模型上下文协议（MCP）开发实战——构建LangChain代理客户端 原创

curl -LsSf https://astral.sh/uv/install.sh | sh
1.

# 为我们的项目创建一个新目录
uv init weather 
cd weather 
# 创建虚拟环境并激活它
uv venv 
source .venv/bin/activate 
# 安装依赖项
uv add "mcp[cli]" httpx 
# 创建我们的服务器文件
touch weather.py
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.

code .
1.

from typing import Any
import httpx
from mcp.server.fastmcp import FastMCP
#初始化FastMCP服务器
mcp = FastMCP("weather")
# 指定常量
NWS_API_BASE = "https://api.weather.gov"
USER_AGENT = "weather-app/1.0"
1.
2.
3.
4.
5.
6.
7.
8.

async def make_nws_request(url: str) -> dict[str, Any] | None:
 """使用恰当的错误处理方式向NWS API提出请求。"""
 headers = {
 "User-Agent": USER_AGENT,
 "Accept": "application/geo+json"
 }
 async with httpx.AsyncClient() as client:
 try:
 response = await client.get(url, headers=headers, timeout=30.0)
 response.raise_for_status()
 return response.json()
 except Exception:
 return None
def format_alert(feature: dict) -> str:
 """将报警功能格式化为可读的字符串。"""
 props = feature["properties"]
 return f"""
Event: {props.get('event', 'Unknown')}
Area: {props.get('areaDesc', 'Unknown')}
Severity: {props.get('severity', 'Unknown')}
Description: {props.get('description', 'No description available')}
Instructions: {props.get('instruction', 'No specific instructions provided')}
"""
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.

@mcp.tool()
async def get_alerts(state: str) -> str:
 """获取美国一个州的天气警报。
 参数:
 state: 两个字母的美国州代码（例如CA，NY)
 """
 url = f"{NWS_API_BASE}/alerts/active/area/{state}"
 data = await make_nws_request(url)
 if not data or "features" not in data:
 return "Unable to fetch alerts or no alerts found."
 if not data["features"]:
 return "No active alerts for this state."
 alerts = [format_alert(feature) for feature in data["features"]]
 return "\n---\n".join(alerts)
@mcp.tool()
async def get_forecast(latitude: float, longitude: float) -> str:
 """获取一个地点的天气预报。
 参数:
 latitude: 位置的纬度
 longitude: 位置的经度
 """
 # 首先获取预测网格端点
 points_url = f"{NWS_API_BASE}/points/{latitude},{longitude}"
 points_data = await make_nws_request(points_url)
 if not points_data:
 return "Unable to fetch forecast data for this location."
 # 从端点响应中获取预测的URL
 forecast_url = points_data["properties"]["forecast"]
 forecast_data = await make_nws_request(forecast_url)
 if not forecast_data:
 return "Unable to fetch detailed forecast."
 # 将时间范围格式化为可读的预测
 periods = forecast_data["properties"]["periods"]
 forecasts = []
 for period in periods[:5]: # 只显示未来5个时段
 forecast = f"""
{period['name']}:
Temperature: {period['temperature']}°{period['temperatureUnit']}
Wind: {period['windSpeed']} {period['windDirection']}
Forecast: {period['detailedForecast']}
"""
 forecasts.append(forecast)
 return "\n---\n".join(forecasts)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.

if __name__ == "__main__" : 
 # 初始化并运行服务器
 mcp.run(transport= 'stdio' )
1.
2.
3.

uv run weather.py
1.

uv add ipykernel
1.

uv run weather.py
1.

!uv add langchain-mcp-adapters
1.

!uv add langgraph langchain-anthropic python-dotenv
1.

