MCP BMI

What is MCP BMI?

MCP BMI is a simple MCP server that exposes two tools: one to calculate a person’s BMI from their weight and height, and another to return the corresponding BMI category. It is built on Python with FastMCP, packaged as a Docker container, and published to the GitHub Container Registry. It is intended for AI assistants or applications that need on‑demand BMI computation.

How to use MCP BMI?

The server can be used in two ways. In Visual Studio Code, add a Docker run command to your settings.json. Locally, run uv sync && uv run bmi_server.py or build and run the Docker image with docker build -t mcp-bmi . followed by docker run --rm -it mcp-bmi. The server provides two tools: calculate_bmi (requires weight and height as float inputs) and bmi_category (requires bmi as a float input).

Key features of MCP BMI

• calculate_bmi tool: returns a numeric BMI index from weight and height.
• bmi_category tool: returns a string category based on the BMI value.
• Built with Python and FastMCP for easy MCP integration.
• Packaged as a Docker image and available on the GitHub Container Registry.
• Ready‑to‑use DevContainer for development in VS Code.
• Includes a test script (test_bmi.py) that can be run with uv run test_bmi.py.

Use cases of MCP BMI

• An AI assistant calculating a user’s BMI from provided weight and height.
• A health‑aware application classifying BMI into underweight, healthy, or obesity categories.
• Developers prototyping health‑related tools within the MCP ecosystem.

FAQ from MCP BMI

What does calculate_bmi return?

It returns a number (bmi_index) representing the Body Mass Index calculated from the input weight and height.

What does bmi_category return?

It returns a string (bmi_category) with one of seven predefined categories: Underweight, Healthy Weight, Overweight, Obesity, Class 1 Obesity, Class 2 Obesity, or Class 3 Severe obesity.

How can I run the server in Visual Studio Code?

Add the following to your settings.json:

{
    "servers": {
      "command": "docker",
      "args": [
         "run",
         "-i",
          "--rm",
         "ghcr.io/oscarrenalias/mcp-server-bmi:latest"
       ]
      }
    }
  }

Can I run the server locally without Docker?

Yes. Run uv sync then uv run bmi_server.py to start the server directly on your machine. Python and uv are required.

What are the runtime dependencies?

The server is built with Python and FastMCP. For local use, you need Python and uv. For Docker usage, you need Docker. The server does not require any external API keys or network access.