The liquidity-enforcement issue arises because the protocol assumes the hunt will be funded with enough ETH to cover all rewards, but the contract itself does not actively enforce that invariant at deployment time. The constructor accepts arbitrary `msg.value` and only validates the verifier address, even though the contract hardcodes a reward of 10 ether and a maximum of 10 treasures, implying an expected full funding target of 100 ether; the README likewise states that the contract is expected to be funded with enough ETH to cover all rewards and notes that the default deployment flow uses 100 ether. Although the deployment script sets `DEFAULT_INITIAL_FUNDING = 100 ether`, it also allows that amount to be overridden via `vm.envOr("INITIAL_FUNDING", DEFAULT_INITIAL_FUNDING)`, and the only post-deployment balance check is that the contract balance equals the chosen `initialFunding`, not that the chosen amount is actually sufficient to fund the full hunt. As a result, the system can be deployed in an underfunded state after which otherwise valid claims will begin reverting with `NotEnoughFunds()` once the balance drops below a single reward payment.