Optimize Core Data Models for iOS Apps

This skill covers Core Data modeling for iOS applications, with deep knowledge of entity design, relationships, performance optimization, and migration strategies. Its entity design fundamentals call for modeling entities around the app's data requirements rather than UI structure, using appropriate attribute types (Integer 16/32/64, Float/Double, String, Boolean, Date, Binary Data), setting default values to avoid null-handling problems, using optional attributes sparingly, and consistent naming conventions. Its relationship best practices include always defining inverse relationships for referential integrity, choosing the right delete rule (Nullify by default, Cascade, Deny, or No Action), using ordered one-to-many sets only when order matters, and considering fetched properties for complex queries that don't fit standard relationships. It covers attribute and index configuration in the Data Model Inspector, one-to-many and many-to-many relationship patterns with generated accessor methods, abstract base entities for shared properties via inheritance, and Transformable attributes backed by NSSecureCoding for complex data types.

Use this skill when designing or optimizing a Core Data model for an iOS app that needs real relationships, complex queries, or migrations - its own guidance is to reach for Core Data only when you need those advanced capabilities, not as a default persistence layer. It covers query optimization (fetchBatchSize, propertiesToFetch with dictionaryResultType, relationshipKeyPathsForPrefetching to avoid faulting), faulting and memory management for batch processing with periodic context refreshes, lightweight migration setup (shouldMigrateStoreAutomatically, shouldInferMappingModelAutomatically, persistent history tracking), and versioning best practices (always add a new model version rather than editing an existing one, test migrations against realistic data volumes). It is not a guide for simple key-value or document persistence - the model is designed around app data requirements with real referential integrity, which is more structure than flat local storage needs.

// Создайте абстрактную базовую сущность для общих свойств
// BaseEntity (Abstract: YES)
@objc(BaseEntity)
public class BaseEntity: NSManagedObject {
    @NSManaged public var createdDate: Date
    @NSManaged public var updatedDate: Date
    @NSManaged public var uuid: String
}

// Конкретные сущности наследуются от BaseEntity
@objc(User)
public class User: BaseEntity {
    @NSManaged public var username: String
    @NSManaged public var email: String
}

@objc(Post)
public class Post: BaseEntity {
    @NSManaged public var title: String
    @NSManaged public var content: String
}

Given an app's data requirements, the skill produces NSManagedObject subclasses with @NSManaged properties and relationships, abstract base entities for shared fields like the one above, generated add/remove accessor extensions for to-many relationships, custom validateValue and validateForInsert overrides for field-level validation (e.g. rejecting an email without an @ or an empty username), optimized NSFetchRequest configurations with batch sizing and relationship prefetching, and lightweight-migration persistent store configuration.

iOS engineers designing or maintaining a Core Data model who need concrete entity, relationship, indexing, and migration patterns rather than starting from the framework docs. It suits apps with real relational data and evolving schemas that need careful migration versioning, and teams that monitor Core Data performance with Instruments and consider NSPersistentCloudKitContainer for CloudKit sync when appropriate.