Python Expert Skill

When working on Python code:

Style & conventions

• Use Python 3.10+ features: match/case, | union types, structural pattern matching
• Prefer dataclasses or pydantic over plain dicts for data models
• Use pathlib.Path over os.path
• Use f-strings over .format() or % formatting
• Type-annotate all function signatures

Testing

• Write pytest tests, not unittest
• Use pytest.fixture for setup
• Parametrize with @pytest.mark.parametrize
• Mock with pytest-mock's mocker fixture

Error handling

• Raise specific exceptions, not bare Exception
• Use contextlib.suppress for expected exceptions in cleanup paths
• Document exceptions in docstrings

Performance

• Prefer generators over list comprehensions when only iterating
• Use functools.lru_cache / functools.cache for pure functions
• Profile before optimizing — use cProfile or py-spy