Semantic interoperability is crucial for the FAIR Principles and strongly relies on Semantic Artefacts that also need to be FAIR. To achieve this, semantic artefacts require rich, structured, and interoperable metadata. The challenge lies in determining the threshold for "rich metadata" and agreeing on a common minimum set. The H2020 FAIRsFAIR project and the RDA Vocabulary Semantic Services Interest Group addressed this question by developing a "minimal metadata model" for semantic artefacts. In this paper, we present background information, methodology, discussions and workshops which contribute to the establishment of the FAIRsFAIR minimum metadata profile for semantic artefacts. We present an extension of the Metadata for Ontology Description and Publication Ontology (MOD2.0) incorporating this profile as well as its implementation (SemanticDCAT-AP) and its use to build FAIRcat, a prototype of a FAIR Data Point harvesting the content of multiple semantic artefact catalogues.