Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股估值分位/历史估值区间分析。当用户说"估值分位"、"历史估值"、"估值贵不贵"、"XX现在贵吗"、"估值百分位"、"PE分位"、"PB分位"、"估值区间"、"历史PE"、"valuation band"时触发。基于 cn-stock-data 历史行情和财务数据，计算个股或指数当前估值在历史区间中的分位数位置，判断估值高低。支持研报风格（formal）和快速判断风格（brief）。不适用于 DCF 估值（用 a-share-dcf）或可比公司（用 a-share-comps）。
.claude/skills/aifinlab-a-share-valuation-band/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-11 | ✗→✓ | ▲ Improved | 5% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 28% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 38% | 0% |
bashSCRIPTS="$SKILLS_ROOT/cn-stock-data/scripts" # 近3-5年K线（计算历史估值区间） python "$SCRIPTS/cn_stock_data.py" kline --code [CODE] --freq daily --start [3年前日期] # 当前行情（最新PE/PB） python "$SCRIPTS/cn_stock_data.py" quote --code [CODE] # 财务数据（历史ROE/利润等） python "$SCRIPTS/cn_stock_data.py" finance --code [CODE]
补充：通过 web 搜索获取指数 PE/PB 历史数据（如沪深300 PE历史中位数）。
| 百分位区间 | 判断 | 含义 | |-----------|------|------| | 0-20% | 极度低估 | 历史底部区域，长期配置价值高 | | 20-40% | 偏低估 | 低于历史均值，具备安全边际 | | 40-60% | 合理区间 | 接近历史中枢，估值中性 | | 60-80% | 偏高估 | 高于历史均值，注意风险 | | 80-100% | 极度高估 | 历史顶部区域，风险收益比差 |
| 维度 | formal | brief | |------|--------|-------| | 输出 | 完整估值分位报告 | 一句话结论 + 核心数据 | | 指标 | PE + PB + PS 多维度 | 仅核心 1-2 个指标 | | 历史对比 | 近 3/5/10 年分别计算 | 仅近 3 年 | | 图表 | 提供估值带数据（可画图） | 无 | | 篇幅 | 500-800 字 | 50-100 字 |
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 | 25,943 | 13,559 | -48% | 1 | 1 | 0% | 3,773 | 1,353 | -64% | 0 | 0 | — |
case-11 | fail→pass | 9,981 | 3,824 | -62% | 1 | 1 | 0% | 1,566 | 1,637 | +5% | 0 | 0 | — |
case-02 | fail→fail | 8,992 | 9,988 | +11% | 1 | 1 | 0% | 1,346 | 1,403 | +4% | 0 | 0 | — |
case-03 | fail→fail | 20,112 | 10,501 | -48% | 1 | 1 | 0% | 3,086 | 1,530 | -50% | 0 | 0 | — |
case-04 | pass→fail | 21,860 | 11,957 | -45% | 1 | 1 | 0% | 3,304 | 1,552 | -53% | 0 | 0 | — |
case-05 | pass→fail | 19,885 | 10,101 | -49% | 1 | 1 | 0% | 2,954 | 1,581 | -46% | 0 | 0 | — |
case-06 | fail→pass | 20,874 | 17,962 | -14% | 1 | 1 | 0% | 2,775 | 3,543 | +28% | 0 | 0 | — |
case-07 | pass→pass | 19,789 | 18,801 | -5% | 1 | 1 | 0% | 2,867 | 3,750 | +31% | 0 | 0 | — |
case-08 | pass→pass | 22,688 | 23,261 | +3% | 1 | 1 | 0% | 3,261 | 4,084 | +25% | 0 | 0 | — |
case-09 | fail→pass | 9,370 | 4,380 | -53% | 1 | 1 | 0% | 1,579 | 1,718 | +9% | 0 | 0 | — |
case-10 | fail→pass | 7,230 | 3,382 | -53% | 1 | 1 | 0% | 1,363 | 1,486 | +9% | 0 | 0 | — |
case-12 | pass→pass | 15,338 | 10,962 | -29% | 1 | 1 | 0% | 2,575 | 2,776 | +8% | 0 | 0 | — |
case-13 | pass→pass | 12,924 | 5,726 | -56% | 1 | 1 | 0% | 1,923 | 1,584 | -18% | 0 | 0 | — |
case-14 | pass→pass | 21,817 | 19,217 | -12% | 1 | 1 | 0% | 3,092 | 3,829 | +24% | 0 | 0 | — |
case-15 | pass→fail | 31,618 | 9,390 | -70% | 1 | 1 | 0% | 6,029 | 1,459 | -76% | 0 | 0 | — |
case-16 | fail→fail | 7,284 | 8,876 | +22% | 1 | 1 | 0% | 1,005 | 1,463 | +46% | 0 | 0 | — |
case-17 | pass→pass | 15,889 | 11,083 | -30% | 1 | 1 | 0% | 2,288 | 2,467 | +8% | 0 | 0 | — |
case-18 | fail→fail | 8,645 | 12,984 | +50% | 1 | 1 | 0% | 1,264 | 1,732 | +37% | 0 | 0 | — |
case-19 | fail→pass | 18,132 | 17,745 | -2% | 1 | 1 | 0% | 2,705 | 3,730 | +38% | 0 | 0 | — |
case-20 | pass→fail | 27,968 | 9,073 | -68% | 1 | 1 | 0% | 5,569 | 1,423 | -74% | 0 | 0 | — |
case-21 | pass→fail | 23,237 | 8,159 | -65% | 1 | 1 | 0% | 4,407 | 1,598 | -64% | 0 | 0 | — |
case-22 | pass→pass | 24,157 | 26,904 | +11% | 1 | 1 | 0% | 3,835 | 4,745 | +24% | 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 12 counted toward the lift figure. The other 10 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 0 percentage points is the difference between those two pass rates over the 12 comparable cases. 7 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.