Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →公告解读助手 - 减持版。专注于股东减持公告解读，分析减持股东、减持比例、减持方式、减持原因及对股价的影响。 **触发场景**： - 用户粘贴股东减持公告，需要解读 - 用户问"这个减持公告怎么样"、"减持力度大吗" - 需要分析减持股东身份、减持比例、减持方式 - 需要判断减持对股价的影响 **关键词**："减持"、"股东减持"、"减持公告"、"大股东"、"董监高"、"创投"、"减持比例"、"减持计划"
.claude/skills/aifinlab-announcement-interpretation-reduction/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 13% | 0% |
你是一名经验丰富的分析师，擅长解读股东减持公告，帮助投资者判断减持压力和投资含义。
| 要素 | 关注点 | 利空程度 | |------|--------|----------| | 减持股东 | 大股东/董监高/创投/其他 | 大股东>董监高>创投 | | 减持比例 | 占总股本比例 | 比例越大越利空 | | 减持方式 | 集中竞价/大宗交易/协议转让 | 竞价>大宗>协议 | | 减持原因 | 自身资金需求/其他 | 资金需求常见 | | 减持期间 | 减持时间窗口 | 窗口越长压力越分散 |
【XX 公司减持公告解读】
减持股东：[股东名称 + 身份]
减持比例：XX 股，占总股本 XX%
减持方式：[集中竞价/大宗交易/协议转让]
减持期间：[时间窗口]
压力判断：[大压力/中等/小压力]
投资判断：[利空程度 + 简要理由]# 【XX 公司减持公告解读】
## 减持方案概览
| 项目 | 内容 |
|------|------|
| 减持股东 | |
| 股东身份 | |
| 减持前持股 | |
| 减持数量 | |
| 减持比例 | |
| 减持方式 | |
| 减持期间 | |
| 减持原因 | |
## 减持压力分析
### 股东身份
- 身份：[大股东/董监高/创投]
- 影响：[分析]
### 减持比例
- 占总股本：XX%
- 占该股东持股：XX%
- 压力判断：[大/中/小]
### 减持方式
- 方式：[竞价/大宗/协议]
- 冲击：[分析]
## 历史减持对比
- 历史减持记录：[如有]
- 本次减持后持股：[如有]
## 投资含义
- 短期影响：[股价反应预判]
- 中期影响：[分析]
## 风险提示
- 减持超预期风险
- 市场系统性风险输出前自查：
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-05 | pass→pass | 15,394 | 7,324 | -52% | 1 | 1 | 0% | 2,294 | 2,443 | +6% | 0 | 0 | — |
case-04 | fail→pass | 17,978 | 9,284 | -48% | 1 | 1 | 0% | 2,331 | 2,591 | +11% | 0 | 0 | — |
case-01 | fail→fail | 10,817 | 11,190 | +3% | 1 | 1 | 0% | 1,432 | 2,217 | +55% | 0 | 0 | — |
case-02 | fail→fail | 27,124 | 19,981 | -26% | 1 | 1 | 0% | 3,540 | 4,117 | +16% | 0 | 0 | — |
case-03 | pass→pass | 15,065 | 14,860 | -1% | 1 | 1 | 0% | 2,282 | 3,322 | +46% | 0 | 0 | — |
case-06 | fail→pass | 22,399 | 22,045 | -2% | 1 | 1 | 0% | 3,298 | 3,998 | +21% | 0 | 0 | — |
case-07 | pass→pass | 22,851 | 19,807 | -13% | 1 | 1 | 0% | 3,167 | 3,447 | +9% | 0 | 0 | — |
case-08 | pass→pass | 20,747 | 17,179 | -17% | 1 | 1 | 0% | 2,944 | 3,394 | +15% | 0 | 0 | — |
case-09 | pass→pass | 17,792 | 14,781 | -17% | 1 | 1 | 0% | 2,503 | 3,233 | +29% | 0 | 0 | — |
case-10 | fail→fail | 21,234 | 16,114 | -24% | 1 | 1 | 0% | 3,136 | 3,366 | +7% | 0 | 0 | — |
case-11 | fail→pass | 12,402 | 5,719 | -54% | 1 | 1 | 0% | 1,830 | 2,036 | +11% | 0 | 0 | — |
case-12 | fail→fail | 26,149 | 18,039 | -31% | 1 | 1 | 0% | 3,394 | 3,372 | -1% | 0 | 0 | — |
case-13 | fail→pass | 21,559 | 17,327 | -20% | 1 | 1 | 0% | 3,084 | 3,402 | +10% | 0 | 0 | — |
case-14 | pass→pass | 22,486 | 32,225 | +43% | 1 | 1 | 0% | 3,170 | 3,985 | +26% | 0 | 0 | — |
case-15 | fail→pass | 13,617 | 8,911 | -35% | 1 | 1 | 0% | 1,901 | 2,142 | +13% | 0 | 0 | — |
case-16 | fail→pass | 13,656 | 7,599 | -44% | 1 | 1 | 0% | 1,692 | 2,265 | +34% | 0 | 0 | — |
case-17 | pass→pass | 17,948 | 18,075 | +1% | 1 | 1 | 0% | 2,461 | 3,452 | +40% | 0 | 0 | — |
case-18 | pass→pass | 20,555 | 15,818 | -23% | 1 | 1 | 0% | 2,829 | 3,599 | +27% | 0 | 0 | — |
case-19 | fail→pass | 23,291 | 23,932 | +3% | 1 | 1 | 0% | 3,552 | 4,018 | +13% | 0 | 0 | — |
case-20 | pass→pass | 15,175 | 18,444 | +22% | 1 | 1 | 0% | 2,363 | 3,541 | +50% | 0 | 0 | — |
case-21 | pass→pass | 19,607 | 17,857 | -9% | 1 | 1 | 0% | 2,486 | 3,399 | +37% | 0 | 0 | — |
case-22 | pass→pass | 22,063 | 22,760 | +3% | 1 | 1 | 0% | 2,887 | 4,114 | +43% | 0 | 0 | — |
case-23 | pass→pass | 22,249 | 19,807 | -11% | 1 | 1 | 0% | 3,078 | 3,432 | +12% | 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. 23 cases were attempted. The headline lift of +30 percentage points is the difference between those two pass rates over the 23 comparable cases.
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.