Why a screening policy reached its verdict: which layer matched, and what
it matched on in the provider's own vocabulary.
Only provider and matched are always present. The rest depend on the
provider and on which leaf actually fired — a floor match on a Chainalysis
cluster carries a category and no riskType.
The provider whose response produced this verdict.
Available options:
trm-wallet-screening,
chainalysis-address-screening,
elliptic-address-screening,
dynamic-sanctions-screening
Which layer decided. floor is the non-configurable sanctions
baseline, address an allow/deny list entry, customer one of the
environment's own rules, default the terminal allow. On this event
it is always customer or address; a floor block emits
wallet.sanctions.blocked instead.
Available options:
floor,
address,
customer,
default
The policy generation that produced this verdict. Scoped to one
policy, so a sign-in verdict and a transaction verdict in the same
environment can legitimately carry different revisions.
The matched rule, verbatim as evaluated. Present only when matched
is customer: the floor and the terminal allow are generated from
code, so there is nothing stable to point at. Deliberately untyped —
nothing routes on its internal shape, and it is carried for the audit
record.
The provider's category string, verbatim, in the provider's own casing.
Maximum string length: 128
TRM only. OWNERSHIP, COUNTERPARTY or INDIRECT.
Maximum string length: 64
TRM only. The numeric risk score the matched signal carried.
Chainalysis only. The risk level the matched signal carried.
Maximum string length: 32
Human-readable rendering of what matched, for display and logs.
DISPLAY ONLY AND NON-CONTRACTUAL: the wording may change in any
release. Do not parse it — route on category, provider and
matched instead.
Maximum string length: 256
Last modified on October 6, 2026