Snowflake Query Optimizer Agent
Analyzes and optimizes Snowflake SQL queries using the QUERY_HISTORY and ACCESS_HISTORY views in ACCOUNT_USAGE. Identifies expensive scans, recommends clustering keys, and monitors warehouse credit consumption via WAREHOUSE_METERING_HISTORY.
Analyzes and optimizes Snowflake SQL queries using the QUERY_HISTORY and ACCESS_HISTORY views in ACCOUNT_USAGE. Identifies expensive scans, recommends clustering keys, and monitors warehouse credit consumption via WAREHOUSE_METERING_HISTORY.
npx skills add agentskillexchange/skills --skill snowflake-query-optimizer-agent
The Snowflake Query Optimizer Agent analyzes SQL performance using Snowflake’s ACCOUNT_USAGE schema views. It queries QUERY_HISTORY for execution metrics including bytes scanned, compilation time, queuing time, and spill-to-disk indicators to identify optimization candidates.
Core analysis includes scanning ACCESS_HISTORY for column-level access patterns to recommend optimal clustering keys via ALTER TABLE ... CLUSTER BY. The skill evaluates micro-partition pruning efficiency using SYSTEM$CLUSTERING_INFORMATION() and recommends re-clustering when overlap depth exceeds thresholds.
Cost management features monitor WAREHOUSE_METERING_HISTORY for credit consumption trends, identify idle warehouses with auto-suspend recommendations, and analyze WAREHOUSE_LOAD_HISTORY for right-sizing decisions. The agent generates query profiles using SYSTEM$EXPLAIN_JSON_TO_TEXT() for visual execution plan analysis. It detects common anti-patterns: SELECT *, unnecessary ORDER BY in subqueries, cross joins, and functions on filter columns preventing pruning. Supports materialized view recommendations based on repeated query patterns from QUERY_HISTORY aggregation. Monitors data sharing and replication costs across regions.