0159-python-type-safety

Catch Python type errors before runtime by enforcing static type correctness through annotations using mypy --strict or Pyright, applying modern union syntax, type narrowing, and bounded TypeVars to surface issues early in CI rather than at runtime.

Type narrowing safely resolves T | None types using guards, while Python protocols define structural interfaces so you can preserve type information across reusable components without requiring explicit inheritance.

Configure mypy strict mode or Pyright for CI type checking by enabling strict rules and adding type annotations to public function signatures, methods, and class interfaces to validate correct usage and surface issues early.

Yes, mypy strict mode supports generics and bounded TypeVars, allowing you to preserve type information across reusable components by applying Generic and defining structural interfaces with Protocol.

You need Python type checking for a service layer when building internal libraries to define repository interfaces and response types, ensuring consumers cannot accidentally call methods with wrong entity types or forget to handle missing results.

The best way to handle nullable return types like User | None in Python is to use modern union syntax and explicit type narrowing with guards, ensuring missing results are handled explicitly during static analysis.