Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Write correct coroutine scopes, lifecycle collection, and dispatcher injection in Android production code. Use for suspend functions, coroutine scopes, and dispatcher mechanics; defer ViewModel StateFlow/LiveData architecture, Fragment lifecycle recipes, persistence/notifications, and unit-test recipes to their specific skills.
.claude/skills/hoangnguyen0403-android-concurrency/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -67% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -22% | 0% |
viewModelScope (VM) or lifecycleScope (Activity/Fragment).DispatcherProvider) for testability. not hardcode Dispatchers.IO.Flow for data streams.StateFlow (State) or SharedFlow (Events).SharedFlow with replay only when late subscribers must receive recent events.collectAsStateWithLifecycle() (Compose) or repeatOnLifecycle (Views).| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 14,931 | 10,801 | -28% | 1 | 1 | 0% | 3,007 | 2,668 | -11% | 0 | 0 | — |
case-02 | fail→fail | 13,264 | 12,124 | -9% | 1 | 1 | 0% | 2,826 | 3,065 | +8% | 0 | 0 | — |
case-03 | fail→pass | 10,079 | 8,456 | -16% | 1 | 1 | 0% | 1,974 | 1,939 | -2% | 0 | 0 | — |
case-04 | pass→pass | 7,546 | 4,973 | -34% | 1 | 1 | 0% | 1,609 | 1,251 | -22% | 0 | 0 | — |
case-19 | pass→pass | 3,949 | 2,885 | -27% | 1 | 1 | 0% | 577 | 729 | +26% | 0 | 0 | — |
case-05 | pass→pass | 7,600 | 7,183 | -5% | 1 | 1 | 0% | 1,575 | 1,583 | +1% | 0 | 0 | — |
case-06 | pass→pass | 6,316 | 2,819 | -55% | 1 | 1 | 0% | 1,329 | 805 | -39% | 0 | 0 | — |
case-07 | pass→pass | 9,770 | 6,515 | -33% | 1 | 1 | 0% | 1,764 | 1,451 | -18% | 0 | 0 | — |
case-08 | pass→pass | 8,785 | 5,520 | -37% | 1 | 1 | 0% | 1,602 | 1,208 | -25% | 0 | 0 | — |
case-09 | pass→pass | 10,206 | 7,866 | -23% | 1 | 1 | 0% | 2,085 | 1,947 | -7% | 0 | 0 | — |
case-10 | pass→pass | 14,388 | 5,702 | -60% | 1 | 1 | 0% | 2,138 | 1,233 | -42% | 0 | 0 | — |
case-11 | pass→pass | 11,069 | 6,565 | -41% | 1 | 1 | 0% | 1,987 | 1,400 | -30% | 0 | 0 | — |
case-12 | pass→pass | 9,766 | 6,005 | -39% | 1 | 1 | 0% | 1,953 | 1,258 | -36% | 0 | 0 | — |
case-13 | pass→pass | 17,503 | 11,750 | -33% | 1 | 1 | 0% | 3,052 | 2,763 | -9% | 0 | 0 | — |
case-14 | fail→pass | 10,389 | 5,999 | -42% | 1 | 1 | 0% | 2,028 | 1,442 | -29% | 0 | 0 | — |
case-15 | pass→pass | 14,049 | 7,238 | -48% | 1 | 1 | 0% | 2,151 | 1,578 | -27% | 0 | 0 | — |
case-16 | pass→pass | 5,959 | 2,264 | -62% | 1 | 1 | 0% | 912 | 612 | -33% | 0 | 0 | — |
case-17 | fail→pass | 12,757 | 3,114 | -76% | 1 | 1 | 0% | 2,283 | 757 | -67% | 0 | 0 | — |
case-18 | pass→pass | 10,534 | 6,879 | -35% | 1 | 1 | 0% | 1,914 | 1,622 | -15% | 0 | 0 | — |
case-20 | pass→pass | 7,732 | 4,356 | -44% | 1 | 1 | 0% | 1,350 | 953 | -29% | 0 | 0 | — |
case-21 | pass→pass | 5,428 | 4,021 | -26% | 1 | 1 | 0% | 861 | 851 | -1% | 0 | 0 | — |
case-22 | pass→pass | 12,502 | 8,294 | -34% | 1 | 1 | 0% | 2,060 | 1,827 | -11% | 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 +18 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.