FastAPI AI Scaffold

Generates a runnable, production-shaped FastAPI AI/ML service so you don't re-derive the same wiring every time. The skeleton encodes the patterns from the fastapi-ai-patterns skill as working code: nothing here is novel framework usage, it's the boring-but-correct baseline an inference service needs before it sees traffic.

Inspired by FastAPI for AI Engineers (AI Engineering Insider, 2026). All generated code is original.

When to use

• "Start/bootstrap a new FastAPI service for serving a model / LLM / RAG / embeddings"
• "Set up a FastAPI project with clean architecture (router/service/repository)"
• The user wants the production basics pre-wired (lifespan model loading, auth, DB, probes, tests, Docker) rather than a single-file main.py

When NOT to use