GuardPay

What we analyze

Limits first: the reliability score is computed from the data available at check time. Whatever could not be fetched is reported as not fetched, never as clean; every report is verifiable by fingerprint at /verify. An address check is dozens of independent questions to lists, the blockchain, and the wallet's behaviour. Below are all groups of checks. Weights and thresholds are not published — scammers would calibrate against them. The full list of triggered and clean factors is visible in every report, free. The score is computed by GuardPay's own engine, not resold from someone else: every factor in the report names its source.

Sanctions and state lists
Address matches against sanctions registers and state lists. A match is an established fact: it is shown as a separate badge and is never softened by good behaviour.
Flagged-address base and community reports
Our continuously growing base of addresses from vetted sources and public reports. A report affects the score only after confirmation — by a quorum of independent submissions or manual moderation.
Stablecoin issuer freezes
A direct on-chain question to the USDT and USDC contracts: is the address frozen by the issuer. A payment to a frozen address technically goes through — and stays there forever; the report says so before the transfer, not after.
Wallet behaviour over time
Inflow and outflow across 24 hours / 7 / 30 days, bursts after dormancy, circadian rhythm (a person or a round-the-clock bot), structured amounts, instant sweeps.
Counterparties and source of funds
Who paid the wallet and whom it paid: senders screened against the bases one and two hops upstream. What matters is the source of funds, not the distance — service conduits do not turn someone else's guilt into yours.
Counterfeit tokens and baits
Tokens carrying a real dollar symbol on a foreign contract, "we will unfreeze your wallet" baits. A counterfeit is identified by the contract address — anyone can type a symbol.
Approvals and draining
Granted approvals, spending through them, drainer patterns: victim wallets and collection addresses.
Service attribution
Recognising exchanges and services: a KYC venue's deposit address and a one-off wallet are different risks, and the report tells them apart.
Age and lifecycle
First visible operation, the wallet's first funder, its role in the chain (personal wallet, service, transit) — from observable history, with the observation limits stated.
What we do NOT claim
  • "Not found" does not mean "clean". The report lists what exactly was and was not checked — in numbers, not fine print.
  • "Source did not answer" does not mean "empty". A check that failed is called failed — it is never passed off as a clean result.
  • Receiving money is not guilt. Anything can be sent to a wallet without its consent; incoming transfers alone do not sink the score.
  • The score is a model estimate, not a verdict. Established facts (sanctions, freezes) are shown separately from behavioural conclusions.
  • Window limits are named. If the history is longer than the window read, the report says which slice it saw and when it read it.
Check an address — free
GuardPay — AML/KYT screening for TRON and EVM networks (Ethereum, BNB Chain, Polygon, Arbitrum One, Base). The report is a model estimate plus public on-chain facts, not a legal opinion. We are not affiliated with stablecoin issuers.
GuardPay