Methodology

How we score crypto risk

Every AML CRYPTOS risk score is a deterministic blend of on-chain attribution, sanctions-list matching and behavioral clustering. Here's exactly what goes into a score.

Data sources

We ingest from 40+ public datasets including OFAC SDN, EU Consolidated, UN Security Council, UK HMT, Chainabuse, GitHub-published cluster dumps and our own honeypot wallets. New entities are reviewed within 24 hours of publication.

Wallet clustering

Common-input heuristics, change-address detection and behavioral fingerprints group related addresses into a single entity. Each cluster is tagged with up to 200 risk categories.

Risk scoring

The 0–100 score weighs direct exposure (you transacted with a flagged wallet) and indirect exposure (your counterparty did). Direct exposure dominates; indirect exposure decays over each hop.

Audit trail

Every report is timestamped and signed. Exports include a JSON payload with the exact data sources, list versions and clustering rules used at the moment of evaluation.

Coverage

Bitcoin, Ethereum, Tron, Solana, BNB Chain, Polygon, Arbitrum, Optimism, Base and Avalanche are fully supported. ERC-20, BEP-20, SPL and TRC-20 tokens are tracked alongside native assets.

Risk levels

Low0 – 29

Normal retail or institutional activity.

Medium30 – 59

Some exposure to higher-risk counterparties.

High60 – 79

Significant indirect exposure. EDD required.

Severe80 – 100

Direct sanctions or darknet exposure.