gcp-bigquery-optimizer

What This Skill Does

The gcp-bigquery-optimizer skill acts as an expert consultant for your Google Cloud BigQuery environment. It is designed to audit, analyze, and provide actionable strategies to reduce excessive billing and improve query performance. By processing data exported from INFORMATION_SCHEMA and billing reports, the AI identifies hidden inefficiencies such as full table scans, ineffective storage class utilization, and missing materialization opportunities. This skill does not connect directly to your GCP project, ensuring your sensitive infrastructure credentials remain secure while you receive expert-level architectural guidance.

Installation

To install this skill, run the following command in your terminal: clawhub install openclaw/skills/skills/anmolnagpal/bigquery-optimizer

Use Cases

• Cost Audit: Analyze the last 30 days of jobs to pinpoint the top 50 most expensive queries impacting your monthly budget.
• Storage Management: Evaluate dataset sizes and identify opportunities to transition data from active storage to long-term storage, which is significantly cheaper after 90 days of inactivity.
• Performance Tuning: Detect queries that lack proper partition pruning or clustering, which are the primary drivers of unnecessary data scanning and high costs.
• Economics Planning: Evaluate whether your current on-demand pricing model is cost-effective compared to committed slot reservations based on your sustained usage patterns.

Example Prompts

1. "I've attached my query logs from INFORMATION_SCHEMA. Can you identify the top 5 queries that are inflating my daily costs and suggest SQL rewrites for them?"
2. "My BigQuery bill has tripled in the last month. Here is my storage report by dataset. Where should I focus my cleanup efforts?"
3. "I have a dashboard that refreshes every hour scanning a 5TB table. Could a materialized view or scheduled query help reduce my costs here?"

Tips & Limitations

• Security: This skill is strictly instruction-based. It will never request your GCP credentials or attempt to run remote commands against your infrastructure. Always provide anonymized data if you are concerned about project names.
• Data Quality: The quality of the optimization analysis is directly proportional to the detail of the input data. Including both jobs and storage usage provides the most comprehensive optimization strategy.
• Actioning: The AI provides the 'how-to' and the SQL logic; you are responsible for testing and implementing these changes in your production environment.