Type-Driven Prompt Programming: From Typed Interfaces to a Calculus of Constraints (2508.12475v1)

Abstract: Prompt programming treats LLM prompts as software components with typed interfaces. Based on a literature survey of 15 recent works from 2023 to 2025, we observe a consistent trend: type systems are central to emerging prompt programming frameworks. However, there are gaps in constraint expressiveness and in supporting algorithms. To address these issues, we introduce the notion of Lambda Prompt, a dependently typed calculus with probabilistic refinements for syntactic and semantic constraints. While this is not yet a full calculus, the formulation motivates a type-theoretic foundation for prompt programming. Our catalog of 13 constraints highlights underexplored areas in constraint expressiveness (constraints 9 through 13). To address the algorithmic gap, we propose a constraint-preserving optimization rule. Finally, we outline research directions on developing a compiler for prompt programs.