aniruddhaadak80/skills/databases-schema-design-basics-postgres

Design schemas for change (PostgreSQL)

Normalize enough to stay consistent, denormalize deliberately where reads demand it.

Track: ⚙️ Backend Engineering · Domain: Databases · Level: foundation · ~35 min

Who this is for: Backend Engineers, API Developers, Platform Engineers, Full-Stack Developers

When to Use This Skill

Normalize enough to stay consistent, denormalize deliberately where reads demand it. Use it whenever a matching task appears in conversation — the agent loads these instructions on demand.

Steps

1. Identify entities and relationships; write 10 queries you must answer well
2. Normalize to 3NF by default; document any intentional denormalization
3. Use UUIDs or bigint PKs; never emails or natural keys
4. Add created_at/updated_at with DB-level defaults on every table
5. Constrain with FKs, CHECKs, and NOT NULL — let the DB defend integrity
6. Review growth estimates: which table hits 100M rows first, plan partitioning now
7. Prefer text with CHECK over varchar(n) unless limits are contractual
8. Use partial indexes for hot filtered subsets

Common Pitfalls

• Soft-delete columns complicating every unique constraint
• Storing money as float instead of integer minor units

Commands

Install with skills CLI

npx skills add aniruddhaadak80/skills --skill databases-schema-design-basics-postgres

Install globally

npx skills add aniruddhaadak80/skills --skill databases-schema-design-basics-postgres -g

Part of aniruddhaadak80/skills · Browse all at https://skills.sh/aniruddhaadak80/skills