Treat this page as a workbench, not a blog post. The goal is to extract a reusable mental model from the source and carry it into the rest of the Fabrknt stack.