Not ruled out - but it costs you the thing the immutability was buying. The fixed addresses aren't there to forbid state. They're there to prove a token is genuine iff its covenant's immutable portion matches the terms its genesis committed, confirmed by one Merkle proof of the genesis transaction. No lineage walk, O(1) in the number of transfers. Genesis-binding is what makes that possible - fetch one transaction, compare, done. A payee accumulated from state can't be checked that way, because by definition it isn't in the genesis. So you lose O(1) verification-from-genesis. That's the real cost.. But an auction doesn't need that property. You're not proving "this covenant pays the right creators forever" - you're proving "this refund goes to the address that bid". That address wasn't committed at genesis and shouldn't be; it was written by an enforced transition one hop back. Different question, different proof, and a much cheaper one. So yes: Profile B with n=1, with the payee read out of the covenant's own state rather than its genesis prefix. The piece that's missing is a standard way to locate and validate that payee inside a script - fixed-width fields at constant offsets, where the push opcode equals the declared width, so a reader can walk the state and refuse if any byte disagrees. Just submitted a proposal yesterday, PR #225 on bsv-blockchain/BRCs. It's deployed and doing exactly your operation on mainnet - a fuel depot reads a racing car's 20-byte owner field out of the car's script, validating every push as it walks, sharing no code with the car at all. That's a third-party read of another covenant's state, in production. Which is to say: your auction wants BRC-226's binding and BRC-X's state. They compose.