Decorators are a powerful way to extend functions, but their flexibility can lead to ambiguity and runtime errors. By using type annotations, we can clarify their behavior and leverage mypy and pyright to catch mistakes early.

This talk explores typing.ParamSpec, TypeVarTuple, Protocol, Concatenate, and other tools to create type-safe decorators that preserve function signatures. You'll learn best practices to improve readability, reduce debugging time, and write more maintainable Python code.