from [langchain.tools](http://langchain.tools) import tool
from langchain.agents import create_agent

subagent1 = create_agent(model="...", tools=[...])

@tool(
    "subagent1_name",
    description="subagent1_description"
)
def call_subagent1(query: str):
    result = subagent1.invoke({
        "messages": [{"role": "user", "content": query}]
    })
    return result["messages"][-1].content

agent = create_agent(model="...", tools=[call_subagent1])

from langchain.agents import AgentState
from [langchain.tools](http://langchain.tools) import tool, ToolRuntime

class CustomState(AgentState):
    example_state_key: str

@tool(
    "subagent1_name",
    description="subagent1_description"
)
def call_subagent1(query: str, runtime: ToolRuntime[None, CustomState]):
    # 쿼리 메세지를 적절한 인풋으로 수정하는 어떤 로직
    subagent_input = some_logic(query, runtime.state["messages"])
    result = subagent1.invoke({
        "messages": subagent_input,
        # 필요한 경우 다른 상태 키도 전달 가능
        # 상태 키의 경우 메인 에이전트와 서브에이전트의 스키마에 추가하는 것 잊지 말 것
        "example_state_key": runtime.state["example_state_key"]
    })
    return result["messages"][-1].content

from typing import Annotated
from langchain.agents import AgentState
from [langchain.tools](http://langchain.tools) import InjectedToolCallId
from langgraph.types import Command

@tool(
    "subagent1_name",
    description="subagent1_description"
)
# 서브에이전트가 도구 호출 결과로 응답할 수 있도록 tool_call_id 전달

def call_subagent1(
    query: str,
    tool_call_id: Annotated[str, InjectedToolCallId],
    # 텍스트 이상의 내용을 반환하려면 Command 객체를 반환해야 함
) -> Command:
    result = subagent1.invoke({
        "messages": [{"role": "user", "content": query}]
    })
    return Command(update={
        # 메인 에이전트에 반환할 상태 키 예시
        "example_state_key": result["example_state_key"],
        "messages": [
            ToolMessage(
                content=result["messages"][-1].content,
                # 올바른 도구 호출과 매칭을 위해 tool_call_id를 포함
                tool_call_id=tool_call_id
            )
        ]
    })

def research_manager_node(state) -> dict:
        history = state["investment_debate_state"].get("history", "")
        market_research_report = state["market_report"]
        sentiment_report = state["sentiment_report"]
        news_report = state["news_report"]
        fundamentals_report = state["fundamentals_report"]

        investment_debate_state = state["investment_debate_state"]

        curr_situation = f"{market_research_report}\n\n{sentiment_report}\n\n{news_report}\n\n{fundamentals_report}"
        past_memories = memory.get_memories(curr_situation, n_matches=2)

        past_memory_str = ""
        for i, rec in enumerate(past_memories, 1):
            past_memory_str += rec["recommendation"] + "\n\n"

        prompt = f"""As the portfolio manager and debate facilitator, your role is to critically evaluate this round of debate and make a definitive decision: align with the bear analyst, the bull analyst, or choose Hold only if it is strongly justified based on the arguments presented.

                                           Summarize the key points from both sides concisely, focusing on the most compelling evidence or reasoning. Your recommendation—Buy, Sell, or Hold—must be clear and actionable. Avoid defaulting to Hold simply because both sides have valid points; commit to a stance grounded in the debate's strongest arguments.
                                           
                                           Additionally, develop a detailed investment plan for the trader. This should include:
                                           
                                           Your Recommendation: A decisive stance supported by the most convincing arguments.
                                           Rationale: An explanation of why these arguments lead to your conclusion.
                                           Strategic Actions: Concrete steps for implementing the recommendation.
                                           Take into account your past mistakes on similar situations. Use these insights to refine your decision-making and ensure you are learning and improving. Present your analysis conversationally, as if speaking naturally, without special formatting. 
                                           
                                           Here are your past reflections on mistakes:
                                           \"{past_memory_str}\"
                                           
                                           Here is the debate:
                                           Debate History:
                                           {history}"""
        response = llm.invoke(prompt)

        new_investment_debate_state = {
            "judge_decision": response.content,
            "history": investment_debate_state.get("history", ""),
            "bear_history": investment_debate_state.get("bear_history", ""),
            "bull_history": investment_debate_state.get("bull_history", ""),
            "current_response": response.content,
            "count": investment_debate_state["count"],
        }

        return {
            "investment_debate_state": new_investment_debate_state,
            "investment_plan": response.content,
        }