What problem does it solve? Unexpectedly high Postgres network transfer bills usually come from application code fetching more data than it uses. This Skill walks you through finding the queries responsible, fixing the anti-patterns behind them, and verifying the data transfer actually dropped. ## Core Features & Use Cases - Egress Diagnosis with pg_stat_statements: Run diagnostic SQL queries to rank statements by total rows returned, rows per call, call frequency, and execution time to pinpoint the biggest transfer contributors. - Anti-Pattern Fixes: Apply targeted fixes for SELECT * overfetching, missing pagination, high-frequency queries on static data, application-side aggregation, and JOIN duplication. - Verification Workflow: Reset stats, re-run diagnostics, and confirm API responses still match expected shapes after fixes. - Use Case: Your Neon bill jumped this month. Use this Skill to identify that an unpaginated products endpoint with a wide JSONB column is transferring hundreds of megabytes daily, then add column selection and LIMIT/OFFSET pagination to cut the cost. ## Quick Start Ask the AI to analyze your database queries for excessive data transfer and fix the query patterns driving up your Postgres egress costs.
