Add information about which LLM/tool you used (and enforce tests/branch protection)

[rugk]

IMHO, as for AI/LLM coded projects, this should be stated/disclosed in the Readme and best also explained what tool you have used.

Given all the subtle bugs I found here and the general feeling etc., I am quite sure, this project is at least partially vibecoded, which itself I also do not want to discourage (especially for scraping etc. it is good and okay), but I'd say disclosing that is the ethical minimum.
It also allows for further developing/contributing on the same toolset and more transparency regarding what is done etc.

What can be improved

Also, generally, I guess these many bugs show a potential problem in good/improved quality and/or testing. One thing I also notice is, that PRs are routinely merged, even though the tests do not pass: https://github.com/Muetze42/hellofresh-database/pull/281 (Tests are there for a reason and proper tests and thus I'd say that is not good - and can even be enforced in GitHub's repo settings.)

What seems to be used

Claude: https://github.com/Muetze42/hellofresh-database/blob/main/CLAUDE.md
https://laravel.com/ai/boost

[Muetze42]

Thanks for the feedback! We already have a CLAUDE.md documenting Claude's involvement, and Claude is listed as a contributor.
All tests are automated and must pass before merging to main. PRs are welcome if you'd like to help improve the code quality or add additional test coverage.

[rugk]

All tests are automated and must pass before merging to main.

Yeah, but that apparently that was not done quite consequently in the past:

As said, to prevent that, you can just add a branch rule to GitHub, sou you cannot accidentally merge a PR without tests passing:
https://docs.github.com/en/repositories/configuring-branches-and-merges-in-your-repository/managing-protected-branches/managing-a-branch-protection-rule#creating-a-branch-protection-rule

[Muetze42]

Deployment is only possible if all tests have been passed.

[rugk]

Ah okay, there is a check. Anyway, IMHO, best practise is to only have code in main that passes tests. If someone else develops/contributes upon that, otherwise, they may not get the tests fixed.