Trust tiers

How a verified mandate and hosted reputation resolve to a trust tier — premium, trusted, standard, cautious — that your policy gates on.

A verified mandate resolves to a TrustTier, and your policy can gate on it (e.g. holdForReviewBelowTier). Tiers are ordered, worst → best.

Tier	How it's earned
premium	A valid, fully-bound mandate — e.g. a Veto mandate with a good signature, `decision = approve`, `exp > now`, authorized amount ≥ cart total, and a merchant match.
`trusted`	A valid mandate that binds at a lower assurance (reserved for AP2 / ACP once implemented).
`standard`	No qualifying mandate, but the agent is known to reputation.
cautious	No mandate and an unknown agent.

Two scales

Be careful not to conflate the two ranking scales:

TrustTier — cautious < standard < trusted < premium. Derived from the mandate.
ReputationTier — risky < standard < trusted < elite. From the hosted lookup.

Your policy's holdForReviewBelowTier reads the trust tier; minReputationTier reads the reputation tier.

Using tiers in policy

import { BALANCED } from '@veto-protocol/checkout';

const policy = {
  ...BALANCED(),
  minReputationTier: 'trusted',     // require at least a trusted reputation
  holdForReviewBelowTier: 'trusted' // anything below trusted → human review
};

Unknown or malformed tiers always rank lowest — a missing tier can never accidentally clear a floor. The gate fails safe toward the most cautious interpretation.

Two scales

Using tiers in policy

On this page