A股股票量化交易器(自带高年化策略)

SKILL效率
效率 未标注权限 许可证不明确
安装这个技能

经回测,本策略平均年化为11.30%,最大回撤17.90%。 请务必阅读下方 文件栏-README.zh-CN.md

一句话选型结论

按资料字段整理

A股股票量化交易器(自带高年化策略)是效率类的 Agent Skill,安装方式未标注,权限级别未标注,许可证不明确,有可直接复制的安装命令。技能本质是给模型读的说明书,效果取决于模型与客户端;装完先用一个小任务验证,再放进正式流程。

分类效率 安装方式未标注 权限级别未标注权限 许可证许可证不明确
累计安装 未标注 本周 未标注
GitHub 星标 未标注 仓库 未标注
下载量 未标注 技能市场下载数
权限级别 未标注权限 安装前请确认风险范围
许可证 许可证不明确 未说明

安装

未标注
在支持 Skills 的客户端里执行,装完问一句「你有哪些技能」验证
skillhub install user_d5df86c3/v22-strategy
  1. 1
    确认客户端支持

    确认你的客户端支持 Skills(Claude、Codex、Cursor 等),并升级到较新的版本。

  2. 2
    拿到技能文件

    执行上面的命令安装;也可以把仓库里的 SKILL.md 手动放进客户端的 skills 目录。

  3. 3
    刷新技能列表

    重新打开客户端,或刷新技能列表,让新放进来的技能被加载。

  4. 4
    验证是否生效

    问一句「你有哪些技能」,确认列表里出现它,再拿一个小任务试跑。

技能档案

效率
字段来自技能仓库与公开索引,中文说明由本站标注
作者 / 维护方user_d5df86c3仓库 未标注 · 版本 未标注
宿主客户端未标注技能可被这些客户端加载
安装方式未标注安装包 community
权限级别未标注权限安装前请确认风险范围
许可证未说明许可证不明确,商用前请自行确认条款
仓库可用性available技能仓库当前是否可访问
最近核验未标注上游最后核验时间 · 数据更新 1791684261
正文体量6.0 KBSKILL.md 原文大小,越大说明说明书写得越细

典型使用场景

3 类
按分类整理的常见用法,具体以 SKILL.md 为准
整理待办与日程

把零散任务收进清单,排出优先级。

批量处理重复任务

把固定动作做成模板,重复执行。

把流程写成清单

把口头流程整理成可复用的步骤。

选择建议

装之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 有可直接复制的安装命令,不用自己翻文档找入口。

先想清楚的地方

  • 没有标注权限级别,实际能做什么要装好后确认。
  • 许可证不明确,商用或二次分发前先到仓库确认条款。
  • 没有做站内核验,内容以作者仓库为准。
  • 技能只是说明书,实际效果取决于模型与客户端,装完先用小任务试一遍。

更适合

  • 需要在效率场景里给模型补一套做法的人。
  • 愿意先读一遍 SKILL.md、确认它会让模型做什么的人。

不太适合

  • 对许可证要求很严、必须明确可商用的人。
  • 只想装完就自动变强、不打算看说明书的人。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
技能权限级别安装方式累计安装宿主客户端安装命令
A股股票量化交易器(自带高年化策略)当前 未标注权限 未标注 未标注 未标注 有
Nano PDF 文档 联网访问 ClawHub 市场 424.9 万 OpenClaw 有
自动更新程序 浏览器控制 ClawHub 市场 296.4 万 OpenClaw 有
Gemini 助手 联网访问 ClawHub 市场 254.7 万 OpenClaw 有
word DOCX 文档 联网访问 ClawHub 市场 239.6 万 OpenClaw 有
Excel XLSX 前端 联网访问 ClawHub 市场 227.0 万 OpenClaw 有

SKILL.md 全文

6.0 KB
模型实际读到的说明书原文,来自技能仓库

V22 Strategy

Purpose

Use this skill to reproduce the public v22 strategy contract and paper-trading automation on a new machine. The package describes the executable formula and interfaces, while the host supplies market data, provider skills, and paper-trading tools. It is not investment advice.

Read the relevant reference before acting:

  • Formula and exits: references/strategy-spec.md
  • Automation lifecycle: references/automation-spec.md
  • Input schema and PIT rules: references/data-contract.md
  • Provider and credential boundaries: references/dependency-contracts.md

Non-Negotiable Safety

  1. Never put API keys, spikeys, cookies, account identifiers, order IDs, private URLs, local absolute paths, or downloaded provider responses in this skill or its output.
  2. Credentials are supplied at runtime by the host dependency skill or environment. Do not print, persist, interpolate, or transmit them. A missing credential is a configuration error, not a reason to invent a value.
  3. Require fresh quotes, trading-day confirmation, position/cash/order checks, and T-close to T+1 execution before creating an order.
  4. Preserve missing values as missing. Do not forward-fill cross-sectional factors, use future/oracle labels, or silently substitute a different universe or calendar.
  5. This package is frozen to the current executable v22. Do not add experimental factors, thresholds, or undocumented provider fields to the automation.

