Use this skill for a bounded conventional code review of the requested Python scope.
For a broad audit asking how the code could be more Pythonic, smaller, lower-maintenance, modernized, replaced by ecosystem libraries, or aligned with large-project practices, use python-engineering:python-quality-audit instead. That entryway fans out StinkySnake, SnakePolish, ecosystem, project-practice, and removal/debt research and synthesizes an actionable report.
Review the requested Python scope with these priorities.
Input
Scope: $ARGUMENTS
Instructions
- Read target files from arguments
- Check each dimension listed below
- Report findings with severity and location
- Suggest fixes with code examples
Shared policy
Load python-engineering:standards-for-python-development before judging the target. Its precedence and contextual defaults are authoritative; this entrypoint must not invent stricter framework migrations or numeric heuristics.
Review for correctness, security, boundary safety, design/cohesion, tests, typing, performance where demonstrated, and documentation affected by the change. Prefer the plugin's defaults for new work, but do not flag coherent existing unittest.mock, absence of Hypothesis without a useful invariant/input space, direct construction, or existing architecture merely because another preferred option exists.
For each finding report severity, file:line, evidence, consequence, suggested correction, and verification. Treat ~500 physical LOC as the default module boundary and inspect/decompose unless evidence shows splitting would worsen cohesion/API boundaries.
Do not edit the reviewed code.