Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Stop the .NET Aspire AppHost and its Docker containers via the developer CLI. Defaults to the current worktree; supports stopping all worktrees or a specific base port.
.claude/skills/platformplatform-aspire-stop/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -69% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -74% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -53% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -51% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -57% | 0% |
bashdotnet run --project developer-cli -- stop [--all] [--port <basePort>]
Use developer-cli exactly as written - do not expand to an absolute worktree path.
Stops the Aspire AppHost and its associated Docker containers (postgres, azurite, mailpit, stripe-cli) - not just the Aspire process. Persistent containers survive a plain aspire stop by design, so this skill cleans them up too.
--all - stop every worktree of this repository (via git worktree list). Only use when the user explicitly asks to stop everything - never as a default--port <basePort> - stop the worktree on that base port (e.g. 9000, 9100, 9200); also cleans up Docker containers for that port even if the worktree was deletedThe flags are mutually exclusive.
aspire-restart is the everyday default).--port).The CLI prints what it stopped (Aspire process trees + each docker rm --force'd container) and exits.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 9,616 | 3,052 | -68% | 1 | 1 | 0% | 1,622 | 499 | -69% | 0 | 0 | — |
case-16 | fail→pass | 13,652 | 1,616 | -88% | 1 | 1 | 0% | 2,172 | 569 | -74% | 0 | 0 | — |
case-02 | fail→pass | 6,344 | 3,245 | -49% | 1 | 1 | 0% | 1,078 | 506 | -53% | 0 | 0 | — |
case-03 | fail→pass | 6,544 | 1,610 | -75% | 1 | 1 | 0% | 1,030 | 502 | -51% | 0 | 0 | — |
case-04 | fail→pass | 7,761 | 1,863 | -76% | 1 | 1 | 0% | 1,311 | 561 | -57% | 0 | 0 | — |
case-05 | fail→pass | 9,172 | 1,791 | -80% | 1 | 1 | 0% | 1,564 | 560 | -64% | 0 | 0 | — |
case-06 | fail→pass | 11,241 | 2,116 | -81% | 1 | 1 | 0% | 1,941 | 657 | -66% | 0 | 0 | — |
case-07 | fail→pass | 12,014 | 2,600 | -78% | 1 | 1 | 0% | 1,840 | 645 | -65% | 0 | 0 | — |
case-08 | pass→pass | 10,111 | 1,841 | -82% | 1 | 1 | 0% | 1,501 | 580 | -61% | 0 | 0 | — |
case-09 | fail→pass | 8,918 | 1,882 | -79% | 1 | 1 | 0% | 1,483 | 561 | -62% | 0 | 0 | — |
case-10 | pass→pass | 10,632 | 2,508 | -76% | 1 | 1 | 0% | 1,708 | 768 | -55% | 0 | 0 | — |
case-11 | fail→pass | 10,790 | 2,137 | -80% | 1 | 1 | 0% | 1,801 | 646 | -64% | 0 | 0 | — |
case-12 | fail→pass | 8,035 | 2,562 | -68% | 1 | 1 | 0% | 1,339 | 643 | -52% | 0 | 0 | — |
case-13 | fail→pass | 8,350 | 2,518 | -70% | 1 | 1 | 0% | 1,435 | 686 | -52% | 0 | 0 | — |
case-14 | pass→pass | 11,047 | 1,725 | -84% | 1 | 1 | 0% | 1,783 | 558 | -69% | 0 | 0 | — |
case-15 | fail→pass | 9,603 | 2,169 | -77% | 1 | 1 | 0% | 1,595 | 648 | -59% | 0 | 0 | — |
case-17 | fail→pass | 14,740 | 3,640 | -75% | 1 | 1 | 0% | 2,351 | 728 | -69% | 0 | 0 | — |
case-18 | pass→pass | 13,660 | 1,988 | -85% | 1 | 1 | 0% | 2,149 | 549 | -74% | 0 | 0 | — |
case-19 | fail→pass | 9,999 | 1,984 | -80% | 1 | 1 | 0% | 1,763 | 585 | -67% | 0 | 0 | — |
case-20 | pass→pass | 10,849 | 4,914 | -55% | 1 | 1 | 0% | 1,755 | 1,154 | -34% | 0 | 0 | — |
case-21 | pass→pass | 5,795 | 3,205 | -45% | 1 | 1 | 0% | 1,016 | 775 | -24% | 0 | 0 | — |
case-22 | pass→pass | 7,216 | 5,930 | -18% | 1 | 1 | 0% | 1,228 | 1,296 | +6% | 0 | 0 | — |
case-23 | pass→pass | 5,809 | 2,446 | -58% | 1 | 1 | 0% | 1,086 | 617 | -43% | 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. 23 cases were attempted. The headline lift of +65 percentage points is the difference between those two pass rates over the 23 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.