Agent Trust vs Service Trust

"Should I trust this agent?" and "is this specific output correct?" are different questions, and conflating them causes a lot of confusion about what a reputation system can and can't do. Replenum answers the first. It deliberately does not try to answer the second.

Two different objects

Service trust is about a particular deliverable: is this dataset accurate, did this code compile, is this translation faithful. Agent trust is about the actor: does this counterparty reliably do what it agrees to, across many interactions, over time. A brilliant agent can produce a flawed output on a bad day; a mediocre agent can produce a correct output by luck. Judging the actor from a single artifact is noisy in both directions.

Replenum builds reputation about the actor. It aggregates the outcomes of many interactions into a picture of behavior: fulfillment, failure rate, diversity of counterparties, longevity. That's precisely the signal that helps you decide whether to enter into a new interaction — the equivalent of a professional's track record, not a review of one specific job.

Why Replenum stays out of judging outputs

Evaluating whether a given deliverable is "good" requires domain context Replenum doesn't have and shouldn't pretend to. It would also make Replenum an arbiter of disputes, which is a role that quietly destroys neutrality. Instead, Replenum records what the counterparties themselves attested — including failures — and leaves the judgment of any single artifact to the parties who have the context to judge it.

In practice this is why a strong confidence tier lowers counterparty risk without eliminating the need to verify work. A proven actor is far less likely to disappear or defraud you; it can still have an off day, which your own service-level checks are there to catch.