🚀 FastMCP Math Server

A modular and extensible Model Context Protocol (MCP) server built using FastMCP in Python. This project goes beyond basic arithmetic and provides a structured mathematical toolkit including algebra, statistics, number theory, and linear algebra.

📌 Project Overview

Project Name: fastmcp_server

This project implements an MCP server exposing multiple math-focused tool groups:

🔢 Core Capabilities

• Calculator

  • add, subtract, multiply, divide

• Equations

  • linear solver
  • quadratic solver
  • system of equations (2 variables)
  • Gaussian elimination (NxN)

• Number Theory

  • gcd, lcm
  • prime check
  • prime factorization

• Statistics

  • mean, median
  • variance
  • standard deviation

🧠 Architecture

+---------------------------+
| VS Code Copilot (Client) |
+------------+-------------+
             |
             |  MCP (HTTP + SSE)
             |
+------------v----------------------+
| FastMCP Server (Python)          |
|----------------------------------|
| calculator tools                 |
| equations solvers                |
| number theory utilities          |
| statistics functions             |
+----------------------------------+

🛠️ Tech Stack

• Python
• FastMCP
• Model Context Protocol (MCP)
• VS Code Copilot (Client)

📂 Project Structure

fastmcp_server/
│
├── .venv/                         # Virtual environment
├── .vscode/
│   └── mcp.json                  # MCP configuration for VS Code
│
├── server_calculator.py           # Basic arithmetic tools
├── server_equations.py           # Linear, quadratic, system solvers
├── server_numberTheory.py        # GCD, LCM, primes, factorization
├── server_statistics.py          # Mean, median, variance, std dev
│
├── requirements.txt              # Dependencies
├── README.md                     # Documentation

⚙️ Setup Instructions

1️⃣ Create Virtual Environment (Recommended)

python -m venv .venv
source .venv/bin/activate     # Linux/Mac
.venv\Scripts\activate        # Windows

2️⃣ Install Dependencies

pip install -r requirements.txt

Or manually:

pip install fastmcp

3️⃣ Configure MCP in VS Code

Create:

.vscode/mcp.json

{
  "servers": {
    "mcpCalculator": {
      "type": "http",
      "url": "http://localhost:3000/mcp"
    },
    "mcpNumberTheory": {
      "type": "http",
      "url": "http://localhost:3001/mcp"
    },
    "mcpEquations": {
      "type": "http",
      "url": "http://localhost:3002/mcp"
    },
    "mcpStatistics": {
      "type": "http",
      "url": "http://localhost:9000/mcp"
    }
  }
}

▶️ Running the Servers

Run each server (First try to run 1/2 servers)

python server_calculator.py

python server_equations.py

python server_statistics.py

python server_numberTheory.py

Servers runs at:

http://localhost:3000/mcp
http://localhost:3001/mcp
http://localhost:3002/mcp
http://localhost:9000/mcp

🔌 Connecting to VS Code Copilot

1. Open VS Code
2. Ensure Copilot is enabled
3. Open Command Palette:

> MCP: List Servers

1. Select:

mcpCalculator
mcpNumberTheory
mcpEquations
mcpStatistics

🔌 Connecting to VS Code Copilot (Easy Way For connection)

• Go in mcp.json
• Above each server there are multiple buttons (Start, Stop, Restart)
• You can connect servers from here
• Start to chat in Copilot window (ctrl + Alt + B)

💬 Example Prompts

🔢 Calculator

Add 10 and 20
Divide 100 by 5

📐 Equations

Solve 2x + 3y = 13 and x - y = 1
Solve quadratic equation 2x^2 + 5x + 3 = 0

🔬 Number Theory

Find gcd of 48 and 18
Check if 29 is prime
Prime factors of 84

📊 Statistics

Mean of [1,2,3,4,5]
Variance of [10,20,30]
Standard deviation of dataset

⚠️ Important Notes

• MCP is not REST
• It Uses:
• JSON-RPC
• Server-Sent Events (SSE)

Expected behavior:

• Accessing / in browser → 404 Not Found (expected)
• Accessing /mcp in browser → 406 Not Acceptable (expected)

✔ This is normal

Use only MCP-compatible clients.

🚀 Future Enhancements

• NxN matrix operations (determinant, inverse)
• LU / QR decomposition
• Symbolic math support
• Graph plotting tools
• Financial/statistical models
• Streaming statistics (real-time data)

🙌 Acknowledgements

• FastMCP
• Model Context Protocol (MCP)
• VS Code Copilot

🛠️ Tech Stack

• Python
• FastMCP
• MCP (Model Context Protocol)
• VS Code Copilot (Client)

📜 License

MIT License