Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Open or report the Spec Kitty dashboard. Use for dashboard URL, localhost daemon metadata in .kittify/.dashboard, --open, --kill, or status views.
.claude/skills/priivacy-ai-spk-admin-dashboard/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -5% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -52% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -47% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -46% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -66% | 0% |
Open the dashboard directly or show the user its URL. Do not treat the dashboard as a generic status abstraction.
spec-kitty dashboard --open so lifecycle validation checks project,token, health, and daemon state before showing a URL.
.kittify/.dashboard exists, treat it as diagnostic metadata only; donot show the URL until validation succeeds.
spec-kitty dashboard --json only for machine-readable mission rows.spec-kitty dashboard --kill only when the user asks to stop it./spec-kitty.status in a text TUI, loadspk-doctrine-show-me and follow its status rendering contract.
Read references/dashboard-daemon.md before diagnosing dashboard metadata, ports, tokens, or stale PID behavior.
The dashboard is a localhost daemon, not a cloud page. Prefer spec-kitty dashboard --open over manually trusting recorded metadata.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-12 | pass→pass | 4,957 | 1,919 | -61% | 1 | 1 | 0% | 646 | 478 | -26% | 0 | 0 | — |
case-01 | fail→fail | 6,052 | 3,020 | -50% | 1 | 1 | 0% | 951 | 565 | -41% | 0 | 0 | — |
case-02 | fail→pass | 5,670 | 2,452 | -57% | 1 | 1 | 0% | 550 | 524 | -5% | 0 | 0 | — |
case-03 | fail→pass | 8,523 | 1,762 | -79% | 1 | 1 | 0% | 870 | 414 | -52% | 0 | 0 | — |
case-04 | pass→pass | 6,694 | 2,317 | -65% | 1 | 1 | 0% | 1,013 | 599 | -41% | 0 | 0 | — |
case-05 | fail→pass | 9,838 | 2,807 | -71% | 1 | 1 | 0% | 1,264 | 674 | -47% | 0 | 0 | — |
case-06 | fail→pass | 8,305 | 2,974 | -64% | 1 | 1 | 0% | 1,090 | 591 | -46% | 0 | 0 | — |
case-07 | fail→pass | 7,860 | 1,456 | -81% | 1 | 1 | 0% | 1,181 | 404 | -66% | 0 | 0 | — |
case-08 | fail→pass | 8,672 | 4,017 | -54% | 1 | 1 | 0% | 1,374 | 644 | -53% | 0 | 0 | — |
case-09 | fail→pass | 13,609 | 3,477 | -74% | 1 | 1 | 0% | 2,066 | 710 | -66% | 0 | 0 | — |
case-10 | pass→pass | 11,396 | 3,439 | -70% | 1 | 1 | 0% | 1,496 | 668 | -55% | 0 | 0 | — |
case-11 | fail→pass | 8,172 | 2,514 | -69% | 1 | 1 | 0% | 1,359 | 499 | -63% | 0 | 0 | — |
case-13 | fail→pass | 14,020 | 2,535 | -82% | 1 | 1 | 0% | 2,063 | 510 | -75% | 0 | 0 | — |
case-14 | fail→pass | 10,576 | 3,133 | -70% | 1 | 1 | 0% | 1,672 | 627 | -63% | 0 | 0 | — |
case-15 | fail→pass | 13,025 | 4,408 | -66% | 1 | 1 | 0% | 1,547 | 838 | -46% | 0 | 0 | — |
case-16 | fail→pass | 12,981 | 2,996 | -77% | 1 | 1 | 0% | 1,916 | 682 | -64% | 0 | 0 | — |
case-17 | fail→pass | 11,689 | 5,719 | -51% | 1 | 1 | 0% | 1,777 | 975 | -45% | 0 | 0 | — |
case-18 | fail→pass | 12,544 | 1,711 | -86% | 1 | 1 | 0% | 1,724 | 490 | -72% | 0 | 0 | — |
case-19 | pass→pass | 5,324 | 2,735 | -49% | 1 | 1 | 0% | 761 | 630 | -17% | 0 | 0 | — |
case-20 | fail→fail | 11,883 | 7,066 | -41% | 1 | 1 | 0% | 1,980 | 1,169 | -41% | 0 | 0 | — |
case-21 | pass→pass | 5,206 | 3,438 | -34% | 1 | 1 | 0% | 714 | 608 | -15% | 0 | 0 | — |
case-22 | pass→pass | 4,841 | 3,102 | -36% | 1 | 1 | 0% | 742 | 701 | -6% | 0 | 0 | — |
case-23 | fail→pass | 6,129 | 3,069 | -50% | 1 | 1 | 0% | 767 | 604 | -21% | 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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/13/2026 | +73% |
Other measured skills in the registry, with their headline benchmark lift.