What problem does it solve? Teams need to enforce organization-specific naming conventions, data type rules, and metadata standards on Snowflake semantic views, but built-in audits cannot cover every custom rule. This Skill lets users express validation rules in plain language and checks the semantic view against them. ## Core Features & Use Cases - Natural Language Rule Parsing: Converts free-form criteria like "all revenue metrics must contain 'revenue' in the name" into concrete validation checks. - Component-Level Inspection: Retrieves tables, columns, measures, time dimensions, and relationships from the semantic view YAML via semantic_view_get.py and flags violations. - Compliance Reporting: Presents per-criterion compliance rates, violation details with expected versus actual values, and prioritized recommendations. - Use Case: A data team requires every measure description to mention its aggregation function. The user enters that rule, the audit scans all measures, and reports which ones fail with fix suggestions. ## Quick Start Ask the assistant to run a custom criteria audit on your semantic view and provide rules such as "verify all date columns use DATE or TIMESTAMP data types".
