FastAPI best practices covering project structure, Pydantic v2 schemas, dependency injection, async handlers, authentication, authorization, transactional service layers, and testing with httpx and pytest. Use when building or reviewing FastAPI apps — Pydantic schemas, dependencies, async handlers,

FastAPI Patterns Modern, production grade FastAPI development: project layout, Pydantic v2 schemas, dependency injection, async patterns, auth, transactional service methods, and testing. Project Structure App Factory and Lifespan Configuration with pydantic settings Pydantic Schemas (v2) Dependency Injection Router and Endpoint Design Service Layer Note on Database Design: Application level unique handling requires an underlying unique database index (e.g., unique=True on your SQLAlchemy mapping attributes). Without underlying constraints, application layer error catching cannot safely prevent concurrent race conditions. Testing with httpx and pytest Anti Patterns Best Practices Always declare a typed response model to prevent accidental PII/data leaks and output clean OpenAPI schemas. Consolidate standard middleware dependency injections via type aliasing: DbDep = Annotated[AsyncSession, Depends(get db)] . Wrap database mutation boundaries gracefully within transactions inside your service layer, catching structural database errors directly. Parse JWT parameters defensively, expecting potential string/integer cast mismatches from modern payload variations. Enforce deterministic sorting (e.g., .order by(Model.id) ) on all offset/limit paginated endpoints to avoid data skips. Isolate authorization checks from core authentication dependencies to provide precise REST status signals ( 401 vs 403 ).

fastapi-patterns