Prisma-Labs-Dev/apple-skills

Apple development skills for Claude Code and Codex — latest iOS 26+ APIs, SwiftUI, Liquid Glass, and more

What it solves

Coding agents often fall back to outdated or deprecated Apple APIs because their training data doesn't include the newest frameworks. Apple Skills gives agents a curated set of up-to-date reference docs and workflow guides, so they can generate code that uses iOS 26+ and Swift 6 conventions like NavigationStack, @Observable, and async/await.

How it works

The project is a collection of "skills" that can be installed into coding agents via plugin marketplaces (like Claude Code) or the Skills CLI (for Codex, Cursor). Each skill contains either framework reference documentation (e.g., SwiftUI, UIKit, HealthKit) or expert workflow guides that teach how to approach tasks (e.g., concurrency patterns, Swift Testing best practices). A coordinator skill (ios-dev) acts as a starting point, routing to the appropriate guides and references. Some skills are intentionally disabled and kept out of the agent's load, because they impose subjective design opinions; you can re-enable them by moving them back into the active skills directory.

Who it’s for

Developers and teams using AI coding agents (Claude Code, Codex, Cursor) to build Apple platform apps. It's especially useful for those targeting the latest iOS and macOS APIs and wanting agents to stay current with modern Swift conventions instead of relying on legacy patterns.

Highlights

  • Covers iOS 26+ / Swift 6 APIs: no legacy patterns, uses @Observable, NavigationStack, async/await, Liquid Glass, and Swift Testing.
  • Broad framework coverage: SwiftUI, UIKit, SwiftData, HealthKit, StoreKit, MapKit, WidgetKit, and more.
  • Expert workflow guides: concurrency patterns, Swift Testing patterns, SwiftData practices, Swift Charts, performance auditing, and macOS SPM packaging without Xcode.
  • Design and guidelines skills for HIG and Liquid Glass API reference.
  • Disabled skills are kept in the repo but not loaded, to respect the model's own design judgment.
  • Simple install and update for multiple agents: Claude Code plugin marketplace, or npx skills for Codex and Cursor.

Related

  • Project
    Chrrxs/robloxstudio-mcp

    An MCP server that enables AI coding agents to control and interact with Roblox Studio, allowing for automated debugging, script execution, performance profiling, and asset management.

  • Project
    kelos-dev/kelos

    An orchestration framework that turns coding agents into Kubernetes workloads for isolated, scalable, and event-driven software development automation.

  • Project
    gethasp/hasp

    A local secret broker for coding agents that keeps sensitive credentials in an encrypted vault and prevents them from entering the agent's context.

  • Project
    Klerith/fernando-skills

    A pair of Claude Code skills that enforce a spec-driven workflow: /spec designs a feature spec through clarifying questions, and /spec-impl implements it step by step only after you manually approve the spec. This prevents silent LLM improvisation, makes decisions explicit, and keeps every change reviewable.