Unity Catalog 0.6.0 added governed metric views, Spark 4.2 support, and a 24-hour token expiry default.

UnityCatalog 0.6.0 introduced metric views and SQL views directly as governed catalog objects for semantic query layers, added Apache Spark 4.2 support, and set access tokens to expire after 24 hours by default [15]. Delta Lake 4.4.0 expanded Unity Catalog Delta API integration to Delta Kernel and an experimental Flink connector [16]. In addition, the Python library release v1.6.3 added support for custom token credentials and dropping NOT NULL constraints on catalog tables [10], while tooling updates enabled Unity Catalog in Remote SSH mode [34].

Databricks widened catalog coverage to agent workflows and business semantics. The platform introduced Pages within Unity Catalog Semantics, allowing teams to curate definitions of terms and metrics so Genie AI can ground responses with verified sources [5]. Architectural guidance detailed methods to govern AI agents, tools, and Model Context Protocol servers through centralized catalog privileges and audit logs [18, 32]. Genie Agents also gained patterns to query across structured tables and unstructured Unity Catalog Volumes under caller identity, preserving row filters and column masks [30].

Community discussions focused on configuration edge cases and permissions. Developers addressed operational snags with Unity Catalog schemas in Declarative Automation Bundles [11], and users reported issues where service credential token requests rejected api-scoped URI strings [29]. Other practitioners troubleshot Azure managed identity failures during initial setup [28], examined storage boundaries for schema secrets [17], and investigated how external applications can query catalog metric views [27].