starrocks-ai-query-autotuner

What problem does it solve?

This Skill removes the manual, error-prone effort of diagnosing slow StarRocks queries by turning execution plans and slow-query signals into concrete optimization actions.

Core Features & Use Cases

• Autonomous EXPLAIN COSTS Diagnosis: Extracts key plan signals, detects full scans, problematic joins (including cartesian patterns), and identifies likely missing partition constraints.
• Stale Statistics Detection & Auto-ANALYZE Guidance: Checks statistics freshness via analysis job status and recommends async ANALYZE when needed.
• Actionable Recommendations: Suggests materialized views, proposes index strategies (bitmap/bloom style), generates join-order and join-method hints, and recommends partition-filter rewrites.

Use Case Example: For a slow analytical query that scans many partitions and joins large datasets without selective filters, the Skill produces a prioritized plan of fixes (partition pruning rewrite, join hinting, and MV/index candidates) to reduce runtime.

Quick Start

Ask the AI to produce an autotuner report for a slow StarRocks SQL query by fetching EXPLAIN COSTS, classifying issues, checking statistics staleness, and returning a prioritized action list.