Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Libu Premarket

v1.3.45

礼部侍郎 - A 股盘前作战地图。主力资金流向 + 财务排雷,每天开盘前 3 分钟给你一份实战级选股报告。

0· 219·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for ygbeyond/libu-premarket.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Libu Premarket" (ygbeyond/libu-premarket) from ClawHub.
Skill page: https://clawhub.ai/ygbeyond/libu-premarket
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install libu-premarket

ClawHub CLI

Package manager switcher

npx clawhub@latest install libu-premarket
Security Scan
Capability signals
CryptoCan make purchases
These labels describe what authority the skill may exercise. They are separate from suspicious or malicious moderation verdicts.
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
medium confidence
Purpose & Capability
Name/description (盘前选股 + 财务排雷) match the included code (data fetch, filters, technical indicators). Payment integration (ClawTip) and SM4 handling are coherent with the 'paid/full' features. Unexpected/extra capabilities: the skill reads/writes to the user's ~/.openclaw orders directory (creates/migrates/deletes order JSON files) and attempts to import a custom Tushare client from ~/.openclaw/skills/tushare-finance/scripts — these are not explicitly obvious from the high-level description.
Instruction Scope
SKILL.md tells the agent to run pre_market.py and to read pre_market_data.json (expected). However the docs advertise 'Lite zero-config' while the runtime performs dependency checks (it will prompt to install gmssl and exit if missing) and will interact with the local ClawTip order paths when unlocking payment features. The Cron/run instructions that read and trust local JSON are consistent with the code.
Install Mechanism
There is no remote install script or arbitrary download; the skill ships Python source and a requirements.txt listing common packages. No external arbitrary URLs in install phase. Risk from install mechanism is low.
!
Credentials
Declared env vars are optional (CLAWTIP_SM4_KEY, TUSHARE_TOKEN) and the code uses them only for payment/Tushare. However pre_market.py enforces gmssl as a required dependency (causing exit if missing) even though Lite mode is advertised as 'zero config'. The skill reads environment variables (env:read) and will need CLAWTIP_SM4_KEY to create/verify encrypted payment credentials — granting a payment key to code that writes/reads local order files is sensitive and should only be done for trusted skills.
Persistence & Privilege
always: false (good). The skill writes cache_data in the skill folder and creates order files under ~/.openclaw/skills/orders/{indicator}/ (expected for a local payment integration). It also migrates legacy order files and may delete incompatible JSON files in a legacy path — this file-deletion behavior under ~/.openclaw could remove artifacts you want to keep and is a behavior to be aware of.
What to consider before installing
What to consider before installing: - The core functionality (pre-market report) is coherent with the included code; network access and local file read/write are expected for fetching data and caching. - 'Lite, zero-config' is partly true, but the script enforces packages (gmssl) and will exit if they are not present; expect to run pip install for dependencies even for Lite mode. - Payment flow uses local order files in ~/.openclaw/skills/orders/{indicator}/ and requires you to set CLAWTIP_SM4_KEY to unlock full features. Only set such a key if you trust the skill and its author — the skill will read that env var and use it to decrypt/verify credentials. - The skill will attempt to import a custom Tushare client from ~/.openclaw/skills/tushare-finance/scripts if that path exists. That means if untrusted code is present there it could be executed by this skill. If you do not use a custom client, ensure that path is empty or reviewed. - The payment migration step may delete incompatible JSON files under a legacy orders path for this slug; back up ~/.openclaw/skills/orders if you rely on other payment/order files. - Recommended precautions: review the two Python files (pre_market.py, payment_utils.py) yourself or run the skill in an isolated environment/container, do not export CLAWTIP_SM4_KEY on a machine you do not fully trust, and back up any ~/.openclaw data before first run.

Like a lobster shell, security has layers — review code before you run it.

latestvk974zjtqrc0hw9d14n89n7xprs85fmh1
219downloads
0stars
22versions
Updated 3d ago
v1.3.45
MIT-0

📈 礼部侍郎 - 盘前作战地图

每天早上开盘前,一份属于你的 A 股核心作战地图。 不看无脑涨停,只看真金白银的主力资金流向。


🎁 免费开箱即用 — Lite 模式零门槛

不用注册、不用配置、不用付费。 安装后直接运行,免费获取:

  • 🌍 全球市场概览 — A50 / 美股 / 港股 / 黄金 / 原油隔夜表现,一眼看清开盘情绪
  • 🔥 昨日热点复盘 — 昨日主力净流入 Top 5 板块 + 热门个股回顾

💡 零配置开箱即用:首次运行即可体验。Lite 模式不需要 Tushare Token,不需要支付密钥。


💎 解锁完整版 — AI 选股 + 财务排雷

支付后解锁以下核心能力:

功能说明
🎯 AI 智能选股从 4500+ 股票中,通过「财务增速 + 资金共振 + 技术面」三重过滤,输出 Top 10 潜力标的
财务排雷自动筛除业绩暴雷、负债过高、现金流异常的个股,帮你避开隐形坑
⚔️ 操盘建议基于资金博弈的实战建议:哪些该追、哪些该躲、哪些该观望

