Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Compatibility alias for OpenAI platform documentation guidance. Delegate to the canonical local `openai-docs` payload while preserving route compatibility.
.claude/skills/foryourhealth111-pixel-openai-knowledge/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | -38% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 34% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -42% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -52% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -54% | 0% |
Provide a stable compatibility alias for callers that still request openai-knowledge, while canonical OpenAI documentation guidance is maintained under the sibling openai-docs skill directory.
This preserves:
openai-knowledgeskills-lock and catalog continuity for legacy callersopenai-docs skill payload first.../openai-docs/SKILL.md../openai-docs/references/**../openai-docs/scripts/**../openai-docs/SKILL.md for the full OpenAI docs workflow.../openai-docs/.openai-knowledge only when a caller explicitly requested this alias.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-08 | fail→pass | 10,708 | 5,073 | -53% | 1 | 1 | 0% | 1,851 | 1,155 | -38% | 0 | 0 | — |
case-01 | pass→pass | 10,343 | 8,868 | -14% | 1 | 1 | 0% | 1,649 | 1,733 | +5% | 0 | 0 | — |
case-02 | pass→pass | 10,915 | 7,888 | -28% | 1 | 1 | 0% | 2,065 | 1,711 | -17% | 0 | 0 | — |
case-03 | fail→pass | 7,658 | 8,205 | +7% | 1 | 1 | 0% | 1,361 | 1,830 | +34% | 0 | 0 | — |
case-04 | fail→pass | 9,021 | 4,271 | -53% | 1 | 1 | 0% | 1,492 | 868 | -42% | 0 | 0 | — |
case-05 | fail→pass | 8,719 | 2,531 | -71% | 1 | 1 | 0% | 1,335 | 640 | -52% | 0 | 0 | — |
case-06 | fail→pass | 10,674 | 3,527 | -67% | 1 | 1 | 0% | 1,737 | 793 | -54% | 0 | 0 | — |
case-07 | fail→pass | 11,110 | 4,364 | -61% | 1 | 1 | 0% | 1,752 | 918 | -48% | 0 | 0 | — |
case-09 | fail→pass | 7,411 | 3,185 | -57% | 1 | 1 | 0% | 1,181 | 765 | -35% | 0 | 0 | — |
case-10 | fail→pass | 8,001 | 1,833 | -77% | 1 | 1 | 0% | 1,330 | 593 | -55% | 0 | 0 | — |
case-11 | fail→pass | 12,497 | 2,974 | -76% | 1 | 1 | 0% | 2,226 | 696 | -69% | 0 | 0 | — |
case-12 | fail→pass | 9,791 | 2,613 | -73% | 1 | 1 | 0% | 1,550 | 676 | -56% | 0 | 0 | — |
case-13 | pass→pass | 13,130 | 2,698 | -79% | 1 | 1 | 0% | 2,193 | 732 | -67% | 0 | 0 | — |
case-14 | fail→pass | 6,022 | 3,571 | -41% | 1 | 1 | 0% | 917 | 872 | -5% | 0 | 0 | — |
case-15 | fail→pass | 10,608 | 3,648 | -66% | 1 | 1 | 0% | 1,908 | 862 | -55% | 0 | 0 | — |
case-16 | fail→pass | 10,432 | 2,782 | -73% | 1 | 1 | 0% | 1,566 | 614 | -61% | 0 | 0 | — |
case-17 | fail→pass | 10,652 | 2,672 | -75% | 1 | 1 | 0% | 1,815 | 653 | -64% | 0 | 0 | — |
case-18 | fail→pass | 7,172 | 2,343 | -67% | 1 | 1 | 0% | 1,173 | 645 | -45% | 0 | 0 | — |
case-19 | fail→pass | 11,870 | 5,829 | -51% | 1 | 1 | 0% | 1,842 | 1,124 | -39% | 0 | 0 | — |
case-20 | pass→pass | 10,068 | 5,430 | -46% | 1 | 1 | 0% | 1,608 | 1,123 | -30% | 0 | 0 | — |
case-21 | fail→pass | 14,923 | 2,593 | -83% | 1 | 1 | 0% | 2,487 | 729 | -71% | 0 | 0 | — |
case-22 | fail→pass | 7,085 | 4,573 | -35% | 1 | 1 | 0% | 1,097 | 957 | -13% | 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 +82 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.