Overview
Judgments
Section titled “Judgments”Judgments: yes or no, with an optional reason, about one item of a finished run’s output, optionally recorded under a judge class (list, get, create, unregister). Each judgment keeps copies of what was judged (the run’s input and output, and the item) so they outlive the run’s own retention. Who judged comes from the authenticated caller, never the body; an app judging for one of its users passes that user’s opaque id as participantId. One live judgment per run, item key, caller and participant: judging again supersedes the earlier one, which stays as history. Caller-plugged via JudgmentRegistryBinding.