socratink-app

socratink-app is an MVP-stage learning product deployed on Vercel serverless.

The repo contains two closely related layers:

• the hosted app: FastAPI backend plus a vanilla JS frontend in public/
• the socratink prompt and skill assets in learnops/ that shape extraction and drill behavior

The product doctrine is stable even while implementation is still moving:

• generation before recognition
• the graph tells the truth
• AI should remove prep friction and increase truthful retrieval reps
• local success does not validate hosted behavior

Start Here

• Read AGENTS.md for repo constraints, agent routing, and workflow ownership.
• Read docs/project/state.md for current MVP priorities and target-user signals.
• Read docs/theta/state.md when the task touches learning claims or product science.
• Read docs/drill/graph-invariants.md before changing drill or graph behavior.

Repo Shape

• main.py and api/ FastAPI app and Vercel entrypoint.
• ai_service.py Model-facing extraction and drill logic.
• public/ Hosted frontend.
• learnops/ Prompt assets and skill artifacts used by the product.
• docs/ Product, drill, project, and agent workflow docs.

Local Run

uvicorn main:app --reload

Then open http://localhost:8000.