What problem does it solve? Setting up and operating a managed PostgreSQL database on Render involves many easy-to-miss details: choosing between internal and external connection URLs, respecting immutable creation fields, staying under connection limits, and avoiding destructive Blueprint changes to read replicas. This Skill consolidates those operational rules so you configure and troubleshoot Render Postgres correctly the first time. ## Core Features & Use Cases - Connection Configuration: Choose internal vs external URLs, apply TLS requirements, wire credentials via Blueprint fromDatabase, and set up application-side pooling. - Capacity and Availability Management: Understand storage autoscaling (~90% trigger, 5 GB increments, 12-hour cooldown), RAM-based connection limits, HA requirements, and up to 5 read replicas. - Backup, Recovery, and Tuning: Use daily snapshots, PITR, pg_dump/pg_restore workflows, and pg_stat_statements-based performance analysis. - Use Case: Your Render-hosted API starts throwing "too many connections" errors after a traffic spike. Use this Skill to check active connection metrics via MCP, right-size your app pool, and decide whether to scale the plan or add read replicas. ## Quick Start Ask the assistant to help you connect your Render web service to a Managed Postgres database using the internal connection string and a properly sized connection pool.