Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →公司研究助手 - 高股息版。专注于高股息公司研究，包括高分红、高股息率、稳定分红的公司。 **触发场景**： - 用户研究高股息公司（股息率>5%、稳定分红、现金流充沛） - 分析公司分红能力、分红意愿、分红可持续性 - 高股息公司估值分析（股息率、分红率） - 高股息公司竞品对标、收益对比 - 写高股息公司研报、收息投资组合建议 **关键词**："高股息"、"分红"、"股息率"、"收息"、"红利"、"现金流"、"稳定分红"、"银行"、"煤炭"、"公用事业"
.claude/skills/aifinlab-company-research-high-dividend/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 18% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 61% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 22% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 122% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 62% | 0% |
你是一名专注于高股息公司的资深研究员，擅长分析分红能力、分红意愿、股息可持续性与收息投资价值。
| 标准 | 指标 | |------|------| | 股息率 | >5%（近 12 个月） | | 分红率 | >30%（分红/归母净利） | | 分红稳定性 | 连续 3 年 + 分红 | | 现金流 | 经营现金流充沛、自由现金流为正 |
1. 公司定位（行业/股息率/分红历史）
2. 分红能力（盈利/现金流/财务健康）
3. 股息率与分红率
4. 股息可持续性
5. 投资亮点与风险1. 公司概况（业务/市值/行业地位）
2. 分红历史分析（连续分红年数、分红率趋势）
3. 分红能力分析（盈利/现金流/财务健康）
4. 股息可持续性分析
5. 竞争格局（同业股息率对比）
6. 估值分析（股息率历史分位、总回报）
7. 催化剂与风险
8. 投资建议（收息价值）| 指标 | 公司 A | 公司 B | 公司 C | 行业平均 | |------|--------|--------|--------|----------| | 市值 (亿) | | | | | | 股息率 (%) | | | | | | 分红率 (%) | | | | | | PE-TTM | | | | | | PB | | | | | | ROE(%) | | | | | | 经营现金流 (亿) | | | | | | 自由现金流 (亿) | | | | | | 资产负债率 (%) | | | | | | 连续分红年数 | | | | |
输出前自查：
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→pass | 32,868 | 34,611 | +5% | 1 | 1 | 0% | 5,159 | 6,108 | +18% | 0 | 0 | — |
case-02 | fail→fail | 19,594 | 24,421 | +25% | 1 | 1 | 0% | 3,370 | 5,430 | +61% | 0 | 0 | — |
case-03 | fail→pass | 26,855 | 32,915 | +23% | 1 | 1 | 0% | 3,943 | 6,335 | +61% | 0 | 0 | — |
case-04 | fail→pass | 18,351 | 16,871 | -8% | 1 | 1 | 0% | 2,856 | 3,494 | +22% | 0 | 0 | — |
case-05 | pass→pass | 17,377 | 14,123 | -19% | 1 | 1 | 0% | 2,652 | 3,288 | +24% | 0 | 0 | — |
case-06 | pass→pass | 12,111 | 14,942 | +23% | 1 | 1 | 0% | 1,921 | 3,058 | +59% | 0 | 0 | — |
case-07 | pass→pass | 19,730 | 23,412 | +19% | 1 | 1 | 0% | 2,845 | 4,368 | +54% | 0 | 0 | — |
case-08 | pass→pass | 14,511 | 19,665 | +36% | 1 | 1 | 0% | 2,076 | 4,133 | +99% | 0 | 0 | — |
case-09 | pass→pass | 15,177 | 35,907 | +137% | 1 | 1 | 0% | 2,397 | 3,448 | +44% | 0 | 0 | — |
case-10 | pass→pass | 26,413 | 30,321 | +15% | 1 | 1 | 0% | 3,791 | 5,752 | +52% | 0 | 0 | — |
case-11 | pass→pass | 22,244 | 22,427 | +1% | 1 | 1 | 0% | 3,246 | 4,173 | +29% | 0 | 0 | — |
case-12 | pass→pass | 16,936 | 19,702 | +16% | 1 | 1 | 0% | 2,580 | 3,937 | +53% | 0 | 0 | — |
case-13 | fail→pass | 18,031 | 37,681 | +109% | 1 | 1 | 0% | 3,278 | 7,284 | +122% | 0 | 0 | — |
case-14 | pass→pass | 22,761 | 27,910 | +23% | 1 | 1 | 0% | 3,113 | 5,350 | +72% | 0 | 0 | — |
case-15 | pass→pass | 19,591 | 24,345 | +24% | 1 | 1 | 0% | 2,797 | 4,875 | +74% | 0 | 0 | — |
case-16 | pass→pass | 19,264 | 21,578 | +12% | 1 | 1 | 0% | 2,712 | 4,052 | +49% | 0 | 0 | — |
case-17 | pass→pass | 20,304 | 21,145 | +4% | 1 | 1 | 0% | 3,188 | 4,580 | +44% | 0 | 0 | — |
case-18 | pass→pass | 18,823 | 17,327 | -8% | 1 | 1 | 0% | 2,774 | 3,674 | +32% | 0 | 0 | — |
case-19 | fail→pass | 16,825 | 20,042 | +19% | 1 | 1 | 0% | 2,502 | 4,057 | +62% | 0 | 0 | — |
case-20 | pass→fail | 24,715 | 27,864 | +13% | 1 | 1 | 0% | 4,417 | 5,913 | +34% | 0 | 0 | — |
case-21 | pass→pass | 25,948 | 28,299 | +9% | 1 | 1 | 0% | 4,208 | 5,164 | +23% | 0 | 0 | — |
case-22 | pass→pass | 17,419 | 16,196 | -7% | 1 | 1 | 0% | 3,067 | 3,904 | +27% | 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 +18 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.