Synthesizing quines with a relational interpreter in miniKanren is a relatively simple but not infrequently arresting example. We exhibit several related, novel synthesis tasks—either pre-existing programming challenges not solved by miniKanren synthesis, or newly formulated challenges of our own. In doing so we exhibit a relational interpreter for a ``mirrored'' language. These examples demonstrate miniKanren’s potential and versatility as a platform or substrate for experimenting on executable program specifications. As with quines, these examples are themselves interesting from a theoretical perspective, may also suggest other, more practical future applications.