Core Data logo

Core Data

CommunityPopular
dpearson2699
core-data

Build, review, or improve Core Data persistence in apps that have not adopted SwiftData. Use when working with NSManagedObject subclasses, NSFetchedResultsController for list-driven UI, NSBatchInsertRequest / NSBatchDeleteRequest / NSBatchUpdateRequest for bulk operations, NSPersistentHistoryChangeRequest for persistent history tracking and multi-target sync, NSStagedMigrationManager for staged schema migrations (iOS 17+), NSCompositeAttributeDescription for composite attributes (iOS 17+), or when integrating Core Data threading with Swift Concurrency. For Core Data + SwiftData coexistence or migration, see the swiftdata skill instead.

Overview

Publisherdpearson2699
Repositoryswift-ios-skills
Skill namecore-data
Stars
1.2K
Forks
61
Bundled files
3
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 3 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by dpearson2699 on GitHub. Read the source before you install it.

Installation

Install the Core Data AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/dpearson2699/swift-ios-skills/tree/main/skills/core-data
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/dpearson2699/swift-ios-skills.git /tmp/swift-ios-skills
mkdir -p .claude/skills
cp -r /tmp/swift-ios-skills/skills/core-data .claude/skills/core-data
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Core Data in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Core Data on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Core Data is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Core Data

Build and maintain data persistence using Core Data for apps that have not adopted SwiftData. Covers stack setup, concurrency, batch operations, NSFetchedResultsController, persistent history tracking, staged migration, and testing.

Contents

Stack Setup

NSPersistentContainer encapsulates the Core Data stack.

Docs: NSPersistentContainer

swift
import CoreData

final class CoreDataStack: @unchecked Sendable {
    static let shared = CoreDataStack()

    let container: NSPersistentContainer

    private init() {
        container = NSPersistentContainer(name: "MyAppModel")
        container.loadPersistentStores { _, error in
            if let error { fatalError("Core Data store failed: \(error)") }
        }
        container.viewContext.automaticallyMergesChangesFromParent = true
        container.viewContext.mergePolicy = NSMergeByPropertyObjectTrumpMergePolicy
    }

    var viewContext: NSManagedObjectContext { container.viewContext }

    func newBackgroundContext() -> NSManagedObjectContext {
        container.newBackgroundContext()
    }
}

For CloudKit sync, use NSPersistentCloudKitContainer instead.

Concurrency and Threading

Core Data contexts are bound to queues. The viewContext is on the main queue; background contexts operate on private queues.

Docs: NSManagedObjectContext

Rules:

  • Always use perform(_:) or performAndWait(_:) when accessing a context off its own queue.
  • Never pass NSManagedObject instances across context or thread boundaries. Pass NSManagedObjectID instead and re-fetch.
  • Set automaticallyMergesChangesFromParent = true on the viewContext.
swift
// Writing on a background context
func updateTrip(id: NSManagedObjectID, newName: String) async throws {
    let context = CoreDataStack.shared.newBackgroundContext()
    try await context.perform {
        guard let trip = try context.existingObject(with: id) as? CDTrip else {
            throw PersistenceError.notFound
        }
        trip.name = newName
        try context.save()
    }
}

Swift Concurrency Integration

NSManagedObjectContext.perform(_:) has an async throws overload (iOS 15+). Avoid marking NSManagedObject subclasses as Sendable.

swift
func importItems(_ records: [ItemRecord]) async throws {
    let context = CoreDataStack.shared.newBackgroundContext()
    try await context.perform {
        for record in records {
            let item = CDItem(context: context)
            item.id = record.id
            item.title = record.title
        }
        try context.save()
    }
    // After save completes, viewContext auto-merges if configured
}

Do not use @unchecked Sendable on managed objects. If you need cross-boundary communication, pass the objectID (which is Sendable) and re-fetch:

swift
let objectID = trip.objectID  // Sendable
Task.detached {
    let bgContext = CoreDataStack.shared.newBackgroundContext()
    try await bgContext.perform {
        let trip = try bgContext.existingObject(with: objectID) as! CDTrip
        trip.isFavorite = true
        try bgContext.save()
    }
}

NSFetchedResultsController

Efficiently drives UITableView / UICollectionView from a Core Data fetch request, with built-in change tracking and optional caching.

Docs: NSFetchedResultsController

swift
import CoreData
import UIKit

class TripsViewController: UITableViewController, NSFetchedResultsControllerDelegate {

    private lazy var fetchedResultsController: NSFetchedResultsController<CDTrip> = {
        let request: NSFetchRequest<CDTrip> = CDTrip.fetchRequest()
        request.sortDescriptors = [
            NSSortDescriptor(keyPath: \CDTrip.startDate, ascending: false)
        ]
        request.fetchBatchSize = 20

        let controller = NSFetchedResultsController(
            fetchRequest: request,
            managedObjectContext: CoreDataStack.shared.viewContext,
            sectionNameKeyPath: nil,
            cacheName: "TripsCache"
        )
        controller.delegate = self
        return controller
    }()

    override func viewDidLoad() {
        super.viewDidLoad()
        try? fetchedResultsController.performFetch()
    }

