Proving you own a payout wallet, without moving funds
Settlement lands directly in a merchant’s own wallet. That makes one question load-bearing: does this merchant actually control the address they gave us?
Get it wrong and you have built a way to route someone’s revenue to an attacker.
Why not a test transfer
Section titled “Why not a test transfer”The usual answer is to ask for a small transfer from the address. It works, and it is a poor experience: it costs money, it costs a transaction, and it means the merchant has to move funds before they have received any.
It also proves less than it looks. It proves someone could spend from that address at that moment, which is what we want, but at the cost of an on-chain action for something that is only ever a question about a key.
What we do instead
Section titled “What we do instead”The merchant asks for a challenge. The server issues one bound to their account with an HMAC, so it cannot be replayed against a different merchant or reused after it expires. They sign it with the wallet’s key. We verify the ed25519 signature against the address they claim.
No transfer, no fee, no on-chain footprint. Ownership of a key is proven by using the key, which is the whole point of a signature.
The challenge comparison is timing-safe, and the same machinery is what handle claims will reuse, proving control of an address is the same problem whether you are registering a payout wallet or claiming a name.
None of this needs a chain at all, which is the point. There is a second path for merchants who want the link recorded on-chain, and that one is exercised against a real cluster rather than a simulation.