While it is widely accepted that AI tools have the ability to significantly increase the productivity of software development, this contribution focuses on its use to improve software evolvability. More specifically, an artifact is presented that leverages Large Language Models (LLMs) through Retrieval-Augmented Generation (RAG} to automatically detect violations of Normalized Systems Theory (NST) evolvability theorems in source code. This artifact is based on a RAG architecture, where a dedicated knowledge base is created with practical examples of NST violations, and an Abstract Syntax Tree (AST) representation is used to convey code structure into natural language. An experimental setup to validate the artifact is described, allowing participants to engage in three experiments through a web application, and the empirical results are presented. Though the limitations of this validation are acknowledged, the results are deemed positive, and some options for future research are identified.