Specialized skill for optimizing cloud data INFRASTRUCTURE costs and resource utilization at the platform level.
Use when: rightsizing warehouses, configuring auto-suspend policies, optimizing storage formats, setting budget alerts, or forecasting cloud spend.
Do NOT use for query-level cost analysis — use `finding-expensive-queries` instead.
Do NOT use for optimizing individual SQL queries — use `optimizing-query-text` or `optimizing-query-by-id` instead.
설치
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
Specialized skill for optimizing cloud data INFRASTRUCTURE costs and resource utilization at the platform level.
Use when: rightsizing warehouses, configuring auto-suspend policies, optimizing storage formats, setting budget alerts, or forecasting cloud spend.
Do NOT use for query-level cost analysis — use `finding-expensive-queries` instead.
Do NOT use for optimizing individual SQL queries — use `optimizing-query-text` or `optimizing-query-by-id` instead.
license
MIT
metadata
{"author":"dataops-agent","version":"2.0.0"}
Cloud FinOps for Data Infrastructure
Platform-level cost optimization. For query-level cost analysis, use finding-expensive-queries. For SQL optimization, use optimizing-query-text.
When to Apply
Use this skill when:
Rightsizing Snowflake warehouses (XS → XL) and scaling policies
Configuring auto-suspend/auto-resume policies
Evaluating storage formats (Parquet, Iceberg, Delta) for cost impact
Setting up budget alerts and cost anomaly detection
Forecasting cloud spend for data platforms
Comparing pricing models (on-demand vs reserved, standard vs enterprise)
Implementing resource tagging/attribution for cost allocation
Scope Boundaries
Task
Correct Skill
"Find our most expensive queries"
finding-expensive-queries
"Optimize this slow query"
optimizing-query-text
"Our warehouse costs are too high"
this skill ✅
"Should we upsize the warehouse?"
this skill ✅
"Set up budget alerts"
this skill ✅
Best Practices
Attribution: Ensure all resources are tagged to attribute costs to teams/products.
Warehouse Suspend/Resume: Configure aggressive auto-suspend policies for unused compute.
Storage vs. Compute: Optimize data formats and partitioning to reduce bytes scanned.
Continuous Monitoring: Set budget alerts and threshold limits.
Execution Pattern
Analyze the infrastructure configuration or query cost patterns.
Identify wasted resources or inefficient scaling policies.
Provide actionable recommendations and configuration code to reduce costs.
