Abstract. The XRP Ledger (XRPL) is a high-performance blockchain built around a fixed-function transaction model rather than general-purpose smart contracts. This architecture provides strong guarantees of determinism, efficiency, and operational predictability, at the cost of expressive flexibility. This talk presents a principled approach to introducing programmability on the XRPL while preserving its core safety properties, with hints at some future directions. We articulate design principles for constrained ledger-level programmability, emphasizing bounded computation, explicit resource accounting, deterministic execution, and composability with existing primitives. As a concrete realization of these principles, we introduce Smart Escrows, a programmable conditional settlement mechanism that extends the XRPL’s native escrow model without introducing unbounded execution semantics. We then describe early exploratory work on zero-knowledge-anchored contract models, in which expressive off-chain computation is paired with succinct on-chain verification. This approach seeks to expand contract expressiveness while maintaining predictable and high-assurance execution. We outline key architectural trade-offs and the challenges they pose. We conclude our talk with some open research questions related to proof verification costs, state binding, and interoperability.