← Back to Skills Marketplace
54lynnn

Stock Watcher Pro

by 54Lynnn · GitHub ↗ · v2.0.10 · MIT-0
cross-platform ⚠ suspicious
104
Downloads
0
Stars
0
Active Installs
11
Versions
Install in OpenClaw
/install stock-tracker
Description
东方财富自选股公告追踪。三级过滤 + LLM分类摘要 + Web仪表盘,支持 agent 定时推送。
README (SKILL.md)

📊 Stock Watcher - 东方财富自选股公告追踪

双模式技能 — Agent 定时自动追踪 + 用户 Web 仪表盘查看。

告诉 agent 你想做什么,例如:

"帮我看看持仓板块这两天有什么重要公告" "每天8点提醒我有价值的自选股公告" "帮我打开公告仪表盘看看"

模式一:Agent Run 模式

Agent 定时运行 run.sh,自动抓取公告、生成摘要,输出有价值公告的 digest 转发给用户。

告诉 agent 即可:

"帮我设置 stock-watcher 定时任务,每天早上8点运行一次,追踪【xx】板块的最新公告,有重要公告通知我"

或者手动运行:

# 用法: bash run.sh [group] [days] [source]
bash run.sh mygroup 15 eastmoney

输出示例

有公告时:

DIGEST_TOTAL:3
1.
【000001平安银行】-【平安银行2026年第一季度报告】
【季度报告】...

2.
【600519贵州茅台】-【贵州茅台关于回购股份的进展公告】
累计回购金额XX亿元...

无公告时:

DIGEST_EMPTY:最近1天test板块无高价值公告

模式二:Dashboard 模式

用户手动运行,抓取公告 + 生成摘要 + 启动 Web 仪表盘,浏览器查看所有公告详情。

告诉 agent 即可:

"帮我打开公告仪表盘,看看 mygroup 板块最近15天的情况"

或者手动运行:

# 用法: bash dashboard.sh [group] [days] [source]
bash dashboard.sh mygroup 15 eastmoney

脚本自动执行三步:

  1. stock_watcher.py --fetch-content — 抓取公告 + 全文 + LLM 分类
  2. daily_summary.py — 生成摘要
  3. dashboard.py — 启动 Flask 仪表盘(默认端口 5001)

启动后浏览器访问 http://localhost:5001,按 Ctrl+C 停止。

仪表盘功能:

  • 股票列表表格:7天/15天/30天/全部的有价值公告比例
  • 点击展开:懒加载公告详情(标题、日期、摘要、正文、原文链接)
  • 类型标签:大类 / 小类(如 股权股本类 / 回购
  • 搜索过滤,响应式设计

首次设置

告诉 agent 即可:

"帮我初始化 stock-watcher,安装依赖、配置 Cookie 和 LLM"

或者按以下步骤手动设置:

1. 安装依赖

cd ~/.openclaw/workspace/skills/stock-watcher
pip install requests pdfplumber flask

2. 配置 Cookie

东方财富网页版登录后的 Cookie,用于获取自选股列表。

  1. 浏览器打开 https://quote.eastmoney.com/zixuan/lite.html 并登录
  2. F12 → Console → 输入 copy(document.cookie) 回车
  3. 粘贴到 cookie.txt

自动续签(可选):

pip install playwright && playwright install chromium
python3 scripts/refresh_cookie.py

3. 配置 LLM(可选)

LLM 用于标题价值判断、公告分类和摘要生成。未配置时仅使用正则过滤。

# .env 文件
echo "LLM_API_KEY=sk-your-api-key" > .env
// config.json
{
  "llm": {
    "enabled": true,
    "base_url": "https://opencode.ai/zen/go/v1",
    "model": "deepseek-v4-flash",
    "timeout": 60,
    "retries": 2
  }
}

4. 验证运行

python3 scripts/stock_watcher.py --group test --days 15 --fetch-content
python3 scripts/stock_watcher.py --stats

5. 配置定时任务

告诉 agent 即可,例如:

"帮我设置 stock-watcher 定时任务,每天早上8点运行一次,追踪【xx】板块的最新公告,有重要公告通知我"

或者手动配置:

openclaw cron add \
  --name stock-watcher \
  --cron "0 1 * * *" \
  --message "运行股票公告扫描:cd {{SKILL_DIR}} && bash run.sh mygroup 15 eastmoney"

完整技术文档、公告分类体系、Token 消耗分析等详见 GitHub

开源

Usage Guidance
Review this skill before installing. Use a dedicated low-risk Eastmoney session if possible, protect cookie.txt like a password, keep .env and the SQLite state out of shared workspaces, and only enable the LLM if you are comfortable sending watchlist-related announcement titles/text to the configured provider. Avoid running scripts/setup.sh unless you intend to install recurring cron jobs, and bind or firewall the dashboard if you do not want LAN access.
Capability Tags
requires-sensitive-credentials
Capability Assessment
Purpose & Capability
Fetching Eastmoney watchlists, announcement content, LLM classification/summarization, local SQLite storage, and a dashboard all fit the stated purpose, but the same purpose uses sensitive session cookies and can transmit watchlist-derived announcement data to a configured LLM endpoint.
Instruction Scope
The docs use broad 'tell the agent' setup and scheduling examples around dependency installation, cookie configuration, LLM setup, dashboard launch, and cron creation, without consistently separating read-only actions from credential or persistence changes.
Install Mechanism
The OpenClaw install metadata only installs standard Python dependencies, but the included setup script installs packages, creates local state/config files, and modifies crontab if a user or agent runs it.
Credentials
Network access to Eastmoney/Cninfo and optional LLM endpoints is purpose-aligned, but the dashboard binds to 0.0.0.0 while documentation presents localhost access, exposing portfolio-derived data on reachable interfaces without authentication.
Persistence & Privilege
The skill stores a raw Eastmoney browser cookie in cookie.txt, writes an LLM key template to .env, stores announcement data in a local SQLite database, refreshes cookies with Playwright, and includes a setup script that silently adds recurring cron jobs.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install stock-tracker
  3. After installation, invoke the skill by name or use /stock-tracker
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v2.0.10
fix: use absolute file:/// paths for references links in README
v2.0.9
chore: sync directory name with GitHub/ClawHub slug
v2.0.8
chore: remove HANDOVER.md from git tracking
v2.0.7
refactor: simplify README.md to ~150 lines, move technical details to references/
v2.0.6
docs: add 开源 section with GitHub link
v2.0.5
refactor: remove 依赖 and 注意事项 sections (belong in README)
v2.0.4
docs: add natural language example for first-time setup
v2.0.3
refactor: move natural language example to top of Agent Run mode
v2.0.2
docs: add natural language usage examples at top of SKILL.md
v2.0.1
docs: add cross-references between GitHub and ClawHub
v2.0.0
v2.0.0: 重构为双模式技能(run/dashboard),优化 digest 输出
Metadata
Slug stock-tracker
Version 2.0.10
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 11
Frequently Asked Questions

What is Stock Watcher Pro?

东方财富自选股公告追踪。三级过滤 + LLM分类摘要 + Web仪表盘,支持 agent 定时推送。 It is an AI Agent Skill for Claude Code / OpenClaw, with 104 downloads so far.

How do I install Stock Watcher Pro?

Run "/install stock-tracker" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Stock Watcher Pro free?

Yes, Stock Watcher Pro is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Stock Watcher Pro support?

Stock Watcher Pro is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Stock Watcher Pro?

It is built and maintained by 54Lynnn (@54lynnn); the current version is v2.0.10.

💬 Comments