castfit code review #
Some reasonable feedback.
I asked both Gemini 2.5 Pro Experimental and ChatGPT o3 to review the code in castfit 0.1.1. I attached the source files for __init__.py, the two test files, and README.md together with the following prompt:
Please review the attached code and produce a comprehensive list of improvements. For any error or edge case, please produce sample code that would trigger the error/edge case. For any improvement, indicate the strength of the improvement. Improvements may include, among other things, performance, documentation, marketing, or ease-of-use.
While both found many similar issues, Gemini's output was both easier to convert to issues (although I should have been more explicit in the prompt) and it also exposed me to the great cattrs library which I should explore more.
I'll implement as many of these changes as I can and then see how the reviews change.
Note: I'm saving these transcripts to a private gist which is not indexed by search engines to try to reduce AI slop.