Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generate a big, wide spread of ideas for anything by running the prompt through many different lenses at once, then clustering and picking. Use when asked to brainstorm ideas for, give me lots of options, help me come up with, or I need ideas for. Produces a high-volume, deliberately varied idea list generated across multiple angles (safe, wild, cheap, ambitious, weird, opposite), grouped into themes, and a shortlist of the most promising — maximizing range so you're not choosing from three obvi
.claude/skills/mohitagw15856-idea-storm/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | 210% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 41% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 5% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 20% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 1% | 0% |
A short brainstorm gives you the three obvious ideas everyone gets. This maximizes range instead: it runs your prompt through many different generative lenses — the safe idea, the wild idea, the cheapest idea, the most ambitious, the weirdest, the opposite — to produce a genuinely wide spread, then clusters and shortlists so the volume is usable. Wide first, then narrow.
Ask for these if not provided:
The storm (by lens)
Clustered
Shortlist: the most promising few]. Wildcard to keep: one out-there option].
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | pass→pass | 23,373 | 20,728 | -11% | 1 | 1 | 0% | 3,180 | 3,336 | +5% | 0 | 0 | — |
case-01 | pass→pass | 21,232 | 18,288 | -14% | 1 | 1 | 0% | 2,569 | 3,084 | +20% | 0 | 0 | — |
case-02 | fail→fail | 36,786 | 34,639 | -6% | 1 | 1 | 0% | 4,171 | 4,464 | +7% | 0 | 0 | — |
case-03 | pass→pass | 25,406 | 21,994 | -13% | 1 | 1 | 0% | 3,316 | 3,347 | +1% | 0 | 0 | — |
case-05 | pass→pass | 21,298 | 26,344 | +24% | 1 | 1 | 0% | 2,624 | 4,049 | +54% | 0 | 0 | — |
case-06 | pass→pass | 22,809 | 27,377 | +20% | 1 | 1 | 0% | 2,754 | 4,080 | +48% | 0 | 0 | — |
case-07 | pass→pass | 20,869 | 21,379 | +2% | 1 | 1 | 0% | 3,299 | 3,072 | -7% | 0 | 0 | — |
case-08 | fail→pass | 5,766 | 14,043 | +144% | 1 | 1 | 0% | 917 | 2,846 | +210% | 0 | 0 | — |
case-09 | pass→pass | 16,993 | 20,109 | +18% | 1 | 1 | 0% | 2,680 | 3,084 | +15% | 0 | 0 | — |
case-10 | fail→pass | 19,494 | 19,860 | +2% | 1 | 1 | 0% | 2,190 | 3,083 | +41% | 0 | 0 | — |
case-11 | pass→pass | 19,567 | 19,415 | -1% | 1 | 1 | 0% | 2,772 | 2,687 | -3% | 0 | 0 | — |
case-12 | pass→pass | 20,647 | 15,944 | -23% | 1 | 1 | 0% | 2,585 | 3,129 | +21% | 0 | 0 | — |
case-13 | pass→pass | 23,571 | 19,644 | -17% | 1 | 1 | 0% | 2,750 | 3,041 | +11% | 0 | 0 | — |
case-14 | pass→pass | 26,754 | 22,307 | -17% | 1 | 1 | 0% | 3,385 | 3,224 | -5% | 0 | 0 | — |
case-15 | pass→pass | 23,111 | 24,889 | +8% | 1 | 1 | 0% | 2,879 | 3,615 | +26% | 0 | 0 | — |
case-16 | pass→pass | 24,135 | 22,053 | -9% | 1 | 1 | 0% | 3,170 | 3,457 | +9% | 0 | 0 | — |
case-17 | pass→pass | 17,488 | 13,048 | -25% | 1 | 1 | 0% | 2,773 | 2,983 | +8% | 0 | 0 | — |
case-18 | pass→pass | 15,280 | 22,132 | +45% | 1 | 1 | 0% | 2,458 | 3,663 | +49% | 0 | 0 | — |
case-19 | pass→pass | 26,268 | 28,461 | +8% | 1 | 1 | 0% | 2,778 | 4,269 | +54% | 0 | 0 | — |
case-20 | pass→pass | 24,609 | 22,716 | -8% | 1 | 1 | 0% | 2,609 | 3,428 | +31% | 0 | 0 | — |
case-21 | pass→pass | 16,713 | 20,628 | +23% | 1 | 1 | 0% | 2,663 | 4,193 | +57% | 0 | 0 | — |
case-22 | pass→pass | 26,801 | 23,788 | -11% | 1 | 1 | 0% | 3,183 | 3,673 | +15% | 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 +9 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.