Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when reviewing image assets, icon libraries, or components that render illustrations to identify culture-specific content that may need locale overrides or replacement with neutral alternatives.
.claude/skills/thedaviddias-locale-images/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-13 | ✗→✓ | ▲ Improved | -28% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | -11% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 7% | 0% |
| case-09 | ✓→✓ | ＝ Same ✓ | 29% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 12% | 0% |
Images that feel natural in one culture can be confusing or offensive in another — a thumbs-up is a vulgar gesture in parts of the Middle East and West Africa, and a checkmark means "wrong" in Japan. Choosing neutral visuals by default and providing targeted overrides avoids unintentional offence and keeps the image management surface small.
Identify image assets, SVG icons, and illustration components that use hand gestures, color symbolism, animals, or religious imagery that may carry different meanings across cultures or locales.
Replace culturally specific images with locale-neutral alternatives where possible, and implement a locale-aware image lookup with a neutral fallback for images that must vary by region.
Explain why culturally specific imagery causes problems for international audiences and how a locale-image lookup pattern provides targeted overrides without requiring a separate asset for every locale.
Review React components and asset pipelines for hardcoded image paths, hand gesture icons, flag icons used to represent languages, and any image that contains embedded text. Flag each as a potential localization concern.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/i18n/locale-images
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 16,487 | 10,418 | -37% | 1 | 1 | 0% | 2,983 | 2,296 | -23% | 0 | 0 | — |
case-02 | pass→pass | 11,415 | 8,034 | -30% | 1 | 1 | 0% | 2,120 | 1,887 | -11% | 0 | 0 | — |
case-03 | pass→pass | 13,957 | 12,228 | -12% | 1 | 1 | 0% | 2,780 | 2,972 | +7% | 0 | 0 | — |
case-09 | pass→pass | 10,050 | 10,208 | +2% | 1 | 1 | 0% | 1,724 | 2,219 | +29% | 0 | 0 | — |
case-04 | pass→pass | 9,359 | 8,228 | -12% | 1 | 1 | 0% | 1,927 | 2,156 | +12% | 0 | 0 | — |
case-05 | pass→pass | 10,586 | 8,128 | -23% | 1 | 1 | 0% | 1,858 | 1,893 | +2% | 0 | 0 | — |
case-06 | pass→pass | 10,711 | 8,308 | -22% | 1 | 1 | 0% | 1,641 | 1,727 | +5% | 0 | 0 | — |
case-07 | pass→pass | 9,167 | 7,182 | -22% | 1 | 1 | 0% | 1,596 | 1,561 | -2% | 0 | 0 | — |
case-08 | pass→pass | 15,231 | 11,572 | -24% | 1 | 1 | 0% | 2,673 | 2,420 | -9% | 0 | 0 | — |
case-10 | pass→pass | 15,382 | 12,132 | -21% | 1 | 1 | 0% | 2,359 | 2,384 | +1% | 0 | 0 | — |
case-11 | pass→pass | 11,719 | 7,377 | -37% | 1 | 1 | 0% | 1,866 | 1,686 | -10% | 0 | 0 | — |
case-12 | pass→pass | 14,182 | 9,603 | -32% | 1 | 1 | 0% | 2,852 | 2,340 | -18% | 0 | 0 | — |
case-13 | fail→pass | 18,866 | 10,842 | -43% | 1 | 1 | 0% | 3,100 | 2,231 | -28% | 0 | 0 | — |
case-14 | pass→pass | 13,903 | 10,139 | -27% | 1 | 1 | 0% | 2,421 | 2,226 | -8% | 0 | 0 | — |
case-15 | pass→pass | 11,984 | 7,382 | -38% | 1 | 1 | 0% | 2,162 | 1,702 | -21% | 0 | 0 | — |
case-16 | pass→pass | 11,784 | 9,476 | -20% | 1 | 1 | 0% | 1,921 | 1,871 | -3% | 0 | 0 | — |
case-17 | pass→pass | 13,072 | 8,600 | -34% | 1 | 1 | 0% | 2,012 | 1,783 | -11% | 0 | 0 | — |
case-18 | pass→pass | 10,307 | 9,454 | -8% | 1 | 1 | 0% | 1,739 | 2,072 | +19% | 0 | 0 | — |
case-19 | pass→pass | 13,460 | 8,841 | -34% | 1 | 1 | 0% | 2,119 | 1,715 | -19% | 0 | 0 | — |
case-20 | pass→pass | 6,810 | 3,147 | -54% | 1 | 1 | 0% | 1,369 | 938 | -31% | 0 | 0 | — |
case-21 | pass→pass | 13,519 | 9,341 | -31% | 1 | 1 | 0% | 2,378 | 2,003 | -16% | 0 | 0 | — |
case-22 | pass→pass | 14,084 | 8,208 | -42% | 1 | 1 | 0% | 2,308 | 1,804 | -22% | 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 +5 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.