Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →提供超越字面的翻译，包含文化禁忌分析与语境优化。
.claude/skills/sdsds222-global-translator-expert/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 0% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 20% | 0% |
<Skill_Start>
你是一个全球顶级的本地化翻译专家。请严格按以下步骤进行内部推演：
系统指令：完成上面的思考与推演后，请务必单独换行输出与开头成对的 Skill_End 标签，然后在标签下方严格按以下格式给出最终结果 ]
【执行浓缩】：(用 1-2 句话高度概括原文的核心难点以及你采用的关键策略，例如：识别出双关、避开了某文化禁忌等) 【最终结果】：(输出经过精心打磨的正式翻译成品；如有必要，可按需提供“商务版”或“口语版”备选)
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 11,893 | 49,717 | +318% | 1 | 1 | 0% | 1,762 | 2,268 | +29% | 0 | 0 | — |
case-02 | fail→pass | 13,199 | 12,540 | -5% | 1 | 1 | 0% | 2,444 | 2,434 | -0% | 0 | 0 | — |
case-03 | fail→pass | 12,705 | 13,618 | +7% | 1 | 1 | 0% | 2,059 | 2,453 | +19% | 0 | 0 | — |
case-04 | pass→fail | 12,417 | 8,847 | -29% | 1 | 1 | 0% | 2,034 | 1,701 | -16% | 0 | 0 | — |
case-05 | pass→fail | 7,853 | 8,669 | +10% | 1 | 1 | 0% | 1,443 | 1,718 | +19% | 0 | 0 | — |
case-06 | pass→fail | 5,294 | 8,414 | +59% | 1 | 1 | 0% | 994 | 1,661 | +67% | 0 | 0 | — |
case-07 | fail→pass | 13,807 | 10,034 | -27% | 1 | 1 | 0% | 2,403 | 2,206 | -8% | 0 | 0 | — |
case-08 | fail→pass | 11,156 | 10,478 | -6% | 1 | 1 | 0% | 1,719 | 2,055 | +20% | 0 | 0 | — |
case-09 | fail→pass | 15,366 | 12,101 | -21% | 1 | 1 | 0% | 2,657 | 2,196 | -17% | 0 | 0 | — |
case-10 | fail→pass | 9,016 | 19,332 | +114% | 1 | 1 | 0% | 1,312 | 1,971 | +50% | 0 | 0 | — |
case-11 | fail→pass | 13,010 | 10,750 | -17% | 1 | 1 | 0% | 2,183 | 2,274 | +4% | 0 | 0 | — |
case-12 | fail→pass | 19,953 | 16,412 | -18% | 1 | 1 | 0% | 3,463 | 2,875 | -17% | 0 | 0 | — |
case-13 | fail→pass | 10,538 | 11,687 | +11% | 1 | 1 | 0% | 1,728 | 2,109 | +22% | 0 | 0 | — |
case-14 | fail→pass | 14,796 | 12,467 | -16% | 1 | 1 | 0% | 2,198 | 2,272 | +3% | 0 | 0 | — |
case-15 | fail→pass | 15,250 | 13,492 | -12% | 1 | 1 | 0% | 2,346 | 2,498 | +6% | 0 | 0 | — |
case-16 | fail→pass | 7,855 | 11,570 | +47% | 1 | 1 | 0% | 1,277 | 2,220 | +74% | 0 | 0 | — |
case-17 | fail→pass | 10,182 | 9,094 | -11% | 1 | 1 | 0% | 1,674 | 1,766 | +5% | 0 | 0 | — |
case-18 | fail→pass | 16,813 | 13,991 | -17% | 1 | 1 | 0% | 2,750 | 2,481 | -10% | 0 | 0 | — |
case-19 | fail→pass | 10,332 | 7,701 | -25% | 1 | 1 | 0% | 1,699 | 1,595 | -6% | 0 | 0 | — |
case-20 | fail→pass | 26,747 | 19,689 | -26% | 1 | 1 | 0% | 4,319 | 3,283 | -24% | 0 | 0 | — |
case-21 | fail→pass | 8,566 | 10,649 | +24% | 1 | 1 | 0% | 1,385 | 1,916 | +38% | 0 | 0 | — |
case-22 | fail→pass | 13,405 | 9,543 | -29% | 1 | 1 | 0% | 2,168 | 1,896 | -13% | 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 +73 percentage points is the difference between those two pass rates over the 22 comparable cases. 3 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.