We’ve found it incredibly useful for figuring out what is going on with a codebase.
I googled a bit (actually DuckDuckGo’d) to find the source code of that tool because I wondered what it considered “evidence”, and I believe this is it. Looks like it looks for:
- log-and-continue catch blocks
- error-obscuring catch blocks (default-return or generic replacement error)
- empty catch blocks
- promise .catch() default fallbacks
- generic status envelopes
- generic record casts
- stringified unknown errors
- pass-through wrappers
- duplicated test mock/setup patterns
We guess you missed, or it wasn’t there at the time, the button that says source code on the site: https://codeberg.org/polyphony/repo-slopscore
Yeah, it wasn’t there at the time. Looks like I found the wrong repo.
Ah, fair. We vaguely remember it being absent and seems like the link you provided confirms it, thank you.
Seems more like a bad coding practice detector than an AI detector. Although obviously one would expect AI code to be more sloppy. I often do log-and-continue catch blocks myself with the intent of coming back to do better error handling later, but often times I don’t because I forgot or am lazy or don’t care.
lol
would be funny if it was the correct repo. Thankfully it isn’t: https://codeberg.org/polyphony/repo-slopscore
