What problem does it solve? Auditing a macOS SwiftUI codebase by hand means checking dozens of domains — API currency, concurrency, SwiftData, Liquid Glass, accessibility, AppKit interop — and reconciling overlapping findings. This Skill automates that by scanning the code, dispatching only the relevant domain audit skills as parallel subagents, deduplicating collisions, and rolling everything into one summary dashboard with a nativeness score. ## Core Features & Use Cases - Scan-steered orchestration: Runs a presence scan to select only the relevant audit skills, then dispatches them in dependency-ordered waves of parallel subagents. - Evidence-backed findings: Every finding passes a LOCATE (lint) → VERIFY (read code plus Apple docs and shipping-app corpus) → FIX pipeline with a strict fix-safety protocol. - Consolidated dashboard: Produces a single _SUMMARY.md with severity rollups, per-skill results, a 0-100 macOS nativeness score, and a deduplicated master finding table. - Use Case: Before shipping a macOS SwiftUI app, run the full audit to get a deterministic report of hard failures, warnings, and advisories across all 28 domains, then apply guarded fixes one commit per finding. ## Quick Start Run a full macOS SwiftUI audit on this project and generate the consolidated findings dashboard with a nativeness score.
