Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when targeting 《中国体育科技》(China Sport Science and Technology) — the CSSCI applied/competitive sport-science journal from the China Institute of Sport Science (国家体育总局体育科学研究所), sibling to 《体育科学》. Best for competitive sport, sports training practice, talent identification, physical conditioning, technical/tactical analysis, training-load and physiological/biochemical monitoring, anti-doping, and sports rehabilitation — i.e. application-oriented exercise science. Route theory-heavy or综合 contributi
.claude/skills/brycewang-stanford-china-sport-science-and-technology/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 8% | 0% |
| case-21 | ✗→✓ | ▲ Improved | 44% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 15% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 32% | 0% |
国家体育总局体育科学研究所主办的 CSSCI 体育学来源刊，与《体育科学》同出体科所 / 学会体系，但分工不同：《体育科学》偏学科理论与综合分量，《中国体育科技》偏应用、竞技体育与运动训练实战。核心要求是面向训练 / 备战 / 运动表现的真问题 + 可落地的科学证据：测试与监控严谨、结论能指导实践。纯理论思辨或与竞技 / 健康实践脱节的稿件在这里会显得"用不上"。
china-sport-science（《体育科学》）；体育社会科学问题 → journal-of-sports-research / journal-of-shanghai-university-of-sport；运动人体科学基础研究亦可看 journal-of-tianjin-university-of-sport。cn-sport-journal-workflow）。正式建议前仍需进入官方核验清单。../../resources/source-basis.md 和 ../../resources/official-source-map.md，从官方来源锚点（体科所 / 期刊官网 / 官方采编系统 / 官方 CNKI 页面）开始核验，并说明核验日期。【匹配度】高 / 中 / 低
【路径】竞技训练 / 运动人体科学应用 / 运动医学康复 / 体育科技应用
【应用落点】<指导何种训练/备战/健康实践一句话>
【证据强度】设计/对照/功效充分 / 不足（需补）
【实践指导】可操作 + 边界条件清晰 / 仅报指标（需补）
【贡献】改进 <训练/监控/康复实践>| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | fail→pass | 24,374 | 20,603 | -15% | 1 | 1 | 0% | 3,299 | 3,639 | +10% | 0 | 0 | — |
case-03 | fail→pass | 29,445 | 16,884 | -43% | 1 | 1 | 0% | 3,067 | 3,313 | +8% | 0 | 0 | — |
case-19 | pass→pass | 20,015 | 21,049 | +5% | 1 | 1 | 0% | 2,788 | 3,615 | +30% | 0 | 0 | — |
case-20 | pass→pass | 17,063 | 19,919 | +17% | 1 | 1 | 0% | 2,416 | 3,595 | +49% | 0 | 0 | — |
case-21 | fail→pass | 26,560 | 21,704 | -18% | 1 | 1 | 0% | 2,467 | 3,558 | +44% | 0 | 0 | — |
case-01 | fail→pass | 25,138 | 17,575 | -30% | 1 | 1 | 0% | 2,658 | 3,047 | +15% | 0 | 0 | — |
case-04 | pass→pass | 23,419 | 18,745 | -20% | 1 | 1 | 0% | 2,843 | 3,444 | +21% | 0 | 0 | — |
case-05 | fail→pass | 28,728 | 24,476 | -15% | 1 | 1 | 0% | 2,947 | 3,894 | +32% | 0 | 0 | — |
case-06 | pass→pass | 15,208 | 15,667 | +3% | 1 | 1 | 0% | 1,733 | 2,982 | +72% | 0 | 0 | — |
case-07 | fail→pass | 25,127 | 20,351 | -19% | 1 | 1 | 0% | 2,958 | 3,644 | +23% | 0 | 0 | — |
case-08 | fail→fail | 27,002 | 19,374 | -28% | 1 | 1 | 0% | 3,298 | 3,702 | +12% | 0 | 0 | — |
case-09 | fail→pass | 19,967 | 25,715 | +29% | 1 | 1 | 0% | 2,843 | 4,158 | +46% | 0 | 0 | — |
case-10 | pass→pass | 24,422 | 16,442 | -33% | 1 | 1 | 0% | 2,913 | 3,771 | +29% | 0 | 0 | — |
case-11 | fail→pass | 27,044 | 20,139 | -26% | 1 | 1 | 0% | 3,178 | 4,336 | +36% | 0 | 0 | — |
case-12 | fail→pass | 23,326 | 27,559 | +18% | 1 | 1 | 0% | 2,761 | 4,257 | +54% | 0 | 0 | — |
case-13 | pass→pass | 25,669 | 23,113 | -10% | 1 | 1 | 0% | 2,895 | 3,748 | +29% | 0 | 0 | — |
case-14 | fail→pass | 43,456 | 28,313 | -35% | 1 | 1 | 0% | 3,086 | 4,623 | +50% | 0 | 0 | — |
case-22 | fail→pass | 21,296 | 8,323 | -61% | 1 | 1 | 0% | 2,498 | 2,548 | +2% | 0 | 0 | — |
case-15 | pass→pass | 28,802 | 28,887 | +0% | 1 | 1 | 0% | 3,357 | 4,463 | +33% | 0 | 0 | — |
case-16 | fail→pass | 23,830 | 13,991 | -41% | 1 | 1 | 0% | 2,663 | 3,101 | +16% | 0 | 0 | — |
case-17 | pass→pass | 26,511 | 30,060 | +13% | 1 | 1 | 0% | 3,603 | 4,626 | +28% | 0 | 0 | — |
case-18 | fail→fail | 26,692 | 21,009 | -21% | 1 | 1 | 0% | 3,107 | 4,071 | +31% | 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 +55 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.