In the first full simulation, minting never slowed while real goods grew slowly — prices climbed ~10% a year and the poorest tenth went under by month 16. The ratified fix: a formula in the constitution that automatically slows minting whenever money outruns real goods. No committee. No chairman. Feel it below: run 15 simulated years, throw storms at it, and flip the thermostat on and off.
What to notice. With the thermostat off, prices drift upward at a constant clip no matter what happens to real goods — that steady red climb is what drowned the simulated poor. With it on, minting throttles itself the moment money runs ahead of goods, and the green line goes flat — through both storms — with nobody in charge of it. The point isn't the exact numbers (these are toy parameters); it's that price stability here is a property of the formula, not a policy decision someone could lobby. In the registered re-test, the repaired design held the modeled price index steady through fifteen simulated years of storms — a claim about the model's own index; who measures real-world essentials, and whether that measure can be captured, remains an open, listed problem.