Good call, Spencer! Right after publishing this I’d been looking into what math.js was capable of (having chosen it rather than unsafe evals, of course) and you just saved me a step. Fixed in 1.1. Thanks!