Paste or upload a course syllabus. An AI agent parses it into topics, searches the web for the best free study resources per topic, and compiles a structured study guide — streamed to you in real time.
1. Clone and configure
git clone <repo-url>
cd ChillSyll
cp .env.example .env
# Edit .env with your API keys2. Get your API keys
OPENAI_API_KEY— from platform.openai.comTAVILY_API_KEY— from app.tavily.com (free tier available)
3. Run
chmod +x start.sh
./start.shOpen http://localhost:5173 in your browser.
| Layer | Technology |
|---|---|
| Frontend | React 18 + Vite + Tailwind CSS |
| Backend | Python + FastAPI |
| AI | OpenAI GPT-4o-mini |
| Web Search | Tavily Search API |
| PDF Parsing | PyPDF2 |
| Streaming | Server-Sent Events (SSE) |
-
Parse — You paste a syllabus (or upload a PDF). Claude extracts every topic, week label, and brief description into structured JSON. You can edit the topic list before proceeding.
-
Search — For each topic, the agent fires two Tavily search queries: one for explanations/tutorials, one for practice problems. Up to 4 topics are processed concurrently.
-
Evaluate — Claude reviews the search results and selects 2–4 of the best free resources per topic, prioritizing reputable sources (Khan Academy, MIT OCW, university pages) and ensuring variety (video + article + practice).
-
Stream — As each topic finishes, its resources appear in the study guide in real time via SSE. You don't wait for the entire guide to generate.
-
Export — Copy the study guide as Markdown or download it as a
.mdfile.
Three realistic syllabi are included in samples/ for testing:
cs101.txt— Intro Computer Science (15 topics)biology.txt— Principles of Biology (12 modules)history.txt— Modern World History (15 weeks)
Education