Theme investing
Skill nexaforgelab/a-share-investment-skill-pack/skills/07_alpha_selection/theme_investing
Open-source A-share investment research skill pack for agent workflows
npx -y skills add nexaforgelab/a-share-investment-skill-pack --skill theme_investingAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
分析政策/产业/资金驱动主题,区分真正受益和概念炒作。
SKILL.md
4.8 KB, ~1.7k tokens by cl100k_base, as published. Nobody here has run it
主题投资分析 Skill
Skill ID
a_share.theme_investing.v1
功能定位
分析政策/产业/资金驱动主题,区分真正受益和概念炒作。
适用场景
主题行情、政策催化、短中期配置。
必填输入
as_of_date: 分析截止日期,必须使用当时已公开可得数据。object_type:stock/industry/portfolio/market/theme/factor。object_id: 股票代码、行业名称、组合ID、主题名称或因子名称。investment_horizon:intraday/1w/1m/3m_6m/1y_3y。benchmark: 沪深300、中证500、中证1000、行业指数或自定义基准。data_bundle: 已加载的数据清单、来源、时间戳和质量评分。
可选输入
portfolio_context: 当前持仓、成本、权重、产品约束。risk_budget: 单票、行业、风格、流动性和回撤约束。previous_thesis: 历史投资假设、上次评级、上次目标价。manual_notes: 研究员补充信息或人工调研纪要。
依赖 Skill
industry_chain_mappingpolicy_impactchip_structure
通用执行流程
- 校验输入参数,确认
as_of_date、对象、投资期限、基准、股票池和组合约束。 - 检查依赖数据是否齐全,并调用或引用数据质量审计结果。
- 按本 Skill 的专业框架进行分析,不跳步,不省略反证。
- 对所有关键判断打分:事实强度、影响程度、可验证性、时效性、组合相关性。
- 生成结构化 JSON、Markdown 报告和审计日志。
- 给出下游建议:进入哪个 workflow、是否需要人工复核、何时再次更新。
本 Skill 专业步骤
- 定义主题边界和产业链映射。
- 按收入相关度、利润弹性、订单确定性、技术壁垒筛选公司。
- 判断主题阶段:萌芽、扩散、主升、拥挤、退潮。
- 输出核心受益、次级受益、概念炒作和风险标的。
判断与打分框架
fact_strength: 事实强度,公告/财报/交易所数据最高,二手新闻和传闻最低。impact_score: 对盈利、估值、现金流、风险偏好、组合风险的影响程度。time_sensitivity: 信息有效期和是否需要盘中处理。verification_score: 是否可被其他数据源交叉验证。portfolio_relevance: 对当前组合、候选池、禁投池的影响程度。
标准输出
同时输出两种形态:
result.json:供下游 Agent / workflow 读取;report.md:供投研人员阅读。
result.json 至少包含:
{
"skill_id": "a_share.theme_investing.v1",
"run_id": "test_run_0001",
"as_of_date": "2026-05-22",
"object_type": "theme",
"object_id": "人工智能",
"conclusion": "待生成,禁止作为投资建议直接执行",
"confidence": 0.0,
"evidence": [],
"counter_evidence": [],
"risks": [],
"action_suggestion": "research_only",
"required_followups": [],
"human_review_required": true,
"data_quality_score": 0.0,
"audit_trail": []
}
报告结构
# 主题投资分析报告
## 1. 一句话结论
## 2. 核心证据
## 3. 关键数据与图表建议
## 4. 正方观点
## 5. 反方观点
## 6. 风险与证伪条件
## 7. 对估值/盈利/组合/交易的影响
## 8. 后续跟踪指标
## 9. 人工复核事项
## 10. 审计日志
强制约束
- 不得把传闻、新闻标题、管理层口头表述当成确定事实;所有关键结论必须标注证据来源与时间戳。
- 不得使用未来函数:财务数据、公告、研报、价格、指数成分必须以
as_of_date当时已公开可得为准。 - 不得绕过人工复核直接给出实盘下单指令;涉及买入、卖出、加仓、减仓、禁投、强制平仓的结论必须输出
human_review_required: true。 - 数据缺失、数据冲突、来源不可靠时,必须降低置信度,并列出不可验证项。
- 对 ST、*ST、停牌、退市整理、重大违法、财务造假、流动性不足标的,必须优先触发风险检查。
- 报告中必须同时给出正方观点、反方观点、证伪条件和后续跟踪指标。
下游流转
- 若输出影响投资评级、仓位或禁投状态,必须传递给
investment_memo、portfolio_risk_monitor和compliance_check。 - 若发现数据异常,必须回流给
data_quality_audit或data_conflict_resolution。 - 若发现投资假设变化,必须回流给
thesis_review。
Gives 0 of the 12 instructions most design systems skills give in ~1.7k tokens
Counted across 528 of the 534 authors here whose files we hold, read 2026-08-06
- create a custom theme if neededin 54 of 528, across 10 files
- read the corresponding theme filein 54 of 528, across 10 files
- ask which theme to applyin 53 of 528, across 9 files
- show the theme showcasein 53 of 528, across 9 files
- maintain visual identity across all slidesin 50 of 528, across 6 files
- apply the specified colors and fontsin 47 of 528, across 3 files
- get explicit confirmationin 45 of 528, across 1 file
- Generate a design system before codingin 19 of 528, across 6 files
- Maintain at least 4.5:1 color contrast ratioin 19 of 528, across 8 files
- Describe component shapes, colors, shadows, and interaction statesin 18 of 528, across 4 files
- Check Python installation and install if missingin 17 of 528, across 4 files
- Default to html-tailwind if stack is unspecifiedin 17 of 528, across 4 files
Said here and by no other author read
- 校验输入参数和数据质量
- 定义主题边界和产业链映射
- 按收入利润等指标筛选公司
- 判断主题所处生命周期阶段
- 区分真实受益标的和概念炒作
- 为关键判断打分
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.