Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股成长质量评估/PEG分析。当用户说"成长性"、"成长质量"、"PEG"、"增长质量"、"growth"、"XX成长性怎么样"、"高增长"、"增速"、"可持续增长"、"增长驱动力"时触发。多维度评估企业成长的质量（可持续性/稳定性/含金量），计算PEG及增长驱动因子分解，区分"好增长"和"差增长"。支持研报风格（formal）和快速评估风格（brief）。
.claude/skills/aifinlab-a-share-growth-quality/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-11 | ✗→✓ | ▲ Improved | 61% | 0% |
| case-04 | ✓→✗ | ▼ Worse | -42% | 0% |
| case-05 | ✓→✗ | ▼ Worse | -36% | 0% |
| case-06 | ✓→✗ | ▼ Worse | -28% | 0% |
| case-07 | ✓→✗ | ▼ Worse | -18% | 0% |
训练数据中的财务数据已过期。 每次执行时必须：
bashSCRIPTS="$SKILLS_ROOT/cn-stock-data/scripts" # 财务指标（多期数据用于计算 CAGR） python "$SCRIPTS/cn_stock_data.py" finance --code [CODE] # 近 3 年日线行情（用于估值分位和趋势） python "$SCRIPTS/cn_stock_data.py" kline --code [CODE] --freq daily --start [3年前日期] # 实时行情（当前 PE / 市值） python "$SCRIPTS/cn_stock_data.py" quote --code [CODE]
参见 references/growth-quality-guide.md 的详细方法论。
根据风格要求选择输出格式：
输出核心结论：
QC 验证：
| 维度 | formal（研报风格） | brief（快速评估） | |------|-------------------|------------------| | 篇幅 | 6-10 页 .docx | 1-2 页 Markdown | | 标题 | "公司名成长质量深度评估" | "XX 成长性速评" | | 结构 | 核心结论→成长指标→质量评估→PEG估值→同行对比→风险→附表 | 成长评分→PEG→关键发现→结论 | | 图表 | 4-6 个（营收/利润增速趋势、ROE分解、PEG同行比较、增长驱动瀑布图） | 成长指标汇总表 1 个 | | 引用 | 必须标注数据来源和分析师出处 | 可省略 | | 免责声明 | 需要 | 不需要 | | 默认风格 | 非默认 | 默认 |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 24,660 | 10,982 | -55% | 1 | 1 | 0% | 3,491 | 2,048 | -41% | 0 | 0 | — |
case-02 | fail→fail | 58,025 | 15,188 | -74% | 1 | 1 | 0% | 8,312 | 2,273 | -73% | 0 | 0 | — |
case-03 | fail→fail | 25,825 | 10,050 | -61% | 1 | 1 | 0% | 3,727 | 2,139 | -43% | 0 | 0 | — |
case-04 | pass→fail | 22,813 | 8,455 | -63% | 1 | 1 | 0% | 3,357 | 1,948 | -42% | 0 | 0 | — |
case-05 | pass→fail | 28,311 | 9,807 | -65% | 1 | 1 | 0% | 4,268 | 2,749 | -36% | 0 | 0 | — |
case-06 | pass→fail | 19,419 | 11,011 | -43% | 1 | 1 | 0% | 2,928 | 2,118 | -28% | 0 | 0 | — |
case-07 | pass→fail | 19,957 | 11,738 | -41% | 1 | 1 | 0% | 2,625 | 2,141 | -18% | 0 | 0 | — |
case-08 | pass→pass | 7,125 | 7,270 | +2% | 1 | 1 | 0% | 1,059 | 2,489 | +135% | 0 | 0 | — |
case-09 | pass→pass | 20,146 | 17,345 | -14% | 1 | 1 | 0% | 2,564 | 4,012 | +56% | 0 | 0 | — |
case-10 | pass→pass | 22,107 | 24,037 | +9% | 1 | 1 | 0% | 3,109 | 4,571 | +47% | 0 | 0 | — |
case-11 | fail→pass | 15,036 | 12,426 | -17% | 1 | 1 | 0% | 2,084 | 3,355 | +61% | 0 | 0 | — |
case-12 | pass→pass | 20,462 | 19,516 | -5% | 1 | 1 | 0% | 2,834 | 4,234 | +49% | 0 | 0 | — |
case-13 | pass→pass | 17,726 | 13,632 | -23% | 1 | 1 | 0% | 2,559 | 3,618 | +41% | 0 | 0 | — |
case-14 | pass→fail | 25,698 | 10,414 | -59% | 1 | 1 | 0% | 3,579 | 2,106 | -41% | 0 | 0 | — |
case-15 | fail→fail | 25,230 | 30,591 | +21% | 1 | 1 | 0% | 3,641 | 5,291 | +45% | 0 | 0 | — |
case-16 | pass→pass | 19,197 | 22,958 | +20% | 1 | 1 | 0% | 2,894 | 4,482 | +55% | 0 | 0 | — |
case-17 | pass→pass | 16,990 | 24,078 | +42% | 1 | 1 | 0% | 2,615 | 4,224 | +62% | 0 | 0 | — |
case-18 | pass→pass | 15,978 | 9,595 | -40% | 1 | 1 | 0% | 2,114 | 2,900 | +37% | 0 | 0 | — |
case-19 | fail→fail | 45,659 | 50,155 | +10% | 1 | 1 | 0% | 6,613 | 2,001 | -70% | 0 | 0 | — |
case-20 | pass→pass | 26,221 | 29,627 | +13% | 1 | 1 | 0% | 3,762 | 5,884 | +56% | 0 | 0 | — |
case-21 | pass→fail | 21,802 | 14,077 | -35% | 1 | 1 | 0% | 3,111 | 1,895 | -39% | 0 | 0 | — |
case-22 | pass→pass | 17,348 | 9,311 | -46% | 1 | 1 | 0% | 2,495 | 3,037 | +22% | 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, and 13 counted toward the lift figure. The other 9 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 -23 percentage points is the difference between those two pass rates over the 13 comparable cases. 9 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.