sql-performance-review
Databricks SQL performance review
Use this skill when optimizing or reviewing SQL in Databricks SQL.
What to ask for (only if missing)
Ask up to 3 questions total:
- The query text (if not provided)
- The table(s) involved + their sizes (rough order of magnitude) OR the query profile / execution plan
- The desired result constraints (correctness, exactness, latency SLA)
If the user can’t provide sizes/plan, proceed with best-effort heuristics and call out assumptions.
Output format
Use the structure in assets/sql-review-output.md.
More from hubert-dudek/medium
team-templates
Generate standardized team templates (project kickoff, data contract, incident postmortem/RCA, weekly status update, Unity Catalog access request). Use when someone asks for a template, PRD/1-pager, design doc, postmortem, weekly update, data contract, or access request.
8lakehouse-pipeline-design
Create a Databricks lakehouse pipeline design doc (bronze/silver/gold, DLT or Jobs), including SLAs, data quality, Unity Catalog governance, monitoring, and an implementation checklist. Use when designing or reviewing ETL/ELT pipelines, DLT pipelines, streaming ingestion, CDC, or batch jobs on Databricks.
8