Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Configure Spec Kitty agent profiles, host-specific paths, command installation, and model/tool routing.
.claude/skills/priivacy-ai-spk-admin-agent-config/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -24% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -56% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -61% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -62% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -48% | 0% |
Use this skill when the user needs to configure Codex, Claude, OpenCode, or another agent for Spec Kitty.
spec-kitty next --agent <name>.spk-doctrine-profile-load when needed.spk-start-agent-surface for host capability differences.Keep host configuration separate from mission content.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 10,782 | 7,894 | -27% | 1 | 1 | 0% | 2,169 | 1,659 | -24% | 0 | 0 | — |
case-02 | fail→fail | 12,141 | 5,817 | -52% | 1 | 1 | 0% | 2,397 | 1,273 | -47% | 0 | 0 | — |
case-03 | fail→pass | 14,764 | 6,687 | -55% | 1 | 1 | 0% | 3,028 | 1,324 | -56% | 0 | 0 | — |
case-04 | fail→pass | 11,604 | 3,661 | -68% | 1 | 1 | 0% | 2,147 | 828 | -61% | 0 | 0 | — |
case-05 | fail→pass | 13,026 | 3,809 | -71% | 1 | 1 | 0% | 2,311 | 881 | -62% | 0 | 0 | — |
case-06 | fail→pass | 10,124 | 3,929 | -61% | 1 | 1 | 0% | 1,729 | 901 | -48% | 0 | 0 | — |
case-07 | fail→pass | 13,759 | 3,916 | -72% | 1 | 1 | 0% | 2,356 | 849 | -64% | 0 | 0 | — |
case-08 | pass→pass | 9,697 | 4,348 | -55% | 1 | 1 | 0% | 1,723 | 857 | -50% | 0 | 0 | — |
case-09 | fail→fail | 9,735 | 3,868 | -60% | 1 | 1 | 0% | 1,739 | 740 | -57% | 0 | 0 | — |
case-10 | fail→pass | 10,205 | 2,625 | -74% | 1 | 1 | 0% | 1,777 | 632 | -64% | 0 | 0 | — |
case-11 | fail→pass | 6,959 | 1,611 | -77% | 1 | 1 | 0% | 1,212 | 338 | -72% | 0 | 0 | — |
case-12 | fail→pass | 10,957 | 4,611 | -58% | 1 | 1 | 0% | 2,020 | 1,010 | -50% | 0 | 0 | — |
case-13 | fail→pass | 6,063 | 3,143 | -48% | 1 | 1 | 0% | 1,082 | 617 | -43% | 0 | 0 | — |
case-14 | pass→pass | 8,883 | 6,361 | -28% | 1 | 1 | 0% | 1,601 | 1,194 | -25% | 0 | 0 | — |
case-15 | fail→pass | 10,614 | 4,798 | -55% | 1 | 1 | 0% | 1,950 | 1,068 | -45% | 0 | 0 | — |
case-16 | fail→pass | 12,314 | 5,443 | -56% | 1 | 1 | 0% | 2,163 | 431 | -80% | 0 | 0 | — |
case-17 | fail→pass | 6,543 | 1,560 | -76% | 1 | 1 | 0% | 1,098 | 407 | -63% | 0 | 0 | — |
case-18 | fail→pass | 7,253 | 1,932 | -73% | 1 | 1 | 0% | 1,383 | 474 | -66% | 0 | 0 | — |
case-19 | fail→fail | 9,647 | 4,097 | -58% | 1 | 1 | 0% | 1,680 | 833 | -50% | 0 | 0 | — |
case-20 | pass→pass | 14,916 | 8,615 | -42% | 1 | 1 | 0% | 3,080 | 1,892 | -39% | 0 | 0 | — |
case-21 | pass→pass | 12,339 | 7,917 | -36% | 1 | 1 | 0% | 2,024 | 1,517 | -25% | 0 | 0 | — |
case-22 | pass→pass | 5,867 | 4,033 | -31% | 1 | 1 | 0% | 1,179 | 925 | -22% | 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 +64 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.