Source: Snowflake Cortex Analyst Verified Query Repository Docs
Snowflake documentation for the Verified Query Repository: curated question-SQL pairs stored in a semantic model that Cortex Analyst reuses when a new question is similar.
Contents
- What: official product documentation
- Author/publisher: Snowflake
- Date: living document; read 2026-11-14
- Original: Verified Query Repository
Summary
Cortex Analyst is Snowflake’s managed natural-language-to-SQL service driven by a semantic model (logical tables, columns, descriptions). The Verified Query Repository (VQR) is a list of human-checked question/SQL pairs in that semantic model. When a user asks something similar to a verified question, Cortex Analyst uses the verified SQL, which raises accuracy and trust.
Key claims
- Each verified query has
name,question,sql, optionalverified_at,verified_byanduse_as_onboarding_question. - Verified SQL must use the semantic model’s logical table and column names, not physical names.
- Invalid or wrong verified queries can hurt accuracy, so run and check each before adding.
- The REST response indicates whether a verified query was used.
- Snowsight proposes new verified queries from user behaviour; a Streamlit generator helps author them.