knowledge-agent/agents/clean_agent/agent.py

from typing import Annotated
from typing_extensions import TypedDict
from langgraph.graph import StateGraph, START, END
from langgraph.graph.message import add_messages
import os
from langchain_openai import ChatOpenAI
from .tools import evaluation_extraction_tool
import uuid

from langgraph.prebuilt import ToolNode, tools_condition
from langgraph.checkpoint.memory import InMemorySaver
import requests
from dotenv import load_dotenv
from utils.logging_config import get_logger

# 配置日志
logger = get_logger('CleanAgent')

# 加载环境变量
load_dotenv()

graph=None
llm_with_tools=None

prompt="""
### 角色 (Role)：
您是一个专业的评估报告检索助手，我的任务是根据用户的查询关键词，从评估报告中提取相关信息。

### 目标 (Goal)：
1.  根据特定的主题（关键词）快速获取相关的评估报告、数据摘要或关键指标，以便您能深入了解某个方面（如产品表现、服务质量、市场反馈、项目评估等）的详细评估情况。
2.  为您的每次查询提供一个唯一的标识符，以便您能轻松追踪和管理您的请求，确保数据的可追溯性。

---
### 工作流 (Workflow)：
1.  从输入信息中提取关键词（query_word）和请求ID（request_id）
2.  调用工具evaluation_extraction_tool，进行评估解析
3.  返回结果

---
### 输入信息：
{input}

### 输出json格式：
{
  "requestId":[请求ID],
  "status":2
}
"""

class State(TypedDict):
    messages: Annotated[list, add_messages]
    name: str
    birthday: str


def chatbot(state: State):
    message = llm_with_tools.invoke(state["messages"])
    # Because we will be interrupting during tool execution,
    # we disable parallel tool calling to avoid repeating any
    # tool invocations when we resume.
    assert len(message.tool_calls) <= 1
    return {"messages": [message]}

def execute_agent_with_api(user_input: str):
            # 生成唯一的线程ID
    import uuid
    thread_id = str(uuid.uuid4())
    logger.info(f"开始执行提取，user_input={user_input}, thread_id={thread_id}")
    global graph, llm_with_tools, prompt
    
    # 替换prompt中的{input}占位符为用户输入
    formatted_prompt = prompt.replace("{input}", user_input)

    try:
        # 如果graph或llm_with_tools未初始化，先初始化
        if graph is None or llm_with_tools is None:
            try:
                # 使用新版本的 ChatOpenAI
                llm = ChatOpenAI(model="gpt-4")
                tools = [evaluation_extraction_tool]
                llm_with_tools = llm.bind_tools(tools=tools)
                
                # 初始化图
                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)
                graph = graph_builder.compile()
            except Exception as e:
                logger.error(f"初始化Agent失败: {str(e)}")
                return f"初始化Agent失败: {str(e)}"
        
        # 执行Agent并收集结果
        results = []
        config = {"configurable": {"thread_id": thread_id}}
        
        # 使用格式化后的prompt作为用户输入
        for event in graph.stream({"messages": [{"role": "user", "content": formatted_prompt}]}, config, stream_mode="values"):
            for value in event.values():
                # 保存消息内容
                if "messages" in event and len(event["messages"]) > 0:
                    message = event["messages"][-1]
                    results.append(message.content)
        
        # 返回结果
        res="\n".join(results) if results else "Agent执行完成，但没有返回结果"
        logger.info(f"Agent执行完成，返回结果: {res}, thread_id={thread_id}")
        return res
    except requests.exceptions.ConnectionError as e:
        return f"OpenAI API 连接错误: {str(e)}\n请检查网络连接或代理设置。"
    except Exception as e:
        return f"执行Agent时出错: {str(e)}"

def execute(query_word: str, request_id: str):
    logger.info(f"开始处理，request_id: {request_id}, query_word: {query_word}")
    result = evaluation_extraction_tool(request_id, query_word)
    return result

def main():
    print(f"开始执行Agent")
    # 设置代理
    proxy_url = os.getenv('DYNAMIC_HTTP_PROXY')
    if proxy_url:
        os.environ["OPENAI_PROXY"] = proxy_url
        os.environ["HTTPS_PROXY"] = proxy_url
        os.environ["HTTP_PROXY"] = proxy_url
    # 执行Agent
    result = execute_agent_with_api('{"query_word":"图文策划方法","request_id":"REQUEST_001"}')
    print(result)

if __name__ == '__main__':
    main()