Nicely done! I wonder how AI would do writing code for this Spinors@Mathematica package for the spinor-helicity formalism: https://arxiv.org/abs/0710.5559.
I did a paper looking at how to extend BCFW recursion to 1-loop unitarity methods in QCD. I wonder if there's any low-hanging fruit for recycling this tree-level result into 1-loop results. I'm trying to remember what Mathematica code I used to do the 1-loop calculations... It's on a dead harddrive somewhere.
