Backend Models

Define and structure database models with clear naming conventions, timestamps, data integrity constraints, appropriate data types, and relationship definitions. Use this skill when creating or modifying database model files in directories like models/, entities/, schemas/, src/models/, app/models/, or any ORM model definition folders. Use this when defining model classes with ORMs like Sequelize, TypeORM, Prisma schema, Django ORM models, Rails ActiveRecord models, SQLAlchemy models, Mongoose schemas, or any database ORM framework, adding or modifying database constraints like NOT NULL, UNIQUE, CHECK constraints, or foreign key relationships to enforce data integrity at the database level, defining model relationships such as hasMany, belongsTo, hasOne, belongsToMany, ManyToMany, or OneToMany associations between entities to represent how data is connected, implementing validation rules at the model layer for data integrity including required fields, format validation, and custom business rule validation, choosing appropriate data types for columns like string, integer, boolean, date, datetime, JSON, text, decimal, UUID, enum based on the data's purpose and how it will be queried, indexing foreign keys and frequently queried columns for query performance to speed up common queries and joins, including createdAt and updatedAt timestamp fields on all tables for auditing and debugging to track when records are created or modified, using singular names for model classes (User, Product, Order) and plural names for database tables (users, products, orders) following ORM conventions, enforcing data integrity rules at the database level using constraints for defense in depth rather than relying solely on application-level validation, defining relationships with appropriate cascade behaviors like CASCADE for dependent deletes where child records should be removed, SET NULL for optional relationships where references can be cleared, or RESTRICT to prevent orphaned records by blocking deletes, balancing database normalization with practical query performance needs to avoid over-normalization that would require excessive joins, and implementing validation at both the model layer and database layer for comprehensive data protection against invalid data.