← 返回 Skills 市场
188
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install a-stock-picker
功能描述
A股三层选股模型 — 量化筛选 → 定性分析 → 择时操作。用于主动选股、持仓诊断、机会扫描。当用户说「选股」「帮我看看有哪些好股」「扫描市场」「符合什么条件才能买」「筛选强势股」时使用。
使用说明 (SKILL.md)
A股三层选股模型
基于「量化筛选 → 定性分析 → 择时操作」三层漏斗模型,从全市场5000+只A股中筛选出最优标的。
工作流程
第一层:量化筛选(机选)
读取 references/screening-rules.md,按以下5维度等权重筛选:
| 维度 | 条件 | 权重 |
|---|---|---|
| 市值 | 50~500亿 | 20% |
| 动量 | MA15 > MA60 | 20% |
| 换手 | 近5日均换手率 > 2% | 20% |
| MACD | MACD金叉且柱状体放大 | 20% |
| 位置 | 股价在筹码低位或突破密集区 | 20% |
过滤规则: 满足 ≥3/5 项 → 进入第二层
数据获取: 使用 AkShare 接口(参考 references/akshare-guide.md)
第二层:定性分析(人选)
对通过第一层的标的,逐只进行定性验证:
- 基本面:净利润增速 > 15%、ROE > 12%、无明显财务恶化
- 行业:属于政策支持主线(AI/半导体/军工/新能源/消费)
- 催化:近期有实质利好(订单/政策/业绩超预期/重组)
- 筹码:主力筹码集中,上方套牢盘不重
- 龙头:细分行业龙头或技术领先优势
过滤规则: 满足 ≥4/5 项 → 进入第三层
第三层:择时操作
对通过第二层的标的,给出具体操作方案:
- 买入区间(精确价格)
- 止损价(结构失效位)
- 第一目标价
- 仓位建议
- 风险等级(低/中/高)
执行顺序
- 读取
references/screening-rules.md— 获取量化筛选参数 - 读取
references/analysis-framework.md— 获取定性分析框架 - 运行
scripts/screen.py— 执行全市场扫描(输出初筛名单) - 对初筛名单逐只做定性验证
- 输出最终买入名单(含操作方案)
输出格式
## 初筛结果
[满足条件的标的列表]
## 定性验证
[每只标的的5维度评分]
## 买入名单(含操作方案)
| 股票 | 代码 | 行业 | 买入区间 | 止损 | 目标 | 仓位 | 风险 |
|------|------|------|---------|------|------|------|------|
数据来源
- 实时行情:
stock_zh_a_spot_em(东方财富) - 日线历史:
stock_zh_a_hist(AkShare) - 财务数据:
stock_financial_analysis_indicator(如果能获取) - 资金流:
stock_individual_fund_flow(如果能获取) - 筹码分布:
stock_cyq_em(如果能获取)
详细接口用法见 references/akshare-guide.md
重要原则
- 量化初筛不选市值<50亿或>500亿的标的(流动性风险/壳价值)
- 只选有催化逻辑的标的,禁止盲目选"超跌"
- 择时方案必须带止损位,没有止损位的标的直接排除
- 每次选股输出不超过5只标的(聚焦)
- 如果市场整体趋势向下(如MA60空头排列),主动提示降低仓位
安全使用建议
This skill appears to be a straightforward stock screener and is internally consistent overall, but review a few items before running: (1) the SKILL.md mentions an AkShare guide file that is missing — verify whether you need AkShare or the included script is sufficient; (2) the docs have small inconsistencies in market‑cap thresholds across files — confirm the numeric policy you want; (3) the Python script performs unauthenticated HTTP requests to public Sina/Tencent endpoints (your IP and request metadata will be visible to those services); and (4) as with any trading tool, test on historical data or in a sandbox before acting on live money. If you want higher assurance, ask the author for the missing akshare guide and for clarification of the intended market‑cap rules, and inspect/run the script in an isolated environment first.
功能分析
Type: OpenClaw Skill
Name: a-stock-picker
Version: 1.0.0
The skill bundle provides a legitimate framework for A-share stock screening and analysis using a three-layer model. The Python script `scripts/screen.py` fetches market data from public financial APIs (Sina and Tencent) using standard libraries and implements basic quantitative scoring logic. The instructions in `SKILL.md` and the reference documents are well-structured and align strictly with the stated purpose of stock picking without any evidence of malicious intent, data exfiltration, or prompt injection.
能力评估
Purpose & Capability
The name/description (A‑share three‑layer stock picker) match the included artifacts: two reference docs describing quantitative and qualitative rules and a Python script that implements a layer‑1 screener. No unrelated credentials, binaries, or install steps are requested. The network calls to public market data providers (Sina / Tencent endpoints) are expected for this purpose.
Instruction Scope
Runtime instructions stay within the stated purpose (read the reference rules, run scripts/screen.py, perform qualitative checks). Minor issues: SKILL.md references using AkShare and an `references/akshare-guide.md` that is not present in the package, and some thresholds/market‑cap bounds differ between files (e.g., screening-rules.md mentions >800亿 exclusion while SKILL.md/script use 500亿). The script performs direct HTTP GETs to public endpoints (expected) and does not read unrelated local files or environment variables.
Install Mechanism
No install spec; the skill is instruction‑only with an included script. No remote downloads or archive extraction are requested. The bundled Python script will execute locally if the agent runs it — there is no automatic installation of third‑party packages in the skill metadata.
Credentials
The skill declares no required environment variables, credentials, or config paths. The script makes unauthenticated calls to public APIs only. No secrets or unrelated service tokens are requested.
Persistence & Privilege
The skill does not request permanent presence (always: false) and does not modify other skills or system configs. It uses normal agent invocation and has no privileged persistence behavior.
如何使用
- 确保已安装 OpenClaw(本地或 Docker 部署)
- 在对话框中输入安装命令:
/install a-stock-picker - 安装完成后,直接呼叫该 Skill 的名称或使用
/a-stock-picker触发 - 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
首发:量化筛选+定性分析+择时
元数据
常见问题
A股三层选股模型 是什么?
A股三层选股模型 — 量化筛选 → 定性分析 → 择时操作。用于主动选股、持仓诊断、机会扫描。当用户说「选股」「帮我看看有哪些好股」「扫描市场」「符合什么条件才能买」「筛选强势股」时使用。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 188 次。
如何安装 A股三层选股模型?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install a-stock-picker」即可一键安装,无需额外配置。
A股三层选股模型 是免费的吗?
是的,A股三层选股模型 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
A股三层选股模型 支持哪些平台?
A股三层选股模型 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。
谁开发了 A股三层选股模型?
由 danpian1(@danpian1)开发并维护,当前版本 v1.0.0。
推荐 Skills