Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Write or review UI copy — buttons, errors, empty states, microcopy, notifications, labels — using the voice & tone system (clear, concise, useful, human, honest) with the what→why→how error formula and inclusive-language rules. Use when the user needs interface copy, error messages, empty-state text, or a copy review.
.claude/skills/plugin87-ux-writing/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 69% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 12% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 34% | 0% |
Produce or critique interface copy in the project's voice.
content/voice-tone.md (voice principles, tone spectrum, formulas, microcopy patterns, inclusive language, pre-ship checklist).Final copy (or a redline review) that reads naturally aloud and passes the checklist. Keep within any character limits for tight UI.
Run every line through the 10-item pre-ship checklist in content/voice-tone.md — do not skip it:
accessibility/i18n-rtl.md).| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 10,741 | 7,941 | -26% | 1 | 1 | 0% | 1,744 | 1,804 | +3% | 0 | 0 | — |
case-02 | fail→pass | 8,387 | 10,703 | +28% | 1 | 1 | 0% | 1,267 | 2,147 | +69% | 0 | 0 | — |
case-03 | fail→fail | 9,860 | 9,713 | -1% | 1 | 1 | 0% | 1,514 | 2,146 | +42% | 0 | 0 | — |
case-04 | pass→pass | 5,698 | 6,512 | +14% | 1 | 1 | 0% | 894 | 1,399 | +56% | 0 | 0 | — |
case-05 | pass→pass | 17,983 | 13,919 | -23% | 1 | 1 | 0% | 2,840 | 2,737 | -4% | 0 | 0 | — |
case-06 | pass→pass | 7,075 | 8,790 | +24% | 1 | 1 | 0% | 1,157 | 1,856 | +60% | 0 | 0 | — |
case-07 | pass→pass | 5,848 | 6,730 | +15% | 1 | 1 | 0% | 898 | 1,562 | +74% | 0 | 0 | — |
case-08 | fail→pass | 8,083 | 7,001 | -13% | 1 | 1 | 0% | 1,360 | 1,540 | +13% | 0 | 0 | — |
case-09 | pass→pass | 8,689 | 9,048 | +4% | 1 | 1 | 0% | 1,286 | 1,911 | +49% | 0 | 0 | — |
case-10 | pass→pass | 14,030 | 11,135 | -21% | 1 | 1 | 0% | 2,325 | 2,193 | -6% | 0 | 0 | — |
case-11 | pass→pass | 6,579 | 9,428 | +43% | 1 | 1 | 0% | 999 | 1,765 | +77% | 0 | 0 | — |
case-12 | pass→pass | 6,402 | 6,287 | -2% | 1 | 1 | 0% | 1,022 | 1,429 | +40% | 0 | 0 | — |
case-13 | pass→pass | 7,272 | 9,462 | +30% | 1 | 1 | 0% | 1,111 | 2,043 | +84% | 0 | 0 | — |
case-14 | fail→pass | 9,025 | 6,693 | -26% | 1 | 1 | 0% | 1,392 | 1,559 | +12% | 0 | 0 | — |
case-15 | pass→pass | 10,592 | 9,768 | -8% | 1 | 1 | 0% | 1,626 | 1,903 | +17% | 0 | 0 | — |
case-16 | pass→pass | 13,579 | 10,334 | -24% | 1 | 1 | 0% | 2,052 | 2,124 | +4% | 0 | 0 | — |
case-17 | fail→pass | 8,383 | 8,383 | 0% | 1 | 1 | 0% | 1,295 | 1,731 | +34% | 0 | 0 | — |
case-18 | pass→pass | 9,635 | 7,444 | -23% | 1 | 1 | 0% | 1,467 | 1,719 | +17% | 0 | 0 | — |
case-19 | pass→pass | 4,655 | 6,567 | +41% | 1 | 1 | 0% | 758 | 1,388 | +83% | 0 | 0 | — |
case-20 | pass→pass | 14,509 | 15,301 | +5% | 1 | 1 | 0% | 2,213 | 2,938 | +33% | 0 | 0 | — |
case-21 | pass→fail | 22,048 | 18,534 | -16% | 1 | 1 | 0% | 4,393 | 3,896 | -11% | 0 | 0 | — |
case-22 | pass→pass | 17,394 | 16,265 | -6% | 1 | 1 | 0% | 3,329 | 3,362 | +1% | 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 +18 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is 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.