What problem does it solve? Python codebases accumulate subtle defects — mutable default arguments, naive datetimes, ambiguous mutation contracts, and type-safety drift — that slip past casual review. This Skill gives an AI agent a structured, impact-prioritized rulebook derived from real PR review patterns so code generation and review follow consistent, defensible engineering standards. ## Core Features & Use Cases - 75 rules across 8 categories: Data Modeling, Error Handling, Type Safety, API Design, Code Simplification, Performance, Naming, and Imports & Structure, each tagged with an impact level from CRITICAL to LOW. - Incorrect/Correct code pairs with counter-signals: every rule shows a bad pattern, the fix, and an explicit "when NOT to apply" paragraph so rules are applied with judgment rather than as blanket refactors. - Version-aware guidance: rules assume Python 3.11+ and call out version-specific features (e.g., warnings.deprecated on 3.13+, assert_never on 3.11+) plus Pydantic-specific applicability tags. - Use Case: While reviewing a new FastAPI service module, the agent flags a shared mutable default argument (CRITICAL), suggests a discriminated union instead of an optional-field bag, and cites the exact rule file for each finding. ## Quick Start Ask the agent to review your Python module or write a new data model using the python-best-practices guidelines and report findings by impact level.
