Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →A股相对强弱/RS分析。当用户说"相对强弱"、"relative strength"、"RS"、"XX比大盘强吗"、"跑赢跑输"时触发。量化分析个股/板块相对大盘的强弱。支持formal和brief风格。
.claude/skills/aifinlab-a-share-relative-strength/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | -44% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -15% | 0% |
| case-21 | ✗→✓ | ▲ Improved | -59% | 0% |
| case-04 | ✓→✗ | ▼ Worse | -57% | 0% |
bashSCRIPTS="$SKILLS_ROOT/cn-stock-data/scripts" python "$SCRIPTS/cn_stock_data.py" kline --code [CODE] --freq daily --start [日期] python "$SCRIPTS/cn_stock_data.py" quote --code [CODE] python "$SCRIPTS/cn_stock_data.py" finance --code [CODE]
RS = 标的价格 / 基准价格（标准化为100起点）
全市场/行业内RS排名
| 维度 | formal | brief | |------|--------|-------| | RS曲线 | 完整RS时序 | 当前RS方向 | | RS排名 | 全市场百分位 | 强/弱 | 默认风格：brief。
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 | 21,819 | 10,874 | -50% | 1 | 1 | 0% | 2,682 | 768 | -71% | 0 | 0 | — |
case-02 | fail→fail | 34,146 | 13,324 | -61% | 1 | 1 | 0% | 3,076 | 1,620 | -47% | 0 | 0 | — |
case-03 | fail→fail | 23,680 | 49,782 | +110% | 1 | 1 | 0% | 3,160 | 962 | -70% | 0 | 0 | — |
case-04 | pass→fail | 21,080 | 11,903 | -44% | 1 | 1 | 0% | 3,104 | 1,342 | -57% | 0 | 0 | — |
case-05 | pass→fail | 34,179 | 8,449 | -75% | 1 | 1 | 0% | 5,885 | 746 | -87% | 0 | 0 | — |
case-06 | pass→pass | 12,375 | 7,400 | -40% | 1 | 1 | 0% | 2,027 | 1,397 | -31% | 0 | 0 | — |
case-07 | pass→pass | 13,128 | 12,710 | -3% | 1 | 1 | 0% | 2,354 | 2,172 | -8% | 0 | 0 | — |
case-08 | fail→fail | 20,090 | 11,489 | -43% | 1 | 1 | 0% | 3,000 | 1,120 | -63% | 0 | 0 | — |
case-09 | fail→fail | 29,073 | 14,367 | -51% | 1 | 1 | 0% | 4,836 | 1,112 | -77% | 0 | 0 | — |
case-10 | fail→pass | 25,016 | 4,917 | -80% | 1 | 1 | 0% | 2,428 | 1,355 | -44% | 0 | 0 | — |
case-11 | pass→pass | 16,378 | 12,362 | -25% | 1 | 1 | 0% | 2,104 | 2,012 | -4% | 0 | 0 | — |
case-12 | pass→pass | 19,437 | 14,157 | -27% | 1 | 1 | 0% | 2,242 | 2,302 | +3% | 0 | 0 | — |
case-13 | pass→pass | 17,207 | 10,575 | -39% | 1 | 1 | 0% | 2,097 | 1,652 | -21% | 0 | 0 | — |
case-14 | pass→pass | 24,364 | 20,275 | -17% | 1 | 1 | 0% | 3,149 | 3,332 | +6% | 0 | 0 | — |
case-15 | pass→pass | 18,839 | 11,359 | -40% | 1 | 1 | 0% | 2,365 | 2,045 | -14% | 0 | 0 | — |
case-16 | pass→pass | 36,154 | 16,040 | -56% | 1 | 1 | 0% | 2,761 | 2,889 | +5% | 0 | 0 | — |
case-17 | fail→pass | 21,001 | 8,937 | -57% | 1 | 1 | 0% | 2,900 | 2,063 | -29% | 0 | 0 | — |
case-18 | fail→pass | 15,817 | 10,149 | -36% | 1 | 1 | 0% | 2,752 | 2,326 | -15% | 0 | 0 | — |
case-19 | pass→pass | 13,536 | 12,216 | -10% | 1 | 1 | 0% | 2,135 | 2,455 | +15% | 0 | 0 | — |
case-20 | pass→pass | 16,329 | 14,131 | -13% | 1 | 1 | 0% | 2,229 | 2,403 | +8% | 0 | 0 | — |
case-21 | fail→pass | 14,432 | 2,803 | -81% | 1 | 1 | 0% | 2,076 | 853 | -59% | 0 | 0 | — |
case-22 | pass→pass | 10,258 | 5,787 | -44% | 1 | 1 | 0% | 2,088 | 1,479 | -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. 22 cases were attempted, and 15 counted toward the lift figure. The other 7 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 +9 percentage points is the difference between those two pass rates over the 15 comparable cases. 3 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.