This server brings discrete math tools into Claude through FastAPI endpoints covering four domains: boolean logic operations with truth table generation and expression simplification, number theory functions for primes and modular arithmetic, cryptography primitives for classical and modern ciphers, and algorithm visualization for sorting and searching. The MCP layer wraps an OpenAI agents implementation that already powers a React learning platform, so you're getting battle-tested calculation logic rather than toy examples. Reach for this when you need Claude to verify proof steps, calculate GCD/LCM in explanations, demonstrate cipher operations, or trace algorithm execution without hallucinating the math. Built on Python 3.11 with the full stack already handling educational workloads at localhost:8000.
An interactive web-based educational platform for learning discrete structures through visualizations, AI-powered assistance, and hands-on tools.
pip install uv)cd backend
uv sync
copy .env.example .env
# Edit .env and add your OPENAI_API_KEY
uv run uvicorn src.main:app --reload
Backend runs on http://localhost:8000
cd frontend
npm install
copy .env.example .env
npm run dev
Frontend runs on http://localhost:5173
discrete-structures-ai-platform/
├── backend/ # FastAPI backend + MCP server
│ ├── src/
│ │ ├── api/ # API endpoints
│ │ ├── agents/ # OpenAI agents
│ │ ├── mcp/ # MCP tools
│ │ ├── models/ # Pydantic models
│ │ └── services/ # Business logic
│ └── tests/ # Backend tests
├── frontend/ # React frontend
│ ├── src/
│ │ ├── components/ # React components
│ │ ├── pages/ # Page components
│ │ ├── services/ # API clients
│ │ ├── hooks/ # Custom hooks
│ │ └── types/ # TypeScript types
│ └── public/ # Static assets
└── specs/ # Project specifications
This project follows a phased implementation approach:
See specs/001-ds-ai-platform/tasks.md for detailed task breakdown.
Educational use only.