Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Write articles, guides, blog posts, tutorials, newsletter issues, and other long-form content in a distinctive voice derived from supplied examples or brand guidance. Use when the user wants polished written content longer than a paragraph, especially when voice consistency, structure, and credibility matter.
.claude/skills/affaan-m-article-writing/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 22% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 20% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-04 | ✓→✗ | ▼ Worse | 29% | 0% |
撰写听起来像真人或真实品牌的长篇内容，而非通用的 AI 输出。
如果用户需要特定的声音，请收集以下一项或多项：
然后提取：
如果未提供声音参考，则默认为直接、运营者风格的声音：具体、实用，且少用夸张宣传。
删除并重写以下任何内容：
交付前：
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 26,172 | 21,686 | -17% | 1 | 1 | 0% | 4,006 | 4,354 | +9% | 0 | 0 | — |
case-02 | fail→fail | 20,754 | 18,483 | -11% | 1 | 1 | 0% | 2,773 | 3,114 | +12% | 0 | 0 | — |
case-03 | fail→pass | 17,083 | 13,521 | -21% | 1 | 1 | 0% | 2,375 | 2,620 | +10% | 0 | 0 | — |
case-04 | pass→fail | 26,135 | 25,399 | -3% | 1 | 1 | 0% | 2,893 | 3,730 | +29% | 0 | 0 | — |
case-05 | fail→pass | 16,608 | 15,500 | -7% | 1 | 1 | 0% | 2,524 | 3,087 | +22% | 0 | 0 | — |
case-06 | fail→fail | 17,950 | 18,342 | +2% | 1 | 1 | 0% | 3,163 | 3,866 | +22% | 0 | 0 | — |
case-07 | fail→fail | 17,912 | 18,621 | +4% | 1 | 1 | 0% | 2,424 | 3,070 | +27% | 0 | 0 | — |
case-08 | fail→fail | 28,621 | 23,451 | -18% | 1 | 1 | 0% | 4,307 | 4,814 | +12% | 0 | 0 | — |
case-09 | fail→pass | 18,428 | 17,249 | -6% | 1 | 1 | 0% | 2,656 | 3,175 | +20% | 0 | 0 | — |
case-10 | fail→fail | 23,792 | 35,388 | +49% | 1 | 1 | 0% | 3,105 | 4,074 | +31% | 0 | 0 | — |
case-11 | pass→pass | 23,386 | 23,023 | -2% | 1 | 1 | 0% | 3,377 | 3,941 | +17% | 0 | 0 | — |
case-12 | fail→fail | 22,557 | 22,745 | +1% | 1 | 1 | 0% | 4,169 | 4,614 | +11% | 0 | 0 | — |
case-13 | fail→fail | 11,358 | 13,124 | +16% | 1 | 1 | 0% | 1,799 | 2,815 | +56% | 0 | 0 | — |
case-14 | fail→fail | 20,409 | 20,619 | +1% | 1 | 1 | 0% | 2,908 | 3,631 | +25% | 0 | 0 | — |
case-15 | fail→fail | 23,627 | 17,851 | -24% | 1 | 1 | 0% | 3,857 | 3,778 | -2% | 0 | 0 | — |
case-16 | fail→fail | 8,331 | 6,474 | -22% | 1 | 1 | 0% | 1,216 | 1,742 | +43% | 0 | 0 | — |
case-17 | fail→pass | 15,869 | 11,452 | -28% | 1 | 1 | 0% | 2,411 | 2,487 | +3% | 0 | 0 | — |
case-18 | fail→fail | 23,248 | 22,440 | -3% | 1 | 1 | 0% | 3,350 | 3,837 | +15% | 0 | 0 | — |
case-19 | fail→fail | 19,928 | 20,959 | +5% | 1 | 1 | 0% | 3,468 | 4,622 | +33% | 0 | 0 | — |
case-20 | pass→pass | 4,837 | 4,703 | -3% | 1 | 1 | 0% | 890 | 1,602 | +80% | 0 | 0 | — |
case-21 | fail→fail | 6,191 | 2,294 | -63% | 1 | 1 | 0% | 945 | 1,051 | +11% | 0 | 0 | — |
case-22 | pass→pass | 3,436 | 2,847 | -17% | 1 | 1 | 0% | 514 | 1,165 | +127% | 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 +14 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is 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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/3/2026 | +23% |
Other measured skills in the registry, with their headline benchmark lift.