Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Plan the localization of a product/content for a new market — beyond translating the words. Use when asked to localize a product, plan market entry localization, prepare a localization brief, or figure out what to adapt for a new region. Produces a brief — target locales, what to translate vs. adapt vs. rebuild (UI, content, formats, imagery, payments, legal), priorities, and the risks/cultural pitfalls.
.claude/skills/mohitagw15856-localization-brief/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 0% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 25% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-04 | ✓→✗ | ▼ Worse | 177% | 0% |
| case-05 | ✓→✗ | ▼ Worse | 9% | 0% |
Localization is not translation — it's making a product feel native in a market, which touches formats, imagery, payment methods, legal norms, and cultural expectations far beyond the strings. This skill plans it: what to translate, what to adapt, what to rebuild for the locale, in priority order, with the cultural and regulatory pitfalls that sink naïve "just translate the UI" launches.
Ask for these only if they aren't already provided:
1. Scope per locale — language + region, and the depth (translate-only vs. full localization).
2. Translate / Adapt / Rebuild — the core matrix; what each element needs:
| Area | Action | Notes | |---|---|---| | UI strings | translate | register, length expansion (DE ~+30%) | | Dates/numbers/currency | adapt | formats, separators, currency + display | | Imagery / examples | adapt | culturally appropriate people, scenarios, names | | Payments | rebuild | local methods (e.g. Alipay/WeChat in CN, iDEAL in NL) | | Legal / privacy | adapt | local consent, terms, data residency | | Content / SEO | adapt | local keywords, not translated ones | | Tone / formality | adapt | formality norms, humour that travels |
3. Priorities — what to do first for the goal (often: UI + payments + legal for a real launch; UI + a landing page for a market test). Sequence by impact.
4. Cultural & regulatory pitfalls — the specific traps for this market: colour/symbol connotations, name/address/phone formats, RTL if relevant, regulated claims, censorship/hosting requirements. The stuff that embarrasses or blocks a launch.
5. Process & QA — who translates (native + in-market review), how strings are managed (don't hard-code), and pseudo-localization / in-context QA before launch.
Localization / internationalization practice — the translate/adapt/rebuild model, locale formats, market-specific payments & legal, in-country QA.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 30,319 | 24,936 | -18% | 1 | 1 | 0% | 5,096 | 4,820 | -5% | 0 | 0 | — |
case-02 | fail→fail | 32,699 | 20,008 | -39% | 1 | 1 | 0% | 5,631 | 4,191 | -26% | 0 | 0 | — |
case-03 | pass→pass | 9,424 | 10,247 | +9% | 1 | 1 | 0% | 1,826 | 2,642 | +45% | 0 | 0 | — |
case-04 | pass→fail | 6,024 | 12,309 | +104% | 1 | 1 | 0% | 1,111 | 3,081 | +177% | 0 | 0 | — |
case-05 | pass→fail | 19,804 | 17,213 | -13% | 1 | 1 | 0% | 3,560 | 3,894 | +9% | 0 | 0 | — |
case-06 | pass→pass | 21,634 | 15,398 | -29% | 1 | 1 | 0% | 3,523 | 3,219 | -9% | 0 | 0 | — |
case-07 | fail→pass | 22,021 | 17,086 | -22% | 1 | 1 | 0% | 3,739 | 3,730 | -0% | 0 | 0 | — |
case-08 | pass→fail | 18,331 | 17,951 | -2% | 1 | 1 | 0% | 2,899 | 3,570 | +23% | 0 | 0 | — |
case-09 | pass→pass | 19,216 | 18,793 | -2% | 1 | 1 | 0% | 2,981 | 3,786 | +27% | 0 | 0 | — |
case-10 | pass→pass | 16,882 | 18,469 | +9% | 1 | 1 | 0% | 2,695 | 3,809 | +41% | 0 | 0 | — |
case-11 | pass→pass | 15,636 | 20,757 | +33% | 1 | 1 | 0% | 2,608 | 3,145 | +21% | 0 | 0 | — |
case-12 | fail→pass | 20,484 | 18,095 | -12% | 1 | 1 | 0% | 2,810 | 3,524 | +25% | 0 | 0 | — |
case-13 | pass→pass | 21,247 | 19,415 | -9% | 1 | 1 | 0% | 3,011 | 3,284 | +9% | 0 | 0 | — |
case-14 | pass→pass | 27,062 | 21,462 | -21% | 1 | 1 | 0% | 3,031 | 4,136 | +36% | 0 | 0 | — |
case-15 | fail→fail | 17,800 | 17,317 | -3% | 1 | 1 | 0% | 3,045 | 3,667 | +20% | 0 | 0 | — |
case-16 | pass→pass | 18,564 | 17,843 | -4% | 1 | 1 | 0% | 3,186 | 3,918 | +23% | 0 | 0 | — |
case-17 | fail→fail | 15,772 | 17,774 | +13% | 1 | 1 | 0% | 2,738 | 3,924 | +43% | 0 | 0 | — |
case-18 | fail→fail | 15,740 | 16,229 | +3% | 1 | 1 | 0% | 2,899 | 3,363 | +16% | 0 | 0 | — |
case-19 | fail→fail | 14,636 | 14,379 | -2% | 1 | 1 | 0% | 2,614 | 3,206 | +23% | 0 | 0 | — |
case-20 | fail→fail | 17,545 | 23,442 | +34% | 1 | 1 | 0% | 2,897 | 4,579 | +58% | 0 | 0 | — |
case-21 | fail→fail | 16,343 | 20,668 | +26% | 1 | 1 | 0% | 2,630 | 4,246 | +61% | 0 | 0 | — |
case-22 | fail→pass | 16,751 | 17,744 | +6% | 1 | 1 | 0% | 3,030 | 3,759 | +24% | 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.