AI-Powered Study Companion - An intelligent platform that combines scheduling, task management, and personalized AI tutoring.
study-agent/
├── frontend/ # React + Vite + TypeScript frontend
├── backend/ # FastAPI backend powered by agentling
└── docs/ # Project guidelines and documentation
The backend is a FastAPI application whose agent layer is powered by agentling — a tiny async tool-calling agent framework:
- agentling - ReAct agent loop, streaming events, skills, and tools
- Claude (Anthropic) - Model provider via the OpenAI-compatible endpoint
- FastAPI + WebSockets - Real-time streaming chat
- MongoDB (Beanie) - Persistence
- Pinecone + OpenAI embeddings - Document RAG
- Mem0 - Long-term memory store
Agent architecture highlights:
- 35 tools (tasks, goals, calendar, search, documents, UK visa sponsors, computation)
- 13 progressive-disclosure skills (SKILL.md folders loaded on demand via
load_skill) - A memory-manager sub-agent exposed as the
manage_memorytool, running a nested agentling agent on a lightweight model - Per-run user identity carried by a contextvar (
app/agent/context.py)
cd backend
python3 -m venv .venv
.venv/bin/pip install -r requirements.txt
cp .env.example .env # fill in your keys
.venv/bin/uvicorn app.main:app --reloadThe frontend is a React application built with:
- React 18 - UI framework
- TypeScript - Type safety
- Vite - Build tool and dev server
- React Router v7 - Client-side routing
- Tailwind CSS v4 - Styling
- Zustand - State management
- Motion - Animations
cd frontend
npm install
npm run devVisit http://localhost:5173 to view the app.
- 🎓 AI-Powered Learning - Personalized study companion
- 📅 Smart Scheduling - Intelligent calendar management
- 🎯 Goal Tracking - Track academic progress
- 📚 Reading Tools - AI annotations and quizzes
- ✨ 5 AI Agents - Different personality types to match learning styles
- 🧠 Long-term Memory - Remembers how you learn across sessions
- Frontend: frontend/README.md
Private project