cleanup-code

name	cleanup-code
description	Polishes recently written or modified Python code for style, dead code, docstrings, type annotations, and checks. Use when coding task is completed.
argument-hint	["directory or file"]
disable-model-invocation	true
allowed-tools	["Read","Glob","Grep","Edit","Bash"]

Clean Up Code

Do a final cleanup polish pass on the code in $ARGUMENTS (or the current package if no argument given).

Use the write-code skill to understand code conventions.

1. Re-read every file that was created or modified.
2. Check for Pythonic style:
   • Use comprehensions instead of loops where clearer
   • Use f-strings instead of .format() or %
   • Use pathlib instead of os.path
   • Use dataclasses or named tuples for structured data
3. Remove dead code, unused imports, and commented-out code
4. Remove unecessary tests, such as one that simply check types or were used for development
5. Verify docstrings on all public APIs (follow write-docstrings skill conventions).
6. Verify complete type annotations on all public functions and methods.
7. Run the full check suite:
   • ruff format .
   • ruff check .
   • ty check
   • pytest
8. All checks must pass before finishing.