About this project
AI generates a lot of code, very quickly. That is mostly the point, but it creates a real problem for review: I was trained to read every line of what my team shipped, and with AI output that doesn't scale. Something has to take the hard edges off before the human review stage, and that is what Quality is for.
It scans a project for the things AIs are typically bad at. God classes. Clumsy credential handling and related security smells. Big unbroken blobs of code that not only read badly but make the AI itself less efficient next time, because it's chewing through tokens re-reading the same wall of text. The findings go back to the bot that wrote the code, and it makes a first pass at cleanup.
The result is that by the time I sit down to review, the obvious stuff has already been dealt with. I will also run it before open-sourcing any of these projects, so if you see one in a public repo, you can assume it has had a Quality pass. Not claiming it catches everything, but at least a real check has happened.
Screenshots
