Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Turn a long-form article, thread, memo, or product narrative into a compact clickable capability map with a workflow loop, use-case matrix, and responsive detail panel.
.claude/skills/nexu-io-codex-interactive-capability-map/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | 497% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 316% | 0% |
| case-20 | ✓→✗ | ▼ Worse | 49% | 0% |
Transform a long article, social thread, memo, product essay, or launch narrative into a visual explainer that people can scan and click through.
Emit one single-file HTML artifact:
html<artifact identifier="codex-interactive-capability-map" type="text/html" title="Codex Interactive Capability Map"> <!doctype html> <html>...</html> </artifact>
Include all CSS and JavaScript inline. Do not use lorem ipsum. Do not leave placeholder cards. If the source does not include enough concrete details, infer a small, clearly labeled conceptual model from the source instead of inventing unrelated content.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 37,866 | 84,525 | +123% | 1 | 1 | 0% | 8,282 | 8,825 | +7% | 0 | 0 | — |
case-02 | fail→fail | 50,488 | 37,765 | -25% | 1 | 1 | 0% | 8,269 | 8,812 | +7% | 0 | 0 | — |
case-15 | pass→pass | 14,937 | 40,566 | +172% | 1 | 1 | 0% | 2,151 | 8,761 | +307% | 0 | 0 | — |
case-03 | fail→fail | 50,884 | 50,660 | -0% | 1 | 1 | 0% | 8,270 | 8,813 | +7% | 0 | 0 | — |
case-04 | fail→fail | 42,514 | 37,483 | -12% | 1 | 1 | 0% | 8,233 | 8,776 | +7% | 0 | 0 | — |
case-05 | fail→pass | 9,529 | 36,034 | +278% | 1 | 1 | 0% | 1,471 | 8,777 | +497% | 0 | 0 | — |
case-06 | pass→pass | 29,611 | 35,547 | +20% | 1 | 1 | 0% | 6,343 | 8,769 | +38% | 0 | 0 | — |
case-20 | pass→fail | 37,364 | 51,495 | +38% | 1 | 1 | 0% | 5,896 | 8,779 | +49% | 0 | 0 | — |
case-07 | fail→fail | 9,002 | 43,152 | +379% | 1 | 1 | 0% | 1,496 | 8,762 | +486% | 0 | 0 | — |
case-08 | fail→pass | 9,715 | 3,084 | -68% | 1 | 1 | 0% | 1,284 | 1,060 | -17% | 0 | 0 | — |
case-09 | pass→pass | 20,802 | 53,027 | +155% | 1 | 1 | 0% | 2,216 | 8,761 | +295% | 0 | 0 | — |
case-10 | pass→pass | 11,400 | 38,219 | +235% | 1 | 1 | 0% | 1,810 | 8,765 | +384% | 0 | 0 | — |
case-11 | pass→pass | 18,534 | 29,049 | +57% | 1 | 1 | 0% | 2,733 | 3,636 | +33% | 0 | 0 | — |
case-12 | fail→pass | 14,733 | 10,697 | -27% | 1 | 1 | 0% | 2,112 | 2,383 | +13% | 0 | 0 | — |
case-13 | pass→pass | 19,915 | 37,583 | +89% | 1 | 1 | 0% | 2,737 | 8,755 | +220% | 0 | 0 | — |
case-14 | pass→pass | 17,100 | 36,269 | +112% | 1 | 1 | 0% | 2,411 | 8,770 | +264% | 0 | 0 | — |
case-16 | fail→pass | 26,675 | 37,751 | +42% | 1 | 1 | 0% | 2,105 | 8,762 | +316% | 0 | 0 | — |
case-17 | pass→fail | 18,990 | 38,174 | +101% | 1 | 1 | 0% | 2,449 | 8,755 | +257% | 0 | 0 | — |
case-18 | pass→pass | 16,786 | 49,357 | +194% | 1 | 1 | 0% | 2,277 | 8,768 | +285% | 0 | 0 | — |
case-19 | fail→fail | 21,390 | 38,679 | +81% | 1 | 1 | 0% | 2,589 | 8,757 | +238% | 0 | 0 | — |
case-21 | pass→fail | 30,723 | 53,735 | +75% | 1 | 1 | 0% | 5,085 | 8,779 | +73% | 0 | 0 | — |
case-22 | pass→fail | 14,600 | 63,780 | +337% | 1 | 1 | 0% | 3,108 | 8,779 | +182% | 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 0 percentage points is the difference between those two pass rates over the 22 comparable cases. 4 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.