Great observation about starting with proofs. This is especially true when you step outside of the code and look at the problems being solved: if you don't alr

Great observation about starting with proofs. This is especially true when you step outside of the code and look at the problems being solved: if you don't already have a proof of concept, and even before that, if you don't have a sound problem to solve, you will not be able to produce intelligible proofs and axioms will be difficult if not impossible. Mechanically, I'll prove to myself something works the way I want or need to in a REPL and then implement the axiom from that... which is sorta in the right direction :)