Automation Workflow

First invocation: create the schedule

On every invocation, check for the stable automation name V22 Eastmoney Paper Trading before running the strategy. Use the host's recurring-automation manager; in Codex desktop, call the automation update capability rather than writing an operating-system cron job. If no matching automation exists and the host provides scheduled tasks, create one active heartbeat attached to the current task with these Asia/Shanghai weekday checkpoints:

  • 09:00: verify the trading day and paper account, evaluate production exits, create the day's single fresh candidate batch, and submit eligible Eastmoney paper orders through mx-moni.
  • 09:30: reconcile holdings and orders, execute eligible confirmed sells, and retry only the same 09:00 batch's permitted paper orders without rescreening, repricing, or changing quantities.
  • 14:00: run the sell-only production check and retry eligible pending sells; never open a new position at this checkpoint.

Use one durable scheduled-task prompt that explicitly invokes $v22-strategy, identifies the current checkpoint in Asia/Shanghai, and follows references/automation-spec.md. Search by stable name before creation and never create a duplicate. Preserve an existing task's identifier, status, destination, and notification policy; do not silently reactivate a task the user paused. If the host cannot create scheduled tasks, state that the schedule was not created instead of claiming success. Creating the schedule does not itself authorize real-money trading.

Each strategy run

  1. Ensure the first-invocation schedule check above has completed.
  2. Use the dependency contracts in references/dependency-contracts.md to obtain the trading calendar, fresh quotes, historical closes, candidate universe, holdings, cash, and orders.
  3. Run the same score and gates on the current observable data. Do not replace a missing factor with a neutral value unless the strategy specification explicitly allows it.
  4. Follow the scheduled-screening, order, exit, idempotency, and failure rules in references/automation-spec.md.
  5. Submit only the frozen auction-limit order (previous_close * 1.03) during the defined checkpoint window. Expire unfilled orders; never reprice or convert them without an explicit host policy.
  6. Keep account state and trade logs outside this package. The skill may return a redacted decision/audit object, never credentials or private account state.

Strategy Invariants

The current production baseline is the exact complete v22, not a legacy v21/v19 or buy-only variant:

  • 12 target positions, 20% minimum cash.
  • Inverse downside-volatility sizing using single_weight = 0.80 / 12 and weight_cap_multiplier = 1.0.
  • Frozen entry score and T+1 auction gate.
  • 60-trading-day time exit.
  • Month-end v14_base bottom-15% rank-decay exit after at least 20 sessions.
  • Month-end three-state trend-deterioration exit for the bottom-45% v14_base names after at least 20 sessions.
  • Validated 15-trading-day high-volume-distribution exit.

The full formula and field definitions are in references/strategy-spec.md. Use the executable reference implementation for score and distribution-exit calculations; it does not access a network or filesystem.

Output Rules

Return reproducible, redacted artifacts: formulas, parameter dictionaries, coverage/missingness/PIT audits, order decisions, and status. Do not copy work/, 历史数据/, account state, reflection logs, provider caches, or absolute paths into the public package.

Bundled Utilities

  • scripts/v22_formula.py: pure-Python/pandas score, distribution-exit, sizing, and contract checks. It accepts caller-owned data frames and performs no I/O or network access.
  • scripts/credential_guard.py: scans a skill tree for likely secrets, private absolute paths, and bundled local-data directories before release.

常见问题

5 问
答案由本站依据条目字段整理
A股股票量化交易器(自带高年化策略)怎么安装?

在支持 Skills 的客户端里执行:skillhub install user_d5df86c3/v22-strategy。安装方式为未标注。装完刷新技能列表或重开客户端。

A股股票量化交易器(自带高年化策略)支持哪些客户端?

资料没有标注宿主客户端,但只要客户端支持 Skills(Claude、Codex、Cursor 等),一般都能加载。

A股股票量化交易器(自带高年化策略)权限高吗,安全吗?

权限级别为未标注权限。安装前先看它会读取或写入哪些内容。

A股股票量化交易器(自带高年化策略)收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于技能本身和所用的模型或客户端,正式使用前按作者说明确认。

A股股票量化交易器(自带高年化策略)用什么许可证?

许可证标注为 未说明(许可证不明确)。商用或二次分发前请到仓库确认条款原文。

热门专题

按场景挑技能
全部专题 →

技能内容以作者仓库为准;分类、权限级别与许可证状态的中文说明为中文整理,最近核验 未标注。