build-macos-apps_skill

Overview

This skill builds professional native macOS apps in Swift using SwiftUI and AppKit from the command line, without Xcode. It covers the full lifecycle: create, debug, test, optimize, and ship apps via CLI tooling and reproducible workflows. The approach is test-driven and outcome-focused: every change is verified and reported. The user describes intent; I implement, verify, and report results.

How this skill works

I run a prescriptive CLI workflow that maps your intent to a concrete sequence of steps (scaffold, build, test, profile, sign, release). After each change I verify with build, test, and launch checks and report concrete outcomes (build success, test counts, app launch). When requirements are unclear I ask clarifying questions before making changes. Tests are added for logic and regressions; UI changes are validated by launching the app.

When to use it

• Start a new macOS app project from the command line without opening an IDE
• Diagnose and fix crashes, build failures, or failing tests in an existing app
• Add a feature or change behavior with incremental, verified steps
• Write tests or run test suites to prove correctness and prevent regressions
• Profile and optimize CPU, memory, or responsiveness issues
• Prepare release artifacts: code signing, notarization, and packaging

Best practices

• Follow small-step workflow: change → verify → report → next change
• Write tests for logic, state changes, bug fixes, and refactors; skip for purely visual exploration
• Always verify builds and tests before committing or moving to the next task
• Ask clarifying questions when requirements are ambiguous or multiple approaches exist
• Leave the repository in a working state at each stopping point (build and tests passing)

Example use cases

• Create a new SwiftUI macOS app scaffolded for documents and persistence, ready to iterate
• Fix a crashing view controller by reproducing the crash in tests and applying a minimal verified fix
• Add a feature such as a menu bar helper or document export with tests and launch verification
• Optimize a slow screen by profiling, fixing hotspots, and measuring before/after improvements
• Package an app for distribution: sign, notarize, and produce release-ready installer artifacts