mcp_server_101

mcp_server_101

3.1

If you are the rightful owner of mcp_server_101 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.

This repository is a practice session for the MIT Hack Decentralized AI summit, focusing on setting up a Claude MCP protocol server quickly.

Claude MCP Protocol Practice

This repository is a practice session for MIT Hack Decentralized AI summit. It includes basic steps to create a Claude MCP protocol up and running in under 5 minutes on Windows.

Weather App Setup

Prerequisites

  • Make sure you pip install uv

Steps

cd weather
uv run weather.py

This starts the server hosted locally on your machine. The python file includes tools available to LLM for getting alerts and forecast from government API.

Claude Configuration

Make sure you changed the claude_config.json available on Claude desktop:

  1. Navigate to File -> Settings -> Developer -> Edit Config
  2. Edit config section and provide absolute path to your python file

Voila! Your 1st MCP server is up and running in 5 mins.

Sample Config File

{
  "mcpServers": {
    "weather": {
      "command": "npx",
      "args": ["-y", "<abs_path_to_file>/mcp-server-weather"]
    }
  }
}

SQL-lite DB Example for Bakery Data

Steps to Run the SQL DB as an MCP Server

  1. Clone the repository:

    git clone https://github.com/modelcontextprotocol/servers.git
    
  2. Navigate to the SQLite directory:

    cd servers
    cd src
    cd sqlite
    
  3. Install Docker if needed:

  4. Build the Docker image:

    docker build -t mcp/sqlite .
    
  5. Update your Claude config file:

    {
      "mcpServers": {
        "sqlite": {
          "command": "docker",
          "args": [
            "run",
            "--rm",
            "-i",
            "-v",
            "mcp-test:/mcp",
            "mcp/sqlite",
            "--db-path",
            "/mcp/test.db"
          ]
        }
      }
    }
    
  6. Restart Claude to apply changes

Bakery Database Setup

  1. Create bakery DB tables:

    • ingredients (id, name, unit, quantity, cost, supplier, min_order_quantity)
  2. Populate with:

    • 10 ingredients with MOQs
  3. Add queries for:

    • Low stock alerts

Hope this is helpful to you in the future!

References

MIT Hack Documentation