artemnovichkov/xcode-27-system-prompts

What it solves

This repository provides a collection of system prompts and documentation extracted from the Xcode 27 beta 5 AI coding assistant. It reveals how Apple implements AI-assisted coding, including the specific instructions used to guide the AI's behavior, reasoning, and code generation for Apple platforms.

How it works

The project organizes extracted files into several categories:

• System Prompt Templates: Core instructions for different modes, such as basic coding assistance, complex reasoning, and rapid code integration.
• Agent Skills: Specialized, on-demand prompts paired with reference documentation for specific tasks like auditing security settings, adopting C bounds safety, or mastering new SwiftUI and App Intents APIs.
• Context Providers: Templates that define how the AI receives information about the current file, selected code, or project structure.
• Tool-Assisted Prompts: Instructions for AI agents that can use search and editing tools to modify the codebase.
• Additional Documentation: Guides on new iOS 26/27 features, on-device LLMs, and framework updates.

Who it’s for

• AI Prompt Engineers: Those looking to understand professional-grade system prompts for coding assistants.
• Apple Developers: Developers wanting to see the internal guidelines and best practices for iOS, macOS, visionOS, and watchOS development.
• LLM Researchers: People studying how large language models are integrated into professional IDEs.

Highlights

• Internal Apple Prompting Principles: Includes guidelines on favoring Apple-first development, platform-aware suggestions, and modern Swift preferences.
• Specialized Agent Skills: Detailed prompts for high-value tasks like security auditing and UIKit modernization.
• Sytactic Validity: Insights into Apple's "Complete File Replacement" philosophy for code editing.
• Comprehensive Framework Guides: Documentation on Liquid Glass design, Visual Intelligence, and on-device LLM integration.