What problem does it solve? AI-generated SwiftUI document apps frequently contain subtle but serious defects: phantom APIs like @FocusedDocument, classes conforming to the value-type FileDocument protocol, missing snapshot(contentType:) methods, undeclared UTTypes, and serialization pinned to the main actor. This Skill systematically detects these document-model defects in macOS SwiftUI projects before they cause data loss or build failures. ## Core Features & Use Cases - 12-Defect Detection Index: Locates document architecture problems from phantom property wrappers (doc-01) to missing UndoManager wiring (doc-12), each with severity ratings and grounded fixes. - Hybrid Lint Engine: Combines tier-1 grep tells with tier-2 ast-grep structural rules (missing snapshot detection, @MainActor-pinned documents) that plain text search cannot express, emitting unified JSON and SARIF output. - Evidence-Grounded Fixes: Every finding cites a consensus code shape from 1,857 real shipping macOS apps plus Apple documentation fetched via Sosumi, never hand-invented snippets. - Use Case: Run this on an AI-scaffolded document-based Mac app to catch a class conforming to FileDocument, a ReferenceFileDocument missing its snapshot method, or DocumentGroupLaunchScene used on a platform where it has no macOS arm. ## Quick Start Audit my macOS SwiftUI document-based app for document model correctness issues and report any findings with fixes.
