Why does this skill recommend BIGINT over UUID for primary keys?

BIGINT GENERATED ALWAYS AS IDENTITY is generally more storage-efficient and provides better indexing performance than UUIDs, which should be reserved for cases requiring global uniqueness or opacity.

Does this skill help with PostgreSQL performance tuning?

Yes, it provides specific guidance on indexing foreign keys, managing MVCC bloat, using fillfactors for update-heavy tables, and implementing partitioning for large datasets.

Can Claude help me use advanced PostgreSQL types like JSONB?

Absolutely. This skill includes best practices for using JSONB for semi-structured data, including when to use it and how to index it effectively with GIN indexes.

How does this skill handle time-series data?

It provides patterns for time-based partitioning, the use of TIMESTAMPTZ, and recommendations for tools like TimescaleDB for automated partitioning and compression.

PostgreSQL Table Design

Name: PostgreSQL Table Design
Author: HermeticOrmus

byHermeticOrmus

0•

Database Management

Architects high-performance PostgreSQL schemas using industry-standard best practices for data integrity and scalability.

This skill provides Claude with comprehensive expertise in PostgreSQL database architecture, covering everything from initial normalization (3NF) to advanced performance tuning. It enforces modern standards like identity columns over serial types, TIMESTAMPTZ for temporal data, and strategic indexing for complex query paths. By incorporating deep knowledge of PostgreSQL 'gotchas'—such as the lack of automatic foreign key indexing and the nuances of MVCC—this skill ensures that your database designs are not only semantically correct but also optimized for production-grade workloads, including insert-heavy time-series data and update-heavy transactional systems.

Key Features

01Expert data type selection including JSONB, Range types, and pgvector for embeddings

02Comprehensive indexing strategies (B-tree, GIN, GiST, BRIN, and Partial indexes)

03Modern partitioning techniques (Range, List, Hash) for massive datasets

040 GitHub stars

05Standardized normalization and strategic denormalization patterns

06Performance optimization for high-concurrency and update-heavy workloads

Use Cases

01Implementing specialized data models using full-text search, geometry, or network types

02Refactoring existing database structures to resolve performance bottlenecks and bloat

03Designing robust relational schemas for new web applications or microservices

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add hermeticormus/floreserlife postgresql

For use in Claude.ai and ChatGPT

Download Skill