ANTHROPIC_API_KEY =sk-xxxxxxxx
1.

from dotenv import load_dotenv
load_dotenv()
import os
api_key=os.environ.get("ANTHROPIC_API_KEY")
1.
2.
3.
4.

from mcp import ClientSession, StdioServerParameters
from mcp.client.stdio import stdio_client

from langchain_mcp_adapters.tools import load_mcp_tools
from langgraph.prebuilt import create_react_agent

from langchain_anthropic import ChatAnthropic
1.
2.
3.
4.
5.
6.
7.

model = ChatAnthropic(model="claude-3-5-sonnet-20241022", api_key=api_key)
1.

server_params = StdioServerParameters( 
 command= "python" , 
 # 确保更新为math_server.py文件的完整绝对路径
 args=[ "./weather.py" ], 
) 

async with stdio_client(server_params) as (read, write): 
 async with ClientSession(read, write) as session: 
 # 初始化连接
 await session.initialize() 

 # 获取工具
 tools = await load_mcp_tools(session) 

 # 创建并运行代理
 agent = create_react_agent(model, tools) 
 agent_response = await agent.ainvoke({ "messages" : "加州目前的天气怎么样" })
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.

agent_response
1.

from IPython.display import display, Markdown
from langchain_core.messages import HumanMessage, ToolMessage, AIMessage
for response in agent_response["messages"]:
 user = ""

 if isinstance(response, HumanMessage):
 user = "**User**"
 elif isinstance(response, ToolMessage):
 user = "**Tool**"
 elif isinstance(response, AIMessage):
 user = "**AI**"

 if isinstance(response.content, list):
 display(Markdown(f'{user}: {response.content[0].get("text", "")}'))
 continue
 display(Markdown(f"{user}: {response.content}"))
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.

if __name__ == "__main__" : 
 # 初始化并运行服务器
 mcp.run(transport= 'sse' )
1.
2.
3.

uv run weather.py
1.

from langchain_mcp_adapters.client import MultiServerMCPClient
1.

async with MultiServerMCPClient(
 {
 "weather": {
 "url": "http://localhost:8000/sse",
 "transport": "sse",
 }
 }
) as client:
 agent = create_react_agent(model, client.get_tools())
 agent_response = await agent.ainvoke({"messages": "what is the weather in nyc?"})

for response in agent_response["messages"]:
 user = ""

 if isinstance(response, HumanMessage):
 user = "**User**"
 elif isinstance(response, ToolMessage):
 user = "**Tool**"
 elif isinstance(response, AIMessage):
 user = "**AI**"

 if isinstance(response.content, list):
 display(Markdown(f'{user}: {response.content[0].get("text", "")}'))
 continue
 display(Markdown(f"{user}: {response.content}"))
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.

async with MultiServerMCPClient( 
 { 
 "weather" : { 
 "url" : "http://localhost:8000/sse" , 
 "transport" : "sse" , 
 } 
 } 
) as client: 
 agent = create_react_agent(model, client.get_tools()) 

 # 流式传输响应块
 async for chunk in agent.astream({ "messages" : "what is the weather in nyc!" }): 
 # 从AddableUpdatesDict结构中提取消息内容
 if 'agent' in chunk and 'messages' in chunk[ 'agent' ]: 
 for message in chunk[ 'agent' ][ 'messages' ]: 
 if isinstance (message, AIMessage): 
 # 处理不同的内容格式
 if isinstance (message.content, list ): 
 # 对于带有文本和工具使用的结构化内容
 for item in message.content: 
 if isinstance (item, dict ) and 'text' in item: 
 display(Markdown( f"**AI**: {item[ 'text' ]} " )) 
 else : 
 # 对于简单文本内容
 display(Markdown( f"**AI**: {message.content} " )) 

 elif 'tools' in chunk and 'messages' in chunk[ 'tools' ]: 
 for message in chunk[ 'tools' ][ 'messages' ]: 
 if hasattr (message, 'name' ) and hasattr (message, 'content' ): 
 # 显示工具响应
 display(Markdown( f"**Tool ( {message.name} )**: {message.content} " ))
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.