For the technical-minded who are following along: part of my solution involves injecting an ES6 Proxy to override the HTMLImageElement component with one which I control, and then using that to replace the content of the jigsaw image with a known image in which the rows and columns (that will eventually become jigsaw “pieces”) are individually labelled with their coordinates. This makes it possible to solve moderately rapidly as a human even if the computer-aided tools don’t work… and because it runs entirely “on top” of the site, it doesn’t depend on any particular version of Jigidi’s code.