modelshow
Double-blind comparison of AI model responses — query models in parallel, judge anonymized outputs, rank on merit. Trigger with "mdls" or "modelshow".