Conm Llama

A Python-based Command Line numerical methods solver that uses the Model Context Protocol (MCP) to find roots of quadratic equations using various numerical methods.

Features

• Bisection Method: Finds roots within a given interval
• Newton-Raphson Method: Fast convergence with single initial guess
• Secant Method: Similar to Newton-Raphson but doesn't require derivative
• Quadratic Formula: Direct calculation of all roots
• Interactive Mode: Custom input support for solving any quadratic equation
• AI-Powered: Uses Groq LLM to intelligently choose and apply methods

Installation

1. Clone project

   git clone https://github.com/Bojji3487/mcp-conm-llama
   

2. Install dependencies:

   uv add -r requirements.txt
   

3. Set up your Groq API key:

   • Get an API key from Groq Console
   • Create a .env file and add your key:

   GROQ_API_KEY=your_key_here
   

Usage

Interactive Mode (Default)

python client.py

This launches an interactive menu where you can:

• Solve custom equations with your choice of method
• Run predefined test examples
• Ask free-form questions about numerical methods

Test Mode

python client.py --test

Runs a suite of predefined test cases to demonstrate all methods.

Example Custom Input Session

🔢 NUMERICAL METHODS SOLVER - Interactive Mode
============================================================

Options:
  1. Solve a custom equation
  2. Run test examples
  3. Free-form question
  4. Exit

Select option (1-4): 1

📐 Enter quadratic equation coefficients (ax² + bx + c = 0):
  a = 1
  b = -5
  c = 6

🔧 Select a method:
  1. Bisection Method
  2. Newton-Raphson Method
  3. Secant Method
  4. Find all roots (Quadratic Formula)
  5. Let AI choose the best method

Enter choice (1-5): 4

🔄 Processing...

📊 Result:
The quadratic formula gives us two real roots: x₁ = 3.000000 and x₂ = 2.000000

Special features

1. Annotated Return Types: Added proper type hints for function returns
2. Descriptive Error Handling: Descriptive error messages with actual values
3. Interactive Client: Menu Driven interactive client that answers using AI
4. Custom Input Support: Users can input any quadratic equation
5. Additional Tool: Added find_all_roots using the quadratic formula
6. Iteration Tracking: Shows number of iterations for convergence
7. uv managed project: This is a uv project that uses uv, a fast alternative to pip...As well as uv starter project using

   uv init
   

Mathematical Background

The solver handles equations of the form: ax² + bx + c = 0

• Bisection Method: Good ol' reliable but slower, requires interval with sign change, best for checking answer when your other methods are uncertain
• Newton-Raphson: Fast convergence, may fail if derivative is zero, easy AF tho....
• Secant Method: Good alternative when derivative is hard to compute, never used it much tbh.
• Quadratic Formula: Direct solution, handles complex roots

Troubleshooting

1. "GROQ_API_KEY not found": Make sure your .env file exists and contains your API key
2. Import errors: Install all requirements using uv add -r requirements.txt

License

MIT License - Feel free to modify and use as needed!