Authenticated write access to the catalogue REST API (organization-scoped), Linked Open Data via djangordf (CIDOC-CRM/SKOS RDF models, content-negotiated Turtle/JSON-LD endpoints, prune + Celery sync, optional Fuseki/SPARQL), and an Elasticsearch search backend for the catalogue (Postgres full-text search stays the default).