Implement or review SwiftData models, containers, queries, relationships, migrations, CloudKit configuration, and background persistence work. Use for SwiftData-backed storage; route legacy Core Data-only work to core-data.
90%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Passed
No findings from the security scan
Design persistence so model identity, relationships, query cost, migration, and context ownership remain explicit.
This skill owns @Model, ModelContainer, ModelContext, @Query, predicates, fetch descriptors, schema migration, @ModelActor, and SwiftData/CloudKit constraints. Route pure Core Data work to core-data, UI state wiring to swiftui-patterns, and general actor diagnostics to swift-concurrency.
Inspect deployment target, Xcode/Swift toolchain, enabled capabilities, and the existing schema before recommending APIs. Preserve the current store and migration path unless the user explicitly authorizes a reset. Verify availability and CloudKit constraints against SDK headers and primary Apple documentation.
Read custom stores, history, and CloudKit, performance, predicates, and modeling, or containers, undo, previews, and errors for advanced workflows. Read Indexing before adding indexes or compound uniqueness.
Install one deliberate root container for an app/scene unless isolation or testing requires another. Inject it into SwiftUI and use in-memory or temporary configurations for previews and tests.
ModelContext is not a general-purpose cross-actor value. Keep UI writes on the UI-owned context. Use @ModelActor or another context created for background persistence work, and pass stable identifiers or PersistentIdentifier values across isolation boundaries rather than live model objects.
Save behavior must be explicit enough for the feature: decide whether a user action saves immediately, a transaction batches changes, or autosave is sufficient. Do not silently swallow persistence errors in production paths where data loss matters.
Use @Query when a SwiftUI view owns a live query whose predicate and sort shape are suitable for property-wrapper construction. Use FetchDescriptor for imperative, reusable, paginated, or actor-owned fetches.
Push filtering and sorting into the store when practical. In-memory filtering is acceptable for a proven small dataset or UI-only transformation, not as a default workaround for predicate limitations.
Read Queries for @Query, FetchDescriptor, sorting, pagination, and dynamic query patterns. Read Predicate Pitfalls before debugging optional relationships, captured values, enums, or unsupported expressions.
Treat schema evolution as a product-data change. Test lightweight and custom stages from the actual previous schema, including failure and rollback/recovery expectations. Never delete the user's store merely to make a migration pass unless the user has explicitly accepted data loss.
Read Core Data Coexistence when both frameworks share a product or when planning an incremental migration.
@Model objects across actor isolation boundaries instead of PersistentIdentifier.ModelContext concurrently from background tasks without @ModelActor.@Query fetches without predicate or fetch limit on large tables.thienqo/ios-skills
6bf8b62
Also appears in
since Sep 6, 2026
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.