Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →expo-ui-swift-ui
.claude/skills/expo-ui-swift-ui/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -21% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -54% | 0% |
> The instructions in this skill apply to SDK 55 only. For other SDK versions, refer to the Expo UI SwiftUI docs for that version for the most accurate information.
@expo/ui/swift-ui.Host, or embedding React Native components with RNHostView.bashnpx expo install @expo/ui
A native rebuild is required after installation (npx expo run:ios).
@expo/ui/swift-ui, modifiers from @expo/ui/swift-ui/modifiers.Host.RNHostView is specifically for embedding RN components inside a SwiftUI tree. Example:jsximport { Host, VStack, RNHostView } from "@expo-ui/swift-ui"; import { Pressable } from "react-native"; <Host matchContents> <VStack> <RNHostView matchContents> // Here, `Pressable` is an RN component so it is wrapped in `RNHostView`. <Pressable /> </RNHostView> </VStack> </Host>;
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 8,126 | 5,716 | -30% | 1 | 1 | 0% | 1,665 | 1,844 | +11% | 0 | 0 | — |
case-02 | fail→pass | 12,435 | 7,614 | -39% | 1 | 1 | 0% | 2,647 | 2,195 | -17% | 0 | 0 | — |
case-03 | fail→pass | 14,852 | 7,250 | -51% | 1 | 1 | 0% | 2,880 | 2,123 | -26% | 0 | 0 | — |
case-04 | fail→pass | 8,352 | 3,518 | -58% | 1 | 1 | 0% | 1,458 | 1,154 | -21% | 0 | 0 | — |
case-05 | fail→pass | 12,185 | 1,932 | -84% | 1 | 1 | 0% | 1,961 | 893 | -54% | 0 | 0 | — |
case-06 | fail→fail | 7,708 | 2,421 | -69% | 1 | 1 | 0% | 1,224 | 895 | -27% | 0 | 0 | — |
case-07 | fail→pass | 14,897 | 3,148 | -79% | 1 | 1 | 0% | 1,986 | 1,189 | -40% | 0 | 0 | — |
case-08 | fail→fail | 6,957 | 2,488 | -64% | 1 | 1 | 0% | 1,356 | 1,013 | -25% | 0 | 0 | — |
case-09 | pass→pass | 4,790 | 1,393 | -71% | 1 | 1 | 0% | 848 | 731 | -14% | 0 | 0 | — |
case-10 | fail→fail | 3,522 | 1,593 | -55% | 1 | 1 | 0% | 553 | 778 | +41% | 0 | 0 | — |
case-11 | fail→pass | 8,936 | 3,958 | -56% | 1 | 1 | 0% | 1,521 | 1,351 | -11% | 0 | 0 | — |
case-12 | fail→pass | 9,945 | 3,203 | -68% | 1 | 1 | 0% | 1,754 | 1,123 | -36% | 0 | 0 | — |
case-13 | fail→pass | 14,901 | 2,666 | -82% | 1 | 1 | 0% | 2,421 | 1,007 | -58% | 0 | 0 | — |
case-14 | fail→fail | 10,088 | 2,613 | -74% | 1 | 1 | 0% | 1,699 | 946 | -44% | 0 | 0 | — |
case-15 | fail→fail | 10,337 | 2,175 | -79% | 1 | 1 | 0% | 1,669 | 950 | -43% | 0 | 0 | — |
case-16 | fail→pass | 8,782 | 3,073 | -65% | 1 | 1 | 0% | 1,599 | 1,114 | -30% | 0 | 0 | — |
case-17 | fail→pass | 15,427 | 5,247 | -66% | 1 | 1 | 0% | 2,485 | 1,697 | -32% | 0 | 0 | — |
case-18 | fail→fail | 16,526 | 6,107 | -63% | 1 | 1 | 0% | 2,643 | 1,723 | -35% | 0 | 0 | — |
case-19 | fail→pass | 13,739 | 6,783 | -51% | 1 | 1 | 0% | 2,770 | 1,941 | -30% | 0 | 0 | — |
case-20 | fail→fail | 11,759 | 2,710 | -77% | 1 | 1 | 0% | 2,262 | 1,017 | -55% | 0 | 0 | — |
case-21 | fail→fail | 13,354 | 6,008 | -55% | 1 | 1 | 0% | 2,761 | 1,579 | -43% | 0 | 0 | — |
case-22 | fail→pass | 7,096 | 6,085 | -14% | 1 | 1 | 0% | 1,653 | 1,864 | +13% | 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 +59 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.