A Claude Code skill that scopes problems like Peter Naur

This project implements a custom skill for the Claude Code agent that operationalizes Peter Naur's 1985 concept of "programming as theory building" by requiring the AI to first construct and document a comprehensive "problem theory" artifact—detailing the problem's core concepts, relationships, and invariants—before generating any code. It is designed for developers and AI researchers who want to improve the depth and correctness of AI-generated software by enforcing a deliberate, theory-first reasoning process. The project is interesting because it bridges a foundational philosophy of software engineering with modern LLM tooling, potentially reducing superficial or brittle code outputs by forcing the AI to build a coherent mental model of the problem domain.