71 lines
2.1 KiB
Python
71 lines
2.1 KiB
Python
"""
|
||
agent 前后
|
||
model 前后
|
||
工具 中
|
||
模型 中
|
||
"""
|
||
|
||
from langchain.agents import create_agent, AgentState
|
||
from langchain.agents.middleware import before_agent, after_agent, before_model, after_model, wrap_model_call, \
|
||
wrap_tool_call
|
||
from langchain_community.chat_models.tongyi import ChatTongyi
|
||
from langchain_core.tools import tool
|
||
from langgraph.runtime import Runtime
|
||
|
||
|
||
@tool(description="查询天气, 传入城市名称字符串，返回字符串天气信息")
|
||
def get_weather(city: str) -> str:
|
||
return f"{city} : 晴天"
|
||
|
||
|
||
@before_agent
|
||
def log_before_agent(state: AgentState, runtime: Runtime) -> None:
|
||
print(f"before agent: info_num: {len(state["messages"])}")
|
||
|
||
|
||
@after_agent
|
||
def log_after_agent(state: AgentState, runtime: Runtime) -> None:
|
||
print(f"after agent: info_num: {len(state["messages"])}")
|
||
|
||
|
||
@before_model
|
||
def log_before_model(state: AgentState, runtime: Runtime) -> None:
|
||
print(f"before model: info_num: {len(state["messages"])}")
|
||
|
||
|
||
@after_model
|
||
def log_after_model(state: AgentState, runtime: Runtime) -> None:
|
||
print(f"after model: info_num: {len(state["messages"])}")
|
||
|
||
|
||
@wrap_model_call
|
||
def model_call_hook(request, handler):
|
||
print(f"model call: {request}")
|
||
return handler(request)
|
||
|
||
|
||
@wrap_tool_call
|
||
def model_tool_hook(request, handler):
|
||
print(f"model tool: {request.tool_call['name']}")
|
||
print(f"args: {request.tool_call['args']}")
|
||
return handler(request)
|
||
|
||
agent = create_agent(
|
||
model=ChatTongyi(model="qwen3-max"),
|
||
tools=[get_weather],
|
||
middleware=[model_call_hook, model_tool_hook, log_before_model,
|
||
log_after_model, log_before_agent, log_after_agent],
|
||
system_prompt="""你是严格遵循ReAct框架的智能体，必须按[思考，行动，观察，再思考]的流程解决问题
|
||
每轮仅能思考并调用1个工具，禁止单词调用多个工具。并告知我你的思考过程，工具调用的原因，按思考、行动
|
||
、观察三个结构告知我"""
|
||
)
|
||
|
||
res = agent.stream(
|
||
{"messages": [{
|
||
"role": "user", "content": "查询北京的天气"
|
||
}]},
|
||
stream_mode="values"
|
||
)
|
||
for chunk in res:
|
||
print(chunk["messages"][-1].content)
|