schema-reviewer_skill

Overview

This skill reviews database schema design focusing on normalization, constraints, indexes, data types, and relationship design. It highlights critical issues, suggests concrete fixes, and produces SQL change snippets you can apply. Use it to validate schema correctness and to plan improvements before deployment.

How this skill works

The skill inspects schema files, migration scripts, and entity definitions to detect normalization problems, missing constraints, poor data types, index gaps, and relationship design flaws. It rates issues by severity, generates specific recommendations and example ALTER/CREATE statements, and can focus on selected review areas (full review, normalization, constraints, index strategy, or data types).

When to use it

• Before DB migrations go to staging or production
• During schema design or table refactoring
• When audits find data integrity or type problems
• Planning index strategy for expected queries
• Validating ORM entity mappings against DB schema

Best practices

• Default to 3NF but selectively denormalize for read-heavy workloads
• Always define PRIMARY KEY and appropriate FOREIGN KEY constraints with ON DELETE/UPDATE rules
• Use DECIMAL for monetary values and TIMESTAMPTZ for timestamps
• Add NOT NULL, CHECK, and UNIQUE where business rules require them
• Index foreign keys and frequent filter/sort columns; prefer composite or partial indexes when appropriate

Example use cases

• Detect repeating group columns and automatically recommend junction or item tables with example SQL
• Find missing primary or foreign keys and produce ALTER TABLE statements to add them
• Identify FLOAT or INT used for money and provide ALTER COLUMN to DECIMAL with checks
• Recommend and render index creation for FK columns, composite filter+order combos, or partial indexes for status-based queries
• Review N:M relations and suggest junction table schema and primary key definitions