The weakest security check in crypto might be: “first 4 characters match.” “last 4 match.” send. That’s not verification. That’s pattern recognition with money attached to it. Address poisoning works because wallets still expect humans to distinguish strings they were never designed to read. And telling people to “double-check harder” doesn’t really fix the UX. That’s why @Americanfort_io Send-to-Name™ approach makes sense to me from a security angle, not just a convenience angle. Instead of: copy address → compare characters → maybe send a test tx → hope the payment starts with a readable @name. On supported flows, the wallet derives a fresh receiving address underneath and the transfer still settles on the native chain. The interesting part is what disappears: the user no longer has to be the checksum. Good security doesn’t just warn people about dangerous actions. It removes the dangerous decision from the flow. If safely sending money depends on humans comparing 42 characters… the UX itself is part of the vulnerability.