Methodology: the behaviour band, in the open
No opaque score. The band — new, established or flagged — is computed from the inputs below, and every verify response carries the inputs that produced it.
Stated inputs only
| Input | Meaning |
|---|---|
| Operator verification state | Whether the operator proved domain control (DNS TXT or well-known file) and when. |
| Passport age | Days since the passport was issued to a verified operator. |
| Verifier reports | Reports filed by paid-tier verifier API keys, in stated categories: abuse, fraud, impersonation, spam, policy_violation, good_behavior, other. Each single verifier's effect on one passport is capped at 3, so one reporter cannot dominate a band. |
| Counted verifier checks | Completed, quota-counted verify/lookup calls naming this passport. Failed calls are never counted. |
| MandateGuard outcome signals | Where MandateGuard's authorize flow checks a passport (it reads this registry), the check itself is a counted verifier check on this passport. MandateGuard outcomes are never written into this registry as scores — this registry computes its own band from its own stated inputs. |
The rules (exactly as implemented)
- flagged — capped negative report weight ≥ 5.
- established — verified operator, age ≥ 30 days, capped negative weight < 2, and (≥ 10 counted verifier checks, or ≥ 1 capped positive report, or age ≥ 90 days).
- new — everything else.
Status is separate from band: a passport is unverified, active, suspended or revoked, with a full public status history. A suspended or revoked passport fails verification regardless of band.
Corrections
If you operate a passport and believe a report is wrong, contact us via the contact page with the passport ID and the report reference. Reports stay attributed to the verifier key that filed them.