Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Start Ouroboros onboarding. Use when the user sends bare `ooo`, asks to start Ouroboros, or is using Ouroboros for the first time.
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-13 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 65% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 203% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 67% | 0% |
Read ../welcome/SKILL.md and follow its instructions exactly. This is the explicit Marketplace-plugin entry point for a bare ooo message before any runtime-specific setup has run.
Your final response MUST end with exactly one breadcrumb footer line:
◆ <current state> → next: <recommended action>Derive <current state> from live session state via ouroboros_session_status when that MCP projection is available; otherwise derive it from this skill's actual outcome. Never use a linear Step N of M footer because Ouroboros is an evolutionary loop. The breadcrumb line must be the last line of the response.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-12 | fail→fail | 5,169 | 6,325 | +22% | 1 | 1 | 0% | 850 | 570 | -33% | 0 | 0 | — |
case-13 | fail→pass | 10,623 | 10,550 | -1% | 1 | 1 | 0% | 1,786 | 1,587 | -11% | 0 | 0 | — |
case-01 | fail→pass | 16,919 | 18,214 | +8% | 1 | 1 | 0% | 1,426 | 2,347 | +65% | 0 | 0 | — |
case-02 | fail→fail | 5,402 | 6,151 | +14% | 1 | 1 | 0% | 1,007 | 541 | -46% | 0 | 0 | — |
case-03 | fail→fail | 14,932 | 6,038 | -60% | 1 | 1 | 0% | 639 | 480 | -25% | 0 | 0 | — |
case-04 | pass→fail | 5,189 | 13,313 | +157% | 1 | 1 | 0% | 971 | 2,270 | +134% | 0 | 0 | — |
case-05 | pass→pass | 4,360 | 7,222 | +66% | 1 | 1 | 0% | 850 | 1,515 | +78% | 0 | 0 | — |
case-06 | pass→fail | 5,467 | 19,996 | +266% | 1 | 1 | 0% | 995 | 3,264 | +228% | 0 | 0 | — |
case-07 | fail→fail | 30,232 | 4,192 | -86% | 1 | 1 | 0% | 2,781 | 399 | -86% | 0 | 0 | — |
case-08 | fail→pass | 17,310 | 20,681 | +19% | 1 | 1 | 0% | 2,971 | 3,058 | +3% | 0 | 0 | — |
case-09 | fail→pass | 4,461 | 13,495 | +203% | 1 | 1 | 0% | 650 | 1,967 | +203% | 0 | 0 | — |
case-10 | fail→pass | 7,844 | 11,293 | +44% | 1 | 1 | 0% | 1,009 | 1,687 | +67% | 0 | 0 | — |
case-11 | fail→fail | 12,989 | 3,903 | -70% | 1 | 1 | 0% | 2,053 | 409 | -80% | 0 | 0 | — |
case-14 | fail→fail | 16,495 | 6,668 | -60% | 1 | 1 | 0% | 2,470 | 436 | -82% | 0 | 0 | — |
case-15 | fail→pass | 8,792 | 13,240 | +51% | 1 | 1 | 0% | 1,442 | 2,005 | +39% | 0 | 0 | — |
case-16 | fail→pass | 10,961 | 9,813 | -10% | 1 | 1 | 0% | 1,838 | 1,878 | +2% | 0 | 0 | — |
case-17 | fail→pass | 14,122 | 13,854 | -2% | 1 | 1 | 0% | 2,093 | 2,061 | -2% | 0 | 0 | — |
case-18 | fail→fail | 3,420 | 5,753 | +68% | 1 | 1 | 0% | 435 | 533 | +23% | 0 | 0 | — |
case-19 | fail→fail | 7,675 | 7,272 | -5% | 1 | 1 | 0% | 1,178 | 466 | -60% | 0 | 0 | — |
case-20 | fail→pass | 18,990 | 10,929 | -42% | 1 | 1 | 0% | 2,461 | 2,063 | -16% | 0 | 0 | — |
case-21 | fail→fail | 15,629 | 5,098 | -67% | 1 | 1 | 0% | 2,894 | 352 | -88% | 0 | 0 | — |
case-22 | fail→pass | 19,344 | 9,728 | -50% | 1 | 1 | 0% | 3,096 | 1,975 | -36% | 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, and 13 counted toward the lift figure. The other 9 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +36 percentage points is the difference between those two pass rates over the 13 comparable cases. 6 cases got worse with the skill loaded, and they are included in that figure.
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.