💰 付费方案

方案价格适合谁
🔹 单次体验¥0.8/次偶尔看看,按需付费
🔥 月度订阅¥9.9/月每天开盘前自动刷新,最划算

☕ 一杯奶茶钱,换每天 30 天精准选股报告。


🚀 3 分钟快速上手

第 1 步:安装

npx clawhub@latest install libu-premarket

第 2 步:运行(免费 Lite 模式)

python3 pre_market.py

就这么简单。零配置,直接出报告。

⚠️ 首次运行需要安装依赖:pip3 install gmssl

第 3 步:(可选)解锁完整版

想获取 AI 选股和财务排雷?运行脚本后会自动创建支付订单,按提示完成支付即可解锁。首次使用需京东 App 扫码授权钱包(仅一次)。

第 4 步:(可选)环境变量配置

完整版支付:解锁 AI 选股和财务排雷后,需配置京东 ClawTip 支付密钥:

export CLAWTIP_SM4_KEY="你的SM4密钥"

💡 该密钥由京东 ClawTip 开发者后台提供,仅完整版支付功能需要。

Tushare 实时数据(可选):在 Tushare Pro 免费注册获取 Token:

export TUSHARE_TOKEN="你的token"

📊 报告长什么样?

每次运行自动生成 pre_market_data.json,你的 AI Agent 基于数据为你输出:

🌍 外围风向:A50期指 -0.2%,美股三大指数涨跌互现,纳斯达克 +0.5%
   → 开盘情绪偏中性,关注科技股延续性

🔥 今日主线:主力净流入 Top 5 板块
   1. 半导体(+32亿)▎量价齐升,重点关注存储芯片方向
   2. 新能源车(+18亿)▎政策利好持续发酵
   3. ...

🎯 重点关注(Top 5)
   • XX股份(000XXX):净利润增速 45% + 主力连续3日净流入 + MACD金叉
   • XX科技(688XXX):ROE 15.2% + 突破年线压力位
   • ...

⚔️ 操盘建议
   → 进攻方向:半导体存储、新能源车零部件
   → 回避方向:前期涨幅过大的AI应用概念
   → 仓位建议:6-7成,关注早盘前30分钟量能

⚙️ 进阶配置

编辑 config.json 可调整选股标准:

{
  "filter_financial": {
    "min_netprofit_yoy": 0.30,   // 净利润增速 ≥30%
    "min_roe": 8                  // ROE ≥8%
  },
  "cache": {
    "root_dir": "./cache_data"
  }
}

🤖 自动化 Cron 部署(推荐)

如果你想让 Agent 每天自动运行并发送报告,请在 OpenClaw Cron 任务中按以下 三步串行 逻辑配置,严禁并行执行

  1. 执行脚本:使用 exec 工具运行 python3 pre_market.py
  2. 验证数据:读取生成的 pre_market_data.json,检查 date 字段是否为今天。如果不是,说明脚本未更新成功,需重试。
  3. 生成报告:数据验证通过后,再让 Agent 读取 JSON 撰写报告。

推荐 Prompt 模板

"请先运行 python3 pre_market.py,读取 pre_market_data.json 确认 date 为今天。确认数据无误后,基于 JSON 内容撰写盘前报告并发送给我。"

避坑指南:不要在 Cron 中让 Agent 同时读取 JSON,这会导致 Agent 读取到昨天的旧缓存。

🔄 数据更新

  • 首次运行:自动从网络获取最新市场数据(需联网)
  • 本地缓存:运行产生的数据保存在 ./cache_data/,下次优先读取,秒级出报告
  • 定期更新:开发者持续优化算法,clawhub update libu-premarket 即可获取最新版

⚠️ 注意事项

  • 本工具仅供数据参考,不构成任何投资建议。股市有风险,入市需谨慎。
  • 运行环境:Python 3.8+,依赖 pandasrequestsnumpygmssl
  • Lite 模式(免费)无需任何密钥配置
  • 完整版需配置 CLAWTIP_SM4_KEY(京东 ClawTip 支付密钥),支持单次 / 月度订阅

🆘 售后支持

方式说明
微信添加 ygbeyond 进专属交流群
Bug 修复24 小时内响应
数据更新群公告 / 朋友圈第一时间通知

📝 更新日志

v1.3.45 — 数据焕新

  • 📦 数据包更新:日线数据更新至 2026-04-24,收盘数据同步更新
  • 🔄 脚本同步:pre_market.py 与 config.json 从内部开发版本同步

v1.3.42 — 安全扫描修复

  • 🔧 manifest.json:新增 optional_env_vars,声明 CLAWTIP_SM4_KEYTUSHARE_TOKEN
  • 📖 SKILL.md:补充环境变量说明,消除「零配置」与实际代码的不一致
  • 📦 数据更新流程优化:移除对不存在的 initial_data.zip 的文档引用,改为网络获取

v13.4.0 — Lite 模式 + 核心安全重构

  • 🎁 新增 Lite 模式:无需配置密钥,免费运行获取全球市场概览与昨日复盘
  • 🔐 支付系统安全重构:SM4 密钥改为纯环境变量,代码零硬编码
  • 首次运行体验优化:零门槛开箱即用,未解锁完整版时引导更友好
  • 📖 文档全面更新:快速开始流程简化,报告示例更直观

v13.3.0 — 零配置体验

  • 开箱即用:安装即运行,无需环境变量
  • 先用后付:未支付自动进入 Lite 模式

v13.2.0 — 安全加固

  • 修复环境变量绕过漏洞

Comments

Loading comments...