Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股集中度风险/分散化分析。当用户说"集中度"、"分散化"、"concentration risk"、"持仓集中"、"行业集中"、"个股集中"、"分散投资"时触发。基于 cn-stock-data 获取数据，评估组合集中度风险。支持 formal/brief 两种输出风格。
.claude/skills/aifinlab-a-share-concentration-risk/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-14 | ✗→✓ | ▲ Improved | -40% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -45% | 0% |
| case-18 | ✓→✗ | ▼ Worse | 9% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 69% | 0% |
通过 cn-stock-data skill 获取数据：
# 组合集中度分析报告
## 一、个股集中度
| 指标 | 数值 | 评估 |
|------|------|------|
| HHI | 0.08 | 中等 |
## 二、行业集中度
[行业偏离分析]
## 三、风险集中度
[风险贡献Top5]
## 四、分散化建议## 集中度速览
- HHI 0.08，有效持仓12只
- Top5占比 48%，偏集中
- 最大行业超配：电子+8%
- 建议：降低Top3权重，增加行业分散参考 references/concentration-risk-guide.md 获取详细方法论与 A股实证研究。
python# 调用 skill result = run_skill({ "param1": "value1", "param2": "value2" })
bashpython scripts/run_skill.py --input data.json
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 36,392 | 51,201 | +41% | 1 | 1 | 0% | 5,045 | 8,885 | +76% | 0 | 0 | — |
case-02 | fail→fail | 15,807 | 21,222 | +34% | 1 | 1 | 0% | 2,418 | 2,824 | +17% | 0 | 0 | — |
case-03 | fail→fail | 30,118 | 42,397 | +41% | 1 | 1 | 0% | 4,821 | 8,875 | +84% | 0 | 0 | — |
case-04 | pass→pass | 5,228 | 6,432 | +23% | 1 | 1 | 0% | 860 | 1,454 | +69% | 0 | 0 | — |
case-05 | pass→pass | 5,853 | 10,484 | +79% | 1 | 1 | 0% | 687 | 2,079 | +203% | 0 | 0 | — |
case-06 | pass→pass | 4,875 | 7,500 | +54% | 1 | 1 | 0% | 755 | 1,891 | +150% | 0 | 0 | — |
case-07 | pass→pass | 16,495 | 15,837 | -4% | 1 | 1 | 0% | 2,661 | 2,738 | +3% | 0 | 0 | — |
case-08 | pass→pass | 13,690 | 13,302 | -3% | 1 | 1 | 0% | 2,196 | 2,196 | 0% | 0 | 0 | — |
case-09 | pass→pass | 14,613 | 11,391 | -22% | 1 | 1 | 0% | 2,116 | 2,278 | +8% | 0 | 0 | — |
case-10 | pass→pass | 16,979 | 16,729 | -1% | 1 | 1 | 0% | 2,292 | 3,079 | +34% | 0 | 0 | — |
case-11 | pass→pass | 17,275 | 22,248 | +29% | 1 | 1 | 0% | 2,745 | 3,616 | +32% | 0 | 0 | — |
case-12 | fail→fail | 15,566 | 10,215 | -34% | 1 | 1 | 0% | 2,046 | 1,940 | -5% | 0 | 0 | — |
case-13 | pass→pass | 13,413 | 6,521 | -51% | 1 | 1 | 0% | 2,017 | 1,509 | -25% | 0 | 0 | — |
case-14 | fail→pass | 29,889 | 10,943 | -63% | 1 | 1 | 0% | 3,515 | 2,119 | -40% | 0 | 0 | — |
case-15 | fail→fail | 21,236 | 26,576 | +25% | 1 | 1 | 0% | 2,830 | 3,962 | +40% | 0 | 0 | — |
case-16 | pass→pass | 16,931 | 4,543 | -73% | 1 | 1 | 0% | 2,348 | 1,185 | -50% | 0 | 0 | — |
case-17 | fail→pass | 18,963 | 16,881 | -11% | 1 | 1 | 0% | 2,967 | 3,357 | +13% | 0 | 0 | — |
case-18 | pass→fail | 11,831 | 11,842 | +0% | 1 | 1 | 0% | 1,888 | 2,058 | +9% | 0 | 0 | — |
case-19 | fail→pass | 12,210 | 3,434 | -72% | 1 | 1 | 0% | 1,852 | 1,027 | -45% | 0 | 0 | — |
case-20 | pass→pass | 23,964 | 29,903 | +25% | 1 | 1 | 0% | 3,642 | 6,600 | +81% | 0 | 0 | — |
case-21 | pass→pass | 16,311 | 14,550 | -11% | 1 | 1 | 0% | 3,042 | 3,343 | +10% | 0 | 0 | — |
case-22 | pass→pass | 29,377 | 33,956 | +16% | 1 | 1 | 0% | 5,863 | 6,798 | +16% | 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 +9 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.
Other measured skills in the registry, with their headline benchmark lift.