Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Expert guidance for Swift Testing: test structure, #expect/#require macros, traits and tags, parameterized tests, test plans, parallel execution, async waiting patterns, and XCTest migration. Use when writing new Swift tests, modernizing XCTest suites, debugging flaky tests, or improving test quality and maintainability in Apple-platform or Swift server projects.
.claude/skills/avdlee-swift-testing-expert/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 14% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 4% | 0% |
| case-14 | ✓→✗ | ▼ Worse | 21% | 0% |
| case-04 | ✓→✓ | = Same ✓ | -27% | 0% |
Use this skill to write, review, migrate, and debug Swift tests with modern Swift Testing APIs. Prioritize readable tests, robust parallel execution, clear diagnostics, and incremental migration from XCTest where needed.
XCUIApplication), performance metrics (XCTMetric), and Objective-C-only test code.#expect as the default assertion and use #require when subsequent lines depend on a prerequisite value..serialized..enabled, .disabled, .timeLimit, .bug, tags) over naming conventions or ad-hoc comments.@available on test functions for OS-gated behavior instead of runtime #available checks inside test bodies; never annotate suite types with @available.Testing in test targets, never in app/library/binary targets.references/fundamentals.md#expect, #require, and throw expectations -> references/expectations.mdreferences/traits-and-tags.mdreferences/parameterized-testing.md.serialized, isolation strategy -> references/parallelization-and-isolation.mdreferences/performance-and-best-practices.mdreferences/async-testing-and-waiting.mdreferences/migration-from-xctest.mdreferences/xcode-workflows.mdreferences/_index.mdtestFooCaseA/testFooCaseB/... methods -> replace with one parameterized @Test(arguments:).try #require(...) then assert on unwrapped value..serialized only as a transition step.withKnownIssue for temporary known failures to preserve signal.CustomTestStringConvertible for focused test diagnostics.#require where failure should stop the test.references/_index.mdreferences/fundamentals.mdreferences/expectations.mdreferences/traits-and-tags.mdreferences/parameterized-testing.mdreferences/parallelization-and-isolation.mdreferences/performance-and-best-practices.mdreferences/async-testing-and-waiting.mdreferences/migration-from-xctest.mdreferences/xcode-workflows.md| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | pass→pass | 19,249 | 7,626 | -60% | 1 | 1 | 0% | 2,975 | 2,169 | -27% | 0 | 0 | — |
case-01 | fail→pass | 14,527 | 12,058 | -17% | 1 | 1 | 0% | 2,220 | 2,867 | +29% | 0 | 0 | — |
case-02 | pass→pass | 12,452 | 6,565 | -47% | 1 | 1 | 0% | 2,062 | 2,132 | +3% | 0 | 0 | — |
case-03 | pass→pass | 14,024 | 9,163 | -35% | 1 | 1 | 0% | 2,416 | 2,527 | +5% | 0 | 0 | — |
case-05 | pass→pass | 18,192 | 9,853 | -46% | 1 | 1 | 0% | 2,296 | 2,575 | +12% | 0 | 0 | — |
case-06 | pass→pass | 10,280 | 6,582 | -36% | 1 | 1 | 0% | 1,742 | 2,107 | +21% | 0 | 0 | — |
case-07 | pass→pass | 11,625 | 6,852 | -41% | 1 | 1 | 0% | 1,793 | 2,043 | +14% | 0 | 0 | — |
case-08 | pass→pass | 10,445 | 6,499 | -38% | 1 | 1 | 0% | 1,646 | 2,078 | +26% | 0 | 0 | — |
case-09 | fail→pass | 12,217 | 6,994 | -43% | 1 | 1 | 0% | 1,811 | 2,069 | +14% | 0 | 0 | — |
case-10 | pass→pass | 12,309 | 4,417 | -64% | 1 | 1 | 0% | 1,750 | 1,661 | -5% | 0 | 0 | — |
case-11 | pass→pass | 13,955 | 9,534 | -32% | 1 | 1 | 0% | 2,157 | 2,443 | +13% | 0 | 0 | — |
case-12 | fail→pass | 15,597 | 9,571 | -39% | 1 | 1 | 0% | 2,473 | 2,568 | +4% | 0 | 0 | — |
case-13 | pass→pass | 13,354 | 6,868 | -49% | 1 | 1 | 0% | 2,210 | 2,075 | -6% | 0 | 0 | — |
case-14 | pass→fail | 17,987 | 14,689 | -18% | 1 | 1 | 0% | 2,807 | 3,390 | +21% | 0 | 0 | — |
case-15 | pass→pass | 14,847 | 9,811 | -34% | 1 | 1 | 0% | 2,305 | 2,550 | +11% | 0 | 0 | — |
case-16 | pass→pass | 7,469 | 5,494 | -26% | 1 | 1 | 0% | 1,230 | 1,855 | +51% | 0 | 0 | — |
case-17 | pass→pass | 13,902 | 7,806 | -44% | 1 | 1 | 0% | 2,007 | 2,164 | +8% | 0 | 0 | — |
case-18 | pass→pass | 8,506 | 6,433 | -24% | 1 | 1 | 0% | 1,468 | 2,052 | +40% | 0 | 0 | — |
case-19 | pass→pass | 10,852 | 4,495 | -59% | 1 | 1 | 0% | 1,713 | 1,727 | +1% | 0 | 0 | — |
case-20 | pass→pass | 8,863 | 5,330 | -40% | 1 | 1 | 0% | 1,507 | 1,944 | +29% | 0 | 0 | — |
case-21 | pass→pass | 12,300 | 9,464 | -23% | 1 | 1 | 0% | 1,781 | 2,561 | +44% | 0 | 0 | — |
case-22 | pass→pass | 14,952 | 19,403 | +30% | 1 | 1 | 0% | 2,526 | 4,178 | +65% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +9 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.