schema-design-advisor

What problem does it solve?

Poorly designed ClickHouse table schemas with suboptimal ORDER BY keys, incorrect data types, missing indexes, and inefficient partitioning lead to slow query performance, excessive storage usage, and degraded merge operations. This skill eliminates these issues by providing evidence-based, workload-aligned schema optimization recommendations.

Core Features & Use Cases

• Evidence-First Inspection: Guides users to gather table schema, parts summary, column compression metrics, and cardinality data before making any recommendations to ensure suggestions are grounded in actual usage.
• Comprehensive Optimization Guidance: Covers ORDER BY design, partition key selection, column type right-sizing, compression codec selection, skip index configuration, and projection/materialized view recommendations for all common ClickHouse schema optimization scenarios.
• Use Case: For a ClickHouse table with slow filter queries and 2x higher than expected storage costs, use this skill to analyze its actual usage patterns and receive actionable ALTER TABLE commands that cut query latency by 40% and reduce storage overhead.

Quick Start

Use the schema-design-advisor skill to analyze your ClickHouse table's current schema, parts, and column metrics, then receive prioritized, actionable recommendations for ORDER BY keys, partition strategies, data types, codecs, skip indexes, and projections to improve query performance and reduce storage costs.