video1.0<iframe src="https://www.loom.com/embed/40be560d55b249ee9c8d15680db677eb" frameborder="0" width="1280" height="960" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>9601280Loomhttps://www.loom.com9601280https://cdn.loom.com/sessions/thumbnails/40be560d55b249ee9c8d15680db677eb-0a1fb18edc1d7c33.gif300.12This Loom explains the Metagraph OCP project, a lineage-aware governed data catalog with a context layer accessible to AI agents via an LCP server. It ingests a company’s business database using an event-driven pipeline that detects updates and syncs the context graph, which includes embeddings and relationships for linked-label RAC retrieval, plus rules-based authentication with admin and analysis roles. The analysis role cannot execute SQL, view PII remote calls, or access PII-marked fields such as email. The demo covers example business tables like query logs and user tables, showing how new tables and downstream dependencies are reflected in the context page as data changes.