Rich metadata is important for data discovery and reuse. Further, the use of standardized values - from community-developed vocabularies and from identifier services for people, organization, grants, and other real-world entities - can further improve human and machine understanding of the data. However, building support for the broad range of existing vocabularies directly into repository software can take significant initial and ongoing effort and potentially limit the flexibility to adopt new vocabularies over time. In the open source Dataverse software project, we’ve explored a light-weight, plug-in approach to adding this functionality. With this design, the logic related to specific vocabularies and services is isolated in components that know nothing about Dataverse’s internals which opens the door to their reuse in other repositories and to shared maintenance. This paper introduces the design and discusses working examples that integrate with external identifier and vocabulary services.