LLMs are often discussed in software engineering as tools for code generation, testing, documentation, repair, and model-driven engineering. This session argues for another important use case: LLM-supported ontology engineering for software engineering knowledge itself. Using the Foundations of Software Languages ontology as a running example, I will show how LLMs can support the discovery, classification, linkage, completion, transformation, and validation of a structured knowledge resource for software languages. The FSL ontology organizes concepts spanning programming languages, modeling languages, tools, formal foundations, methods, and their embedding into software engineering activities; its first release explicitly leveraged GenAI within an ontology-engineering workflow.

The session is intended as both a case report and a discussion starter for the LLM4SE community. The FSL repository exposes not only ontology artifacts, queries, and validation material, but also a release history with documented phases and summaries of ChatGPT conversations used during development, making it a concrete evidence base for discussing LLM-supported engineering practice.

The broader question is how software engineers should govern, validate, and reuse LLM-generated or LLM-shaped knowledge artifacts—an issue that aligns closely with LLM4SE’s interest in LLMs across the software lifecycle, model-driven engineering, and the quality of LLM-generated software artifacts.