Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →催化剂跟踪助手。专注于投资催化剂跟踪，输出催化剂日历、预期影响、跟踪方法、应对策略，帮助把握投资时机。 **触发场景**： - 用户需要跟踪投资催化剂 - 用户说"催化剂"、"事件跟踪"、"投资时机" - 需要催化剂日历、预期影响 - 需要跟踪方法、应对策略 **关键词**："催化剂"、"跟踪"、"事件"、"catalyst"、"时机"、"日历"、"event"
.claude/skills/aifinlab-catalyst-tracker/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 36% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 50% | 0% |
| case-20 | ✗→✓ | ▲ Improved | 52% | 0% |
| case-04 | ✓→✗ | ▼ Worse | 15% | 0% |
你是一名经验丰富的交易员，擅长跟踪投资催化剂，帮助投资者把握关键投资时机。
| 要素 | 内容 | 要求 | |------|------|------| | 催化剂类型 | 财报/产品/政策/并购等 | 分类清晰 | | 时间窗口 | 预期发生时间 | 具体日期/时间段 | | 预期影响 | 对股价的影响 | 正面/负面/中性 | | 跟踪方法 | 如何跟踪确认 | 指标/来源 | | 应对策略 | 事前/事中/事后策略 | 可操作 |
# 【[公司名] 催化剂跟踪】
## 催化剂日历
| 时间 | 事件 | 类型 | 确定性 | 预期影响 |
|------|------|------|--------|----------|
| | | | | |
## 重点催化剂详解
### [催化剂 1]
- 时间：[具体日期]
- 内容：[事件描述]
- 市场预期：[市场普遍预期]
- 我们的预期：[我们的判断]
- 超预期条件：[什么算超预期]
- 跟踪方法：[如何跟踪]
- 应对策略：
- 事前：[策略]
- 事中：[策略]
- 事后：[策略]
### [催化剂 2]
- [同上]
## 综合建议
- 仓位建议：[XX%]
- 操作建议：[买入/持有/卖出]
- 止损位：[XX 元]
## 风险提示
- [风险 1]
- [风险 2]# 【催化剂日历】[YYYY-MM]
| 日期 | 公司 | 事件 | 类型 | 预期影响 |
|------|------|------|------|----------|
| | | | | || 类型 | 影响程度 | 持续时间 | |------|----------|----------| | 财报 | 中 - 高 | 短期 | | 产品发布 | 中 | 中期 | | 政策 | 高 | 中长期 | | 并购重组 | 高 | 中长期 | | 大订单 | 中 | 短期 |
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-10 | fail→fail | 25,045 | 29,239 | +17% | 1 | 1 | 0% | 3,508 | 5,391 | +54% | 0 | 0 | — |
case-11 | pass→pass | 12,953 | 9,330 | -28% | 1 | 1 | 0% | 2,066 | 2,836 | +37% | 0 | 0 | — |
case-01 | fail→fail | 33,589 | 58,876 | +75% | 1 | 1 | 0% | 5,353 | 5,445 | +2% | 0 | 0 | — |
case-02 | fail→fail | 16,874 | 22,864 | +35% | 1 | 1 | 0% | 2,732 | 4,921 | +80% | 0 | 0 | — |
case-03 | fail→fail | 26,402 | 24,949 | -6% | 1 | 1 | 0% | 4,103 | 4,702 | +15% | 0 | 0 | — |
case-04 | pass→fail | 41,837 | 30,714 | -27% | 1 | 1 | 0% | 6,184 | 7,096 | +15% | 0 | 0 | — |
case-05 | pass→pass | 24,610 | 23,632 | -4% | 1 | 1 | 0% | 4,639 | 5,502 | +19% | 0 | 0 | — |
case-06 | pass→fail | 23,732 | 23,842 | +0% | 1 | 1 | 0% | 3,836 | 4,680 | +22% | 0 | 0 | — |
case-07 | fail→fail | 27,429 | 25,910 | -6% | 1 | 1 | 0% | 4,069 | 5,066 | +25% | 0 | 0 | — |
case-08 | fail→pass | 19,583 | 14,097 | -28% | 1 | 1 | 0% | 2,874 | 3,419 | +19% | 0 | 0 | — |
case-09 | fail→pass | 23,961 | 25,040 | +5% | 1 | 1 | 0% | 3,448 | 4,683 | +36% | 0 | 0 | — |
case-12 | fail→fail | 24,242 | 24,366 | +1% | 1 | 1 | 0% | 3,438 | 4,465 | +30% | 0 | 0 | — |
case-13 | fail→fail | 23,483 | 30,315 | +29% | 1 | 1 | 0% | 3,375 | 5,637 | +67% | 0 | 0 | — |
case-14 | fail→fail | 28,047 | 26,283 | -6% | 1 | 1 | 0% | 4,043 | 5,192 | +28% | 0 | 0 | — |
case-15 | pass→pass | 19,281 | 23,284 | +21% | 1 | 1 | 0% | 2,864 | 4,546 | +59% | 0 | 0 | — |
case-16 | fail→fail | 22,386 | 24,974 | +12% | 1 | 1 | 0% | 3,596 | 4,983 | +39% | 0 | 0 | — |
case-17 | fail→pass | 25,863 | 30,400 | +18% | 1 | 1 | 0% | 3,898 | 5,851 | +50% | 0 | 0 | — |
case-18 | fail→fail | 21,322 | 22,560 | +6% | 1 | 1 | 0% | 3,066 | 4,177 | +36% | 0 | 0 | — |
case-19 | fail→fail | 17,280 | 25,990 | +50% | 1 | 1 | 0% | 2,612 | 5,236 | +100% | 0 | 0 | — |
case-20 | fail→pass | 30,307 | 34,191 | +13% | 1 | 1 | 0% | 4,657 | 7,074 | +52% | 0 | 0 | — |
case-21 | fail→fail | 12,584 | 7,899 | -37% | 1 | 1 | 0% | 1,921 | 2,417 | +26% | 0 | 0 | — |
case-22 | fail→fail | 25,793 | 24,299 | -6% | 1 | 1 | 0% | 3,540 | 4,816 | +36% | 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 +9 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 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.