Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Write empty-state content that turns a blank screen into a next step. Use when asked to write an empty state, a zero-data / first-run state, a no-results state, or onboarding placeholder content. Produces empty-state copy — a clear headline, a helpful line, and a primary action — for each type (first-use, user-cleared, no-results, error/permission), so a blank screen guides instead of confuses.
.claude/skills/mohitagw15856-empty-state-writer/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 231% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 973% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 44% | 0% |
| case-02 | ✓→✓ | = Same ✓ | 63% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 133% | 0% |
An empty state is the most-missed onboarding moment: the user arrives and there's nothing there. Done well, it explains the value, removes confusion, and offers the one action that fills the screen. This skill writes empty states that teach and activate — not blank voids or generic "No data" labels.
Given "the empty state for a projects list", write it anyway — infer why the screen is empty, the value of the feature, and the best first action, labelling assumptions. Cover the distinct empty-state types that apply. Never hand back a question instead of copy.
Ask for these only if they aren't already provided (else infer and label):
Write the relevant types (skip those that don't apply):
For each: Headline · Supporting line · Action(s), plus a one-line note on the intended tone/illustration.
UX writing & onboarding practice — empty states as activation moments, differentiated by type, with value framing and a single clear action.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 13,040 | 7,895 | -39% | 1 | 1 | 0% | 2,110 | 2,109 | -0% | 0 | 0 | — |
case-02 | pass→pass | 8,381 | 9,183 | +10% | 1 | 1 | 0% | 1,394 | 2,270 | +63% | 0 | 0 | — |
case-03 | fail→pass | 4,347 | 7,853 | +81% | 1 | 1 | 0% | 658 | 2,176 | +231% | 0 | 0 | — |
case-04 | pass→pass | 6,114 | 9,300 | +52% | 1 | 1 | 0% | 973 | 2,267 | +133% | 0 | 0 | — |
case-05 | fail→pass | 1,856 | 11,804 | +536% | 1 | 1 | 0% | 240 | 2,575 | +973% | 0 | 0 | — |
case-06 | pass→pass | 6,223 | 7,546 | +21% | 1 | 1 | 0% | 1,130 | 2,027 | +79% | 0 | 0 | — |
case-07 | fail→fail | 11,384 | 6,670 | -41% | 1 | 1 | 0% | 2,031 | 1,914 | -6% | 0 | 0 | — |
case-08 | fail→fail | 14,716 | 11,369 | -23% | 1 | 1 | 0% | 2,122 | 2,481 | +17% | 0 | 0 | — |
case-09 | fail→pass | 8,375 | 8,432 | +1% | 1 | 1 | 0% | 1,467 | 2,109 | +44% | 0 | 0 | — |
case-10 | fail→fail | 7,512 | 12,161 | +62% | 1 | 1 | 0% | 1,255 | 2,698 | +115% | 0 | 0 | — |
case-11 | pass→pass | 6,404 | 6,799 | +6% | 1 | 1 | 0% | 1,172 | 1,781 | +52% | 0 | 0 | — |
case-12 | fail→fail | 7,814 | 7,695 | -2% | 1 | 1 | 0% | 1,236 | 2,113 | +71% | 0 | 0 | — |
case-13 | pass→pass | 8,595 | 8,009 | -7% | 1 | 1 | 0% | 1,299 | 1,878 | +45% | 0 | 0 | — |
case-14 | pass→pass | 7,681 | 9,124 | +19% | 1 | 1 | 0% | 1,314 | 2,053 | +56% | 0 | 0 | — |
case-15 | pass→pass | 9,974 | 9,029 | -9% | 1 | 1 | 0% | 1,646 | 2,045 | +24% | 0 | 0 | — |
case-16 | fail→fail | 9,674 | 8,552 | -12% | 1 | 1 | 0% | 1,331 | 2,099 | +58% | 0 | 0 | — |
case-17 | pass→pass | 6,632 | 7,633 | +15% | 1 | 1 | 0% | 1,059 | 1,991 | +88% | 0 | 0 | — |
case-18 | pass→pass | 7,391 | 9,234 | +25% | 1 | 1 | 0% | 1,098 | 2,248 | +105% | 0 | 0 | — |
case-19 | pass→pass | 6,926 | 8,809 | +27% | 1 | 1 | 0% | 1,087 | 2,162 | +99% | 0 | 0 | — |
case-20 | pass→pass | 4,833 | 6,295 | +30% | 1 | 1 | 0% | 783 | 1,708 | +118% | 0 | 0 | — |
case-21 | pass→pass | 15,480 | 14,052 | -9% | 1 | 1 | 0% | 2,407 | 2,947 | +22% | 0 | 0 | — |
case-22 | pass→pass | 7,869 | 9,530 | +21% | 1 | 1 | 0% | 1,375 | 2,232 | +62% | 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 +14 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.