Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股同行财务对比/两股PK。当用户说"对比"、"比较"、"PK"、"XX和YY比"、"谁更好"、"财务对比"、"同行比较"、"peer compare"、"哪个更值得买"时触发。基于 cn-stock-data 获取两家或多家公司的财务和行情数据，进行多维度对比。与 a-share-comps 的区别：comps 侧重估值对标（PE/PB/EV），本 skill 侧重财务质量全面对比（盈利/成长/现金流/效率/风险）。支持研报风格（formal）和快速对比风格（brief）。
.claude/skills/aifinlab-a-share-peer-compare/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-12 | ✗→✓ | ▲ Improved | 107% | 0% |
| case-09 | ✓→✗ | ▼ Worse | -67% | 0% |
| case-17 | ✓→✗ | ▼ Worse | -56% | 0% |
| case-18 | ✓→✗ | ▼ Worse | -53% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | 23% | 0% |
通过 cn-stock-data skill 获取数据：
finance 路由：财务三表、财务指标（ROE/毛利率/净利率/资产负债率等）quote 路由：实时行情、PE/PB 等估值指标按以下 6 维度逐项对比，每项标注"优/中/弱"：
| 维度 | 核心指标 | 更优判断 | |------|---------|---------| | 盈利能力 | ROE、毛利率、净利率 | 数值越高越优 | | 成长性 | 营收增速、净利润增速 | 增速越高越优，兼看持续性 | | 现金流 | 经营现金流/净利润、自由现金流 | 比值>1 且稳定为优 | | 运营效率 | 存货周转率、应收账款周转率 | 周转越快越优 | | 财务安全 | 资产负债率、流动比率、利息保障倍数 | 负债率低+流动比率>1.5 为优 | | 估值水平 | PE、PB、PEG | 结合成长性，低估值+高成长为优 |
formal 风格（默认）：
brief 风格：
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-12 | fail→pass | 18,481 | 40,631 | +120% | 1 | 1 | 0% | 3,468 | 7,180 | +107% | 0 | 0 | — |
case-01 | fail→fail | 26,663 | 8,622 | -68% | 1 | 1 | 0% | 4,712 | 1,326 | -72% | 0 | 0 | — |
case-02 | fail→fail | 14,773 | 9,488 | -36% | 1 | 1 | 0% | 2,099 | 1,286 | -39% | 0 | 0 | — |
case-03 | fail→fail | 40,417 | 8,505 | -79% | 1 | 1 | 0% | 6,412 | 1,466 | -77% | 0 | 0 | — |
case-04 | fail→fail | 27,877 | 35,536 | +27% | 1 | 1 | 0% | 4,034 | 6,962 | +73% | 0 | 0 | — |
case-05 | pass→pass | 22,838 | 25,944 | +14% | 1 | 1 | 0% | 3,711 | 4,583 | +23% | 0 | 0 | — |
case-06 | pass→pass | 24,181 | 19,792 | -18% | 1 | 1 | 0% | 3,320 | 3,512 | +6% | 0 | 0 | — |
case-07 | fail→fail | 17,213 | 11,723 | -32% | 1 | 1 | 0% | 2,572 | 1,361 | -47% | 0 | 0 | — |
case-08 | fail→fail | 20,804 | 42,807 | +106% | 1 | 1 | 0% | 3,393 | 6,555 | +93% | 0 | 0 | — |
case-09 | pass→fail | 21,097 | 26,805 | +27% | 1 | 1 | 0% | 3,807 | 1,241 | -67% | 0 | 0 | — |
case-10 | fail→fail | 9,296 | 11,842 | +27% | 1 | 1 | 0% | 1,478 | 1,495 | +1% | 0 | 0 | — |
case-11 | fail→fail | 15,834 | 22,804 | +44% | 1 | 1 | 0% | 2,863 | 4,751 | +66% | 0 | 0 | — |
case-25 | fail→fail | 27,725 | 13,428 | -52% | 1 | 1 | 0% | 4,579 | 1,711 | -63% | 0 | 0 | — |
case-13 | fail→fail | 12,160 | 25,937 | +113% | 1 | 1 | 0% | 1,727 | 5,306 | +207% | 0 | 0 | — |
case-14 | fail→fail | 12,721 | 7,034 | -45% | 1 | 1 | 0% | 2,352 | 1,138 | -52% | 0 | 0 | — |
case-15 | pass→pass | 28,898 | 51,394 | +78% | 1 | 1 | 0% | 4,680 | 7,854 | +68% | 0 | 0 | — |
case-16 | pass→pass | 23,510 | 24,956 | +6% | 1 | 1 | 0% | 3,994 | 4,812 | +20% | 0 | 0 | — |
case-17 | pass→fail | 23,536 | 11,646 | -51% | 1 | 1 | 0% | 3,390 | 1,484 | -56% | 0 | 0 | — |
case-18 | pass→fail | 17,051 | 10,317 | -39% | 1 | 1 | 0% | 2,779 | 1,313 | -53% | 0 | 0 | — |
case-19 | fail→fail | 26,098 | 38,421 | +47% | 1 | 1 | 0% | 4,678 | 8,311 | +78% | 0 | 0 | — |
case-20 | fail→fail | 15,847 | 9,449 | -40% | 1 | 1 | 0% | 2,370 | 1,275 | -46% | 0 | 0 | — |
case-21 | fail→fail | 17,213 | 18,602 | +8% | 1 | 1 | 0% | 3,042 | 3,978 | +31% | 0 | 0 | — |
case-22 | pass→pass | 27,564 | 48,273 | +75% | 1 | 1 | 0% | 4,336 | 8,060 | +86% | 0 | 0 | — |
case-23 | pass→pass | 22,693 | 29,125 | +28% | 1 | 1 | 0% | 4,352 | 6,627 | +52% | 0 | 0 | — |
case-24 | fail→fail | 10,863 | 8,416 | -23% | 1 | 1 | 0% | 1,746 | 1,235 | -29% | 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. 25 cases were attempted, and 13 counted toward the lift figure. The other 12 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of -8 percentage points is the difference between those two pass rates over the 13 comparable cases. 5 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.