Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use for marketing, social, replies, product copy, launch copy, and writing in the user's voice when their Emulo writing profile should guide the task. Do not use for unrelated execution or UI/UX design alone.
.claude/skills/ohad6k-write/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 69% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 75% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 51% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 21% | 0% |
emulo.py two directories above this skill; fall back to ./emulo.py only for a direct repo checkout.EMULO_PY, then run python "$EMULO_PY" plugin profile-path --domain write.## Voice laws, ## Casual register, ## Professional register) loads voice laws always, plus exactly one register:run emulo and deepen write to mine it.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 7,130 | 17,451 | +145% | 1 | 1 | 0% | 1,178 | 2,316 | +97% | 0 | 0 | — |
case-02 | fail→fail | 6,281 | 6,393 | +2% | 1 | 1 | 0% | 828 | 598 | -28% | 0 | 0 | — |
case-03 | fail→fail | 11,289 | 33,652 | +198% | 1 | 1 | 0% | 1,524 | 2,059 | +35% | 0 | 0 | — |
case-04 | fail→fail | 11,017 | 11,343 | +3% | 1 | 1 | 0% | 1,804 | 1,037 | -43% | 0 | 0 | — |
case-05 | fail→fail | 5,444 | 7,430 | +36% | 1 | 1 | 0% | 742 | 831 | +12% | 0 | 0 | — |
case-06 | fail→pass | 18,004 | 6,918 | -62% | 1 | 1 | 0% | 893 | 1,507 | +69% | 0 | 0 | — |
case-07 | fail→fail | 22,665 | 22,226 | -2% | 1 | 1 | 0% | 3,760 | 3,794 | +1% | 0 | 0 | — |
case-08 | fail→pass | 11,740 | 26,589 | +126% | 1 | 1 | 0% | 1,402 | 2,460 | +75% | 0 | 0 | — |
case-09 | fail→pass | 10,939 | 16,150 | +48% | 1 | 1 | 0% | 1,805 | 1,637 | -9% | 0 | 0 | — |
case-10 | fail→pass | 7,758 | 8,537 | +10% | 1 | 1 | 0% | 1,178 | 1,776 | +51% | 0 | 0 | — |
case-11 | fail→pass | 13,489 | 11,116 | -18% | 1 | 1 | 0% | 1,851 | 2,245 | +21% | 0 | 0 | — |
case-12 | fail→pass | 7,168 | 7,402 | +3% | 1 | 1 | 0% | 1,132 | 1,565 | +38% | 0 | 0 | — |
case-13 | fail→pass | 9,272 | 13,433 | +45% | 1 | 1 | 0% | 1,504 | 2,286 | +52% | 0 | 0 | — |
case-14 | fail→pass | 6,232 | 22,247 | +257% | 1 | 1 | 0% | 985 | 3,386 | +244% | 0 | 0 | — |
case-15 | pass→fail | 13,005 | 8,932 | -31% | 1 | 1 | 0% | 1,738 | 625 | -64% | 0 | 0 | — |
case-16 | pass→fail | 12,806 | 15,097 | +18% | 1 | 1 | 0% | 1,944 | 2,335 | +20% | 0 | 0 | — |
case-17 | fail→pass | 13,749 | 15,900 | +16% | 1 | 1 | 0% | 1,819 | 1,788 | -2% | 0 | 0 | — |
case-18 | fail→fail | 16,117 | 7,332 | -55% | 1 | 1 | 0% | 2,465 | 644 | -74% | 0 | 0 | — |
case-19 | fail→pass | 8,856 | 6,448 | -27% | 1 | 1 | 0% | 1,275 | 1,252 | -2% | 0 | 0 | — |
case-20 | pass→pass | 17,567 | 13,257 | -25% | 1 | 1 | 0% | 2,803 | 3,196 | +14% | 0 | 0 | — |
case-21 | pass→pass | 7,651 | 5,388 | -30% | 1 | 1 | 0% | 1,015 | 1,081 | +7% | 0 | 0 | — |
case-22 | pass→pass | 3,736 | 7,180 | +92% | 1 | 1 | 0% | 563 | 1,384 | +146% | 0 | 0 | — |
case-23 | pass→pass | 7,660 | 14,701 | +92% | 1 | 1 | 0% | 1,519 | 2,513 | +65% | 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, and 17 counted toward the lift figure. The other 6 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 +35 percentage points is the difference between those two pass rates over the 17 comparable cases. 2 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.