agentsclimarketplace
Sign in

Macos swift review

Skill yarlson/yarstack/plugins/yarstack/skills/macos-swift-review

Supplement code review with macOS, Swift, and SwiftUI-specific correctness analysis. Use alongside code-review whenever the requested scope contains meaningful Swift production or test code, a Swift package, an Xcode project, or a macOS Swift app; skip only when no meaningful Swift code is in scope.From its SKILL.md

Install
npx -y skills add yarlson/yarstack --skill macos-swift-review

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.0 KB, 328 tokens by cl100k_base, as published. Nobody here has run it

macOS Swift Review

Run the required Swift and macOS-specific pass alongside general review and find platform failures it is unlikely to catch. Keep the review read-only.

Workflow

  1. Confirm whether the scope is a diff, flow, module, or full app and identify the actual distribution model.
  2. Trace SwiftUI state ownership, identity, navigation, presentation, windows, restoration, and model lifetime through affected flows.
  3. Check actor isolation, @MainActor, task ownership, cancellation, async streams, and blocking main-thread work.
  4. Check platform-specific persistence, atomic file access, security-scoped resources, sandbox containers, and corrupted-state recovery where used.
  5. Check entitlements, hardened runtime, keychain, helpers, XPC or IPC, Apple Events, URL schemes, permissions, signing, updates, and notarization only where the selected scope depends on them.
  6. Distinguish source evidence from checks requiring identities, entitlements, release artifacts, or a live macOS environment. Do not build, sign, notarize, access the network, or control the UI without separate authority.

Use code-review for generic concerns, security-review for exploit analysis, ui-control for direct UI evidence, test-gap-review for coverage, and rollout-readiness-review for release readiness. Follow-ups remain read-only unless separately authorized.

Report severity, affected location, realistic platform scenario, impact, smallest safe correction, verification, and confidence limits. State directly when no platform-specific finding exists.

Finish when the selected Apple-platform surface is exhausted or a named environment limitation blocks further evidence.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.