Companion sketchnote from the patch-gap paper (CC BY 4.0). Full-size PNG.
A Python, stdlib-only toy companion to The Governance Patch-Gap. It generates a discovery rate Rd from a label-blind interface searcher on a Highly Optimized Tolerance (HOT) rule graph. The patch-gap G = Rd/Rp and the stocks Sknown and Sactive are outputs of that toy, not fitted measurements of any legal system.
The paper defines G and maps legal open texture onto HOT / robust-yet-fragile design. This software exists so those qualitative distinctions can be run: a searcher produces Rd, a patcher produces Rp, and G and S emerge. It does not estimate Congress, courts, or any statute book.
Version 0.2.0 adds four gates without changing the toy dynamics: Lean 4 proofs of the paper's §4.3 stock algebra; a recomputation of the §4.4 regime arithmetic against the paper text; a 162-cell searcher sweep frozen before its first run; and a figure/prediction ledger that fails on drift. Cite the paper concept for the argument; cite this software concept for the runnable artifact.
AI governance · open texture · reward hacking · regulatory arbitrage · highly optimized tolerance · governance patch-gap · legal exploit discovery · computational claim gate · synthetic rule graph