    // MARK: - UITableViewDataSource

    override func numberOfSections(in tableView: UITableView) -> Int {
        fetchedResultsController.sections?.count ?? 0
    }

    override func tableView(_ tableView: UITableView, numberOfRowsInSection section: Int) -> Int {
        fetchedResultsController.sections?[section].numberOfObjects ?? 0
    }

    override func tableView(_ tableView: UITableView, cellForRowAt indexPath: IndexPath) -> UITableViewCell {
        let cell = tableView.dequeueReusableCell(withIdentifier: "TripCell", for: indexPath)
        let trip = fetchedResultsController.object(at: indexPath)
        cell.textLabel?.text = trip.name
        return cell
    }

    // MARK: - NSFetchedResultsControllerDelegate (diffable)

    func controller(
        _ controller: NSFetchedResultsController<any NSFetchRequestResult>,
        didChangeContentWith snapshot: NSDiffableDataSourceSnapshotReference
    ) {
        let snapshot = snapshot as NSDiffableDataSourceSnapshot<String, NSManagedObjectID>
        dataSource.apply(snapshot, animatingDifferences: true)
    }
}

Key points:

  • The fetch request must have at least one sort descriptor.
  • Call deleteCache(withName:) before changing the fetch request predicate or sort descriptors, or set cacheName to nil.
  • The diffable snapshot delegate method (didChangeContentWith:) is available iOS 13+ and is preferred over the older per-change callbacks.
  • After a context reset(), call performFetch() again.

Batch Operations

Batch operations execute at the SQL level, bypassing the managed object context. They are fast but don't trigger context notifications automatically.

NSBatchInsertRequest (iOS 13+)

Docs: NSBatchInsertRequest

swift
func batchImport(_ records: [[String: Any]]) async throws {
    let context = CoreDataStack.shared.newBackgroundContext()
    try await context.perform {
        let request = NSBatchInsertRequest(
            entity: CDTrip.entity(),
            objects: records
        )
        request.resultType = .objectIDs
        let result = try context.execute(request) as? NSBatchInsertResult
        if let ids = result?.result as? [NSManagedObjectID] {
            NSManagedObjectContext.mergeChanges(
                fromRemoteContextSave: [NSInsertedObjectsKey: ids],
                into: [CoreDataStack.shared.viewContext]
            )
        }
    }
}

NSBatchDeleteRequest (iOS 9+)

Docs: NSBatchDeleteRequest

swift
func deleteOldTrips(before cutoff: Date) async throws {
    let context = CoreDataStack.shared.newBackgroundContext()
    try await context.perform {
        let fetchRequest: NSFetchRequest<NSFetchRequestResult> = CDTrip.fetchRequest()
        fetchRequest.predicate = NSPredicate(format: "endDate < %@", cutoff as NSDate)
        let request = NSBatchDeleteRequest(fetchRequest: fetchRequest)
        request.resultType = .resultTypeObjectIDs
        let result = try context.execute(request) as? NSBatchDeleteResult
        if let ids = result?.result as? [NSManagedObjectID] {
            NSManagedObjectContext.mergeChanges(
                fromRemoteContextSave: [NSDeletedObjectsKey: ids],
                into: [CoreDataStack.shared.viewContext]
            )
        }
    }
}

NSBatchUpdateRequest (iOS 8+)

swift
func markAllTripsAsNotFavorite() async throws {
    let context = CoreDataStack.shared.newBackgroundContext()
    try await context.perform {
        let request = NSBatchUpdateRequest(entity: CDTrip.entity())
        request.propertiesToUpdate = ["isFavorite": false]
        request.resultType = .updatedObjectIDsResultType
        let result = try context.execute(request) as? NSBatchUpdateResult
        if let ids = result?.result as? [NSManagedObjectID] {
            NSManagedObjectContext.mergeChanges(
                fromRemoteContextSave: [NSUpdatedObjectsKey: ids],
                into: [CoreDataStack.shared.viewContext]
            )
        }
    }
}

Always merge changes back into relevant contexts after batch operations. Batch delete does not enforce the Deny delete rule.

For destructive or retryable batch work, use a proof loop: preflight the predicate and expected count, execute with an object-ID result type, merge IDs into live contexts, refetch, and assert the postcondition. On failure, restore a pristine fixture or prove the operation is idempotent before retrying; never blindly rerun a partially completed batch.

Persistent History Tracking

Track store-level changes across targets (app, extensions, widgets) and processes. The core workflow is:

Docs: NSPersistentHistoryChangeRequest

  1. Enable persistent history and remote-change notifications before loading the store.
  2. Observe changes and fetch transactions after the target's durable token.
  3. Merge transaction notifications into live contexts, then persist the new token.
  4. Purge only history that every relevant consumer has processed.

Load persistent-history.md when implementing the store options, observer, token persistence, merge loop, or purge policy.

Staged Migration

NSStagedMigrationManager (iOS 17+) sequences schema migrations through ordered lightweight or custom stages. Stage inputs use compiled model-version checksums, not model names. Apps supporting systems below iOS 17 need the lightweight migration or mapping-model path.

