How do I use WeatherServer?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@WeatherServer What's the weather in London?" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

LangChain MCP Weather Application

A complete example demonstrating how to build and consume a Model Context Protocol (MCP) server using:

Python
FastMCP
LangChain
OpenAI
Open-Meteo APIs

The project exposes weather information through an MCP server and allows a LangChain agent to discover and invoke weather tools dynamically.

Features

The weather tool returns:

Latitude
Longitude
Timezone
Humidity
Wind Speed
Weather Conditions
Sunrise Time
Sunset Time
7-Day Forecast

The application demonstrates:

Creating an MCP server with FastMCP
Registering MCP tools
Connecting to MCP servers with LangChain
Discovering tools dynamically
Building AI agents that invoke MCP tools
Integrating external REST APIs

Architecture

+-------------------+
|   LangChain Agent |
+---------+---------+
          |
          v
+-------------------+
| MCP Client        |
| MultiServerMCP    |
+---------+---------+
          |
          | stdio
          |
          v
+-------------------+
| MCP Server        |
| FastMCP           |
+---------+---------+
          |
          v
+-------------------+
| Weather Service   |
+---------+---------+
          |
          v
+-------------------+
| Open-Meteo APIs   |
+-------------------+

Project Structure

project/
│
├── server.py
├── client.py
├── agent.py
├── weather_service.py
├── requirements.txt
├── .env
└── README.md

Components

1. weather_service.py

Contains the business logic responsible for:

Geocoding

Converts a city name into:

Latitude
Longitude
Timezone

Uses:

https://geocoding-api.open-meteo.com

Weather Retrieval

Fetches:

Current humidity
Wind speed
Weather conditions
Sunrise
Sunset
7-day forecast

Uses:

https://api.open-meteo.com

Example Output

{
  "city": "Matadi",
  "latitude": -5.799,
  "longitude": 13.440,
  "timezone": "Africa/Kinshasa",
  "humidity": 82,
  "wind_speed": 12.4,
  "weather_conditions": 1,
  "sunrise": "2026-06-01T06:03",
  "sunset": "2026-06-01T17:58",
  "forecast": [
    {
      "date": "2026-06-01",
      "min_temp": 20.2,
      "max_temp": 29.1
    }
  ]
}

2. server.py

Creates the MCP server.

MCP Server Initialization

mcp = FastMCP("WeatherServer")

Tool Registration

@mcp.tool()
def weather(city: str):
    return get_weather(city)

Start Server

mcp.run()

The server exposes the weather tool to any MCP-compatible client.

3. client.py

Demonstrates connecting to the MCP server and discovering tools.

Create MCP Client

client = MultiServerMCPClient(
    {
        "weather_server": {
            "transport": "stdio",
            "command": "python",
            "args": ["server.py"]
        }
    }
)

Discover Tools

tools = await client.get_tools()

Example Output

TOOLS
[StructuredTool(name='weather', ...)]

4. agent.py

Builds an AI agent capable of invoking MCP tools.

Load Environment Variables

load_dotenv()

Create OpenAI Model

llm = ChatOpenAI(
    model="gpt-5-nano"
)

Retrieve MCP Tools

tools = await client.get_tools()

Create Agent

agent = create_agent(
    model=llm,
    tools=tools,
    system_prompt="You are a weather assistant."
)

Invoke Agent

result = await agent.ainvoke(
    {
        "messages": [
            {
                "role": "user",
                "content": "What is the weather in Matadi?"
            }
        ]
    }
)

Installation

Clone Repository

git clone https://github.com/your-org/weather-mcp.git

cd weather-mcp

Create Virtual Environment

Using venv

python -m venv .venv

Activate:

Linux / macOS

source .venv/bin/activate

Windows

.venv\Scripts\activate

Install Dependencies

pip install -r requirements.txt

requirements.txt

mcp
langchain
langchain-openai
langchain-mcp-adapters
python-dotenv
requests
openai

Environment Variables

Create a .env file:

OPENAI_API_KEY=your_openai_api_key

Running the Application

Option 1: Run MCP Server

python server.py

The server starts and waits for MCP client requests.

Option 2: Test MCP Client

python client.py

Example:

TOOLS
[StructuredTool(name='weather', ...)]

Option 3: Run LangChain Agent

python agent.py

Example:

{
  "city": "Matadi",
  "latitude": -5.799,
  "longitude": 13.440,
  "timezone": "Africa/Kinshasa",
  "humidity": 82,
  "wind_speed": 12.4,
  "weather_conditions": 1,
  "sunrise": "2026-06-01T06:03",
  "sunset": "2026-06-01T17:58",
  "forecast": [
    {
      "date": "2026-06-01",
      "min_temp": 20.2,
      "max_temp": 29.1
    }
  ]
}

MCP Workflow

User Question
      |
      v
LangChain Agent
      |
      v
MCP Tool Discovery
      |
      v
Weather Tool
      |
      v
Open-Meteo APIs
      |
      v
Weather Data
      |
      v
Agent Response

Example Queries

What is the weather in Paris?

What is the humidity in New York?

Give me the 7-day forecast for Tokyo.

When is sunrise in London?

What is the wind speed in Matadi?

Future Enhancements

Multiple MCP servers
Weather alerts
Historical weather data
Air quality information
LangGraph integration
Redis caching
Azure deployment
Streaming responses
RAG integration
Multi-agent orchestration

Technologies Used

Python
FastMCP
LangChain
OpenAI
Open-Meteo API
AsyncIO
MCP (Model Context Protocol)

Learning Objectives

This project teaches:

MCP Server Development
MCP Tool Registration
MCP Client Integration
LangChain Tool Discovery
AI Agent Tool Calling
REST API Integration
Async Python Programming
OpenAI + LangChain Integration

License

MIT License

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files to deal in the Software without restriction.

LangChain MCP Weather Application

Features

Architecture

Project Structure

Components

1. weather_service.py

Geocoding

Weather Retrieval

Example Output

2. server.py

MCP Server Initialization

Tool Registration

Start Server

3. client.py

Create MCP Client

Discover Tools

Example Output

4. agent.py

Load Environment Variables

Create OpenAI Model

Retrieve MCP Tools

Create Agent

Invoke Agent

Installation

Clone Repository

Create Virtual Environment

Using venv

Linux / macOS

Windows

Install Dependencies

requirements.txt

Environment Variables

Running the Application

Option 1: Run MCP Server

Option 2: Test MCP Client

Option 3: Run LangChain Agent

MCP Workflow

Example Queries

Future Enhancements

Technologies Used

Learning Objectives

License

Resources

Looking for Admin?

Latest Blog Posts

MCP directory API