This is for anyone designing Cassandra or DynamoDB schemas who keeps running into hot partitions or wondering why their queries are slow. It walks through the core mental shift from relational modeling to query-first design, explaining partition keys, sort keys, single-table patterns, and when to denormalize. The comparison table between SQL and NoSQL mindsets is genuinely helpful, and the anti-pattern checklist (like avoiding ALLOW FILTERING in Cassandra) will save you from common mistakes. If you're moving from Postgres to a distributed system and still thinking in joins and normalization, this reorients you quickly.
npx skills add https://github.com/davila7/claude-code-templates --skill nosql-expert