LangGraph had been in development and use for some time before the LangGraph Platform announcement, but the official initial release of LangGraph itself was on January 17, 2024.
请注意，这些字段现在反映在状态中：
snapshot = graph.get_state(config)
{k: v for k, v in snapshot.values.items() if k in ("name", "birthday")}
const snapshot = await graph.getState(config);
const relevantState = Object.fromEntries(
Object.entries(snapshot.values).filter(([k]) =>
["name", "birthday"].includes(k)
)
);
这使得下游节点（例如进一步处理或存储信息的节点）可以轻松访问它们。
5. Manually update the state¶
LangGraph 对应用程序状态提供了高度控制。例如，在任何时候（包括中断时），你都可以使用 graph.update_state 手动覆盖一个键：
{'configurable': {'thread_id': '1',
'checkpoint_ns': '',
'checkpoint_id': '1efd4ec5-cf69-6352-8006-9278f1730162'}}
LangGraph 对应用程序状态提供了高度控制。例如，在任何时候（包括中断时），你都可以使用 graph.updateState 手动覆盖一个键：
{
configurable: {
thread_id: '1',
checkpoint_ns: '',
checkpoint_id: '1efd4ec5-cf69-6352-8006-9278f1730162'
}
}
6. View the new value¶
如果你调用 graph.get_state，你可以看到新值已反映：
snapshot = graph.get_state(config)
{k: v for k, v in snapshot.values.items() if k in ("name", "birthday")}
如果你调用 graph.getState，你可以看到新值已反映：
const updatedSnapshot = await graph.getState(config);
const updatedRelevantState = Object.fromEntries(
Object.entries(updatedSnapshot.values).filter(([k]) =>
["name", "birthday"].includes(k)
)
);
手动状态更新将在 LangSmith 中生成追踪。如果需要，它们也可以用于控制人机交互工作流。通常建议使用 interrupt 函数，因为它允许在人机交互中独立于状态更新传输数据。
恭喜！ 你已向状态添加了自定义键以促进更复杂的工作流，并学会了如何从工具内部生成状态更新。
查看下面的代码片段以回顾本教程中的图：
from typing import Annotated
from langchain_tavily import TavilySearch
from langchain_core.messages import ToolMessage
from langchain_core.tools import InjectedToolCallId, tool
from typing_extensions import TypedDict
from langgraph.checkpoint.memory import InMemorySaver
from langgraph.graph import StateGraph, START, END
from langgraph.graph.message import add_messages
from langgraph.prebuilt import ToolNode, tools_condition
from langgraph.types import Command, interrupt
class State(TypedDict):
messages: Annotated[list, add_messages]
name: str
birthday: str
@tool
def human_assistance(
name: str, birthday: str, tool_call_id: Annotated[str, InjectedToolCallId]
) -> str:
"""Request assistance from a human."""
human_response = interrupt(
{
"question": "Is this correct?",
"name": name,
"birthday": birthday,
},
)
if human_response.get("correct", "").lower().startswith("y"):
verified_name = name
verified_birthday = birthday
response = "Correct"
else:
verified_name = human_response.get("name", name)
verified_birthday = human_response.get("birthday", birthday)
response = f"Made a correction: {human_response}"
state_update = {
"name": verified_name,
"birthday": verified_birthday,
"messages": [ToolMessage(response, tool_call_id=tool_call_id)],
}
return Command(update=state_update)
tool = TavilySearch(max_results=2)
tools = [tool, human_assistance]
llm_with_tools = llm.bind_tools(tools)
def chatbot(state: State):
message = llm_with_tools.invoke(state["messages"])
assert(len(message.tool_calls) <= 1)
return {"messages": [message]}
graph_builder = StateGraph(State)
graph_builder.add_node("chatbot", chatbot)
tool_node = ToolNode(tools=tools)
graph_builder.add_node("tools", tool_node)
graph_builder.add_conditional_edges(
"chatbot",
tools_condition,
)
graph_builder.add_edge("tools", "chatbot")
graph_builder.add_edge(START, "chatbot")
memory = InMemorySaver()
graph = graph_builder.compile(checkpointer=memory)
import {
Command,
interrupt,
MessagesZodState,
MemorySaver,
StateGraph,
END,
START,
} from "@langchain/langgraph";
import { ToolNode, toolsCondition } from "@langchain/langgraph/prebuilt";
import { ChatAnthropic } from "@langchain/anthropic";
import { TavilySearch } from "@langchain/tavily";
import { ToolMessage } from "@langchain/core/messages";
import { tool } from "@langchain/core/tools";
import { z } from "zod";
const State = z.object({
messages: MessagesZodState.shape.messages,
name: z.string(),
birthday: z.string(),
});
const humanAssistance = tool(
async (input, config) => {
// Note that because we are generating a ToolMessage for a state update, we
// generally require the ID of the corresponding tool call. This is available
// in the tool's config.
const toolCallId = config?.toolCall?.id as string | undefined;
if (!toolCallId) throw new Error("Tool call ID is required");
const humanResponse = await interrupt({
question: "Is this correct?",
name: input.name,
birthday: input.birthday,
});
// We explicitly update the state with a ToolMessage inside the tool.
const stateUpdate = (() => {
// If the information is correct, update the state as-is.
if (humanResponse.correct?.toLowerCase().startsWith("y")) {
return {
name: input.name,
birthday: input.birthday,
messages: [
new ToolMessage({ content: "Correct", tool_call_id: toolCallId }),
],
};
}
// Otherwise, receive information from the human reviewer.
return {
name: humanResponse.name || input.name,
birthday: humanResponse.birthday || input.birthday,
messages: [
new ToolMessage({
content: `Made a correction: ${JSON.stringify(humanResponse)}`,
tool_call_id: toolCallId,
}),
],
};
})();
// We return a Command object in the tool to update our state.
return new Command({ update: stateUpdate });
},
{
name: "humanAssistance",
description: "Request assistance from a human.",
schema: z.object({
name: z.string().describe("The name of the entity"),
birthday: z.string().describe("The birthday/release date of the entity"),
}),
}
);
const searchTool = new TavilySearch({ maxResults: 2 });
const tools = [searchTool, humanAssistance];
const llmWithTools = new ChatAnthropic({
model: "claude-3-5-sonnet-latest",
}).bindTools(tools);
const memory = new MemorySaver();
const chatbot = async (state: z.infer<typeof State>) => {
const message = await llmWithTools.invoke(state.messages);
return { messages: message };
};
const graph = new StateGraph(State)
.addNode("chatbot", chatbot)
.addNode("tools", new ToolNode(tools))
.addConditionalEdges("chatbot", toolsCondition, ["tools", END])
.addEdge("tools", "chatbot")
.addEdge(START, "chatbot")
.compile({ checkpointer: memory });
Next steps¶
在完成 LangGraph 基础教程之前，还有一个概念需要回顾：将 checkpointing 和 state updates 连接到时间旅行。