What problem does it solve? Creating and exploring Fabric IQ Ontology items requires composing complex base64-encoded definition envelopes, managing long-running operations, and correctly mapping entity types, data bindings, and relationship types against lakehouse and eventhouse sources, which is error-prone when done by hand. ## Core Features & Use Cases - Ontology Authoring: Create Ontology items, add entity and relationship types, and bind them to lakehouse or eventhouse tables through a preview-and-confirm flow before any write is persisted. - Ontology Consumption: Enumerate entity types, bindings, lineage, and relationships, extract grounding context JSON, and perform graph walks over an existing ontology in read-only mode. - Query Routing: Translate ontology property names into physical source columns and delegate data queries to the matching per-datasource skill (SQL endpoint, Spark, or Eventhouse KQL). - Use Case: A data engineer asks to create an ontology with Customer and Order entity types bound to a lakehouse; the skill resolves workspace and item IDs, composes the definition parts, previews the change set, and persists it via updateDefinition after confirmation. ## Quick Start Ask the assistant to list the entity types and their data bindings in a named Fabric ontology, or to create a new ontology item with specified entity types bound to a lakehouse table.