Docs: NSStagedMigrationManager

Load staged-migration.md when building the ordered stages, model references, custom handler, and persistent-store option.

Composite Attributes

iOS 17+ supports composite attributes: groups of sub-attributes on an entity that act as a single logical unit. Define them in the model editor by adding a Composite type attribute and nesting sub-attributes beneath it.

Docs: NSCompositeAttributeDescription

Composite attributes map to Codable structs in SwiftData coexistence scenarios.

SwiftData Boundary

Use the swiftdata skill for Core Data + SwiftData coexistence or migration implementation. Before handing off, preserve these Core Data boundaries:

  • SwiftData must point at the existing persistent store URL when it is meant to share or migrate Core Data data.
  • Shared persisted data must keep entity names, property names, types, and schema compatible across the Core Data model and SwiftData @Model classes.
  • Map renamed persisted properties with SwiftData @Attribute(originalName:).

Testing

In-Memory Store for Tests

swift
import CoreData
import Testing

struct CoreDataTests {
    func makeTestContainer() throws -> NSPersistentContainer {
        let container = NSPersistentContainer(name: "MyAppModel")
        let description = NSPersistentStoreDescription()
        description.type = NSInMemoryStoreType
        container.persistentStoreDescriptions = [description]

        var loadError: Error?
        container.loadPersistentStores { _, error in loadError = error }
        if let loadError { throw loadError }
        return container
    }

    @Test func createAndFetchTrip() throws {
        let container = try makeTestContainer()
        let context = container.viewContext

        let trip = CDTrip(context: context)
        trip.name = "Test Trip"
        trip.startDate = .now
        try context.save()

        let request: NSFetchRequest<CDTrip> = CDTrip.fetchRequest()
        let trips = try context.fetch(request)
        #expect(trips.count == 1)
        #expect(trips.first?.name == "Test Trip")
    }
}

Tips:

  • Share the NSManagedObjectModel instance across tests to avoid "duplicate entity" warnings.
  • Use a single shared model loaded once:
swift
private let sharedModel: NSManagedObjectModel = {
    let url = Bundle.main.url(forResource: "MyAppModel", withExtension: "momd")!
    return NSManagedObjectModel(contentsOf: url)!
}()

func makeTestContainer() throws -> NSPersistentContainer {
    let container = NSPersistentContainer(name: "MyAppModel",
                                          managedObjectModel: sharedModel)
    // ... configure in-memory store
}

Common Mistakes

MistakeFix
Passing NSManagedObject across threadsPass objectID and re-fetch in the target context
Forgetting to merge batch operation resultsCall mergeChanges(fromRemoteContextSave:into:)
Calling save() without checking hasChangesGuard with context.hasChanges first
Using deprecated init(concurrencyType:) confinement typeUse .privateQueueConcurrencyType or .mainQueueConcurrencyType
Not setting mergePolicy on viewContextSet NSMergeByPropertyObjectTrumpMergePolicy to avoid conflict crashes
Modifying fetch request on live NSFetchedResultsController without deleting cacheCall deleteCache(withName:) first or use cacheName: nil
Batch delete ignoring Deny delete ruleBatch delete bypasses delete rules; validate manually
Marking NSManagedObject as @unchecked SendableDo not. Pass objectID instead
Pointing SwiftData at a fresh store during coexistenceUse the existing store URL and compatible schema when SwiftData should share or migrate Core Data data

Review Checklist

  • NSPersistentContainer is initialized once and shared
  • viewContext used only on main queue; background contexts for writes
  • perform(_:) or performAndWait(_:) wraps all off-queue context access
  • automaticallyMergesChangesFromParent set on viewContext
  • mergePolicy set on viewContext to prevent conflict crashes
  • Batch operation results merged into relevant contexts
  • NSFetchedResultsController fetch requests have sort descriptors
  • Persistent history tracking enabled for multi-target apps
  • Core Data + SwiftData handoff preserves store URL, schema compatibility, entity/property names, and rename mappings
  • Tests use in-memory stores with shared NSManagedObjectModel
  • No NSManagedObject instances cross thread boundaries

References

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Core Data AI skill do?

Build, review, or improve Core Data persistence in apps that have not adopted SwiftData. Use when working with NSManagedObject subclasses, NSFetchedResultsController for list-driven UI, NSBatchInsertRequest / NSBatchDeleteRequest / NSBatchUpdateRequest for bulk operations, NSPersistentHistoryChangeRequest for persistent history tracking and multi-target sync, NSStagedMigrationManager for staged schema migrations (iOS 17+), NSCompositeAttributeDescription for composite attributes (iOS 17+), or when integrating Core Data threading with Swift Concurrency. For Core Data + SwiftData coexistence...

Why use Core Data on TypingMind?

Because you install it once and use it with any model. Core Data is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Core Data in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/dpearson2699/swift-ios-skills/tree/main/skills/core-data. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Core Data?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Core Data?

As many as you like. As long as a model supports skills, you can use Core Data with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Core Data AI skill free?

It is published on GitHub by dpearson2699. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