mcp_server

kazhian/mcp_server

3.2

If you are the rightful owner of mcp_server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

The Model Context Protocol (MCP) server is a framework designed to facilitate communication between language models and external tools or APIs, enabling seamless integration and execution of tasks.

Tools
1
Resources
0
Prompts
0

Install libs & run

brew install uv

uv add mcp[cli]
uv run weather.py -- Server started. But you wont see anythiong
-- Now run
mcp dev weather.py -- Run a MCP server with the MCP Inspector. 
mcp run weather.py -- Run a MCP server.

Hard Way: Add this in Claude settings

  • Go to Claude settings and add this
  • Open Claude
  • Open Settings
  • Select Developer
  • Click Edit Config button
  • This will launch explorer/finder
  • Open the config file and add this
{
  "mcpServers": {
    "weather": {
      "command": "uv",
      "args": [
        "--directory",
        "/Users/kazhian/Lab/mcp_server",
        "run",
        "weather.py"
      ]
    }
  }
}

Automatic: Add the mcp server to Claude settings

$ mcp install weather.py
 INFO     Added server 'WeatherAPI' to Claude config     claude.py:136
 INFO     Successfully installed WeatherAPI in Claude app   cli.py:485

Now Claude settings JSON will look like this

{
  "mcpServers": {
    "weather": {
      "command": "uv",
      "args": [
        "--directory",
        "/Users/kazhian/Lab/mcp_server",
        "run",
        "weather.py"
      ]
    },
    "WeatherAPI": {
      "command": "/usr/local/bin/uv",
      "args": [
        "run",
        "--with",
        "mcp[cli]",
        "mcp",
        "run",
        "/Users/kazhian/Lab/mcp_server/weather.py"
      ]
    }
  }
}

sequenceDiagram
    participant User
    participant ClaudeDesktop as Claude Desktop<br/>(Orchestrator)
    participant LLM as Claude LLM
    participant MCPClient as MCP Client
    participant MCPServer as MCP Server (Weather)
    participant WeatherAPI as Weather API/Tool

    User->>ClaudeDesktop: "Get me the weather for Chennai"
    ClaudeDesktop->>LLM: Send user message
    LLM->>LLM: Analyze request & decide tool needed
    LLM-->>ClaudeDesktop: Return tool call request:<br/>weather:weather(location="Chennai")
    Note over ClaudeDesktop: Claude Desktop executes<br/>the tool call (not LLM)
    ClaudeDesktop->>MCPClient: Execute tool via MCP Client
    MCPClient->>MCPServer: Forward tool call via MCP protocol
    MCPServer->>WeatherAPI: Fetch weather data for Chennai
    WeatherAPI->>WeatherAPI: Retrieve current conditions<br/>(temp, humidity, condition)
    WeatherAPI-->>MCPServer: Return weather data:<br/>{condition: "foggy", temp: 7°C,<br/>humidity: 27%, description: "..."}
    MCPServer-->>MCPClient: Send response via MCP protocol
    MCPClient-->>ClaudeDesktop: Return tool result
    ClaudeDesktop->>LLM: Send tool results to LLM
    LLM->>LLM: Process results & format response<br/>into readable paragraph
    LLM-->>ClaudeDesktop: Return formatted response
    ClaudeDesktop-->>User: Display: '"'The current weather in Chennai<br/>is foggy and windy with a temperature<br/>of 7°C and humidity at 27%...'"'