PyBe is a scenario-driven Python learning prototype built from the supplied PRD and breakdown document. It has no login flow for now.
📚 View the project wiki: WIKI.md
- Scenario browser with difficulty, concept, and search filters
- Interactive learning session: learner reasoning, abstraction mapping, conversational prompts, Python construct generation, prompt scoring, and reflection capture
- Dashboard with progress, prompt maturity, concept mastery, misconceptions, and recent sessions
- Roadmap view covering V0 through V3 from the source documents
- JSON-file backed API with seed data
- JSON file storage
- Express + Node.js
- React + Vite
- Plain CSS, no auth
- Node.js 18+
- Install dependencies:
npm run installAll- Configure the server environment:
cp server/.env.example server/.envThe default values work for local development.
- Seed sample data:
npm run seed- Run the app:
npm run dev- Frontend: http://localhost:5173
- API: http://localhost:5000/api
The AI behavior in this prototype is deterministic and local. The abstraction mapper, prompt evaluator, and Python construct generator use rule-based logic so you can run everything without external AI keys. Later phases can replace those services with OpenAI, RAG, or TinyLLM components.
Learning data is stored in server/src/data/db.json. This keeps the prototype simple and fully local, without MongoDB, Docker, Atlas, or any external database.