GetCassis

Audits a dbt project for AI agent readiness, identifying potential errors an agent might make with the data. It generates a prioritized report detailing failure modes like incorrect numbers, tables, or joins, using a scalable two-pass architecture with parallel subagents.

Use this skill to query data with Cassis (which grounds the question against the user's project ontology and returns SQL, a plan to approve, or an ontology gap to clarify) and Snowflake (which executes the SQL).