Menu
Run the full evidence-backed macOS SwiftUI audit suite over a project. Use only when explicitly requested with $swiftui-audit.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Audit macOS SwiftUI accessibility for correctness, current APIs, and production conventions. Use for that domain or as part of a full audit.
Audit macOS SwiftUI animation motion for correctness, current APIs, and production conventions. Use for that domain or as part of a full audit.
Audit macOS SwiftUI api currency for correctness, current APIs, and production conventions. Use for that domain or as part of a full audit.
Audit macOS SwiftUI appearance color for correctness, current APIs, and production conventions. Use for that domain or as part of a full audit.
Audit macOS SwiftUI appkit interop for correctness, current APIs, and production conventions. Use for that domain or as part of a full audit.
Audit macOS SwiftUI appkit overuse for correctness, current APIs, and production conventions. Use for that domain or as part of a full audit.
Based on SOC occupation classification
| name | swiftui-audit |
| description | Run the full evidence-backed macOS SwiftUI audit suite over a project. Use only when explicitly requested with $swiftui-audit. |
Let <swiftui-plugin-root> be the absolute plugin directory two levels above this SKILL.md. Resolve that path before running commands and shared references.
Resolve the requested target, defaulting to the current directory. Create the output directory and persist the relevance scan:
mkdir -p swiftui-audits
python3 "<swiftui-plugin-root>/scripts/audit-scan.py" "<target>" --json swiftui-audits/_scan.json
Read <swiftui-plugin-root>/skills/audit-macos-swiftui-full/SKILL.md completely. Follow that orchestrator using _scan.json and preserve its dependency waves: guards; state and data; UI domains; boundaries and scoring.
When collaboration tools are available, run independent auditors in the same wave concurrently. Otherwise run them serially in the listed order.
Write per-finding Markdown under swiftui-audits/ using <swiftui-plugin-root>/references/_shared/finding-schema.md, then create _SUMMARY.md and its 0–100 nativeness score.
Read .swiftui-plugin/settings.md, falling back to .claude/swiftui.local.md. With strict_audit: true or no setting, label hard findings release-blocking. With strict_audit: false, label the same findings advisory. The mechanical audit-gate.sh accepts matching --strict and --advisory modes for CI exit behavior.
Apply fixes only when the user requested them and the shared fix-safety protocol permits them. Keep all other findings open.
The lint engine locates candidates; every reported finding still requires source inspection and verification through the bundled corpus and current Apple documentation.