This project was built with AI but with real human love (6 hour API time; 3 days wall time). I personally care for this project only because this project enabled me to have a language when reasoning about agent behavior. Slop is trivial in 2026's economy and but projects made with human love rarely converge to local minimas.
This project is very meta. It is about studying behaviors of dynamic programs whose graph is not even statically parsable. So any form of tutorials / studies added as a part of this project are AI "instructed" but with 0 value for slop since its a evaluation of the AI's behavior itself (whats the point if claude is juding claude.). The role of AI in this project is like paint to a painter. AI the medium not the subject.