← 返回 Skills 市场
291
总下载
0
收藏
1
当前安装
1
版本数
在 OpenClaw 中安装
/install 12-v2-2
功能描述
AI量化交易系统,8因子信号系统、多数据源容错、实时行情分析
使用说明 (SKILL.md)
AI量化交易技能 v2.2
完整的量化交易系统,提供股票分析、信号生成、回测和实时行情监控。
技能描述
V2.2完整版量化交易系统,提供8因子信号系统、多数据源容错(AKShare+Efinance)、实时行情分析(量比/换手率)、筹码分布分析、大盘环境判断和邮件推送通知。
使用场景
- 用户:"分析贵州茅台" → 返回完整的技术分析报告
- 用户:"今天有哪些买入信号?" → 扫描市场返回符合条件的股票
- 用户:"回测这个策略" → 运行回测系统返回收益报告
- 用户:"大盘环境如何?" → 分析市场环境判断强势/震荡/弱势
工具和依赖
工具列表
位于 core/ 和 modules/ 目录:
core/分析流水线.py:完整分析流程modules/01-多数据源管理器.py:AKShare+Efinance容错modules/02-大盘环境分析器.py:市场环境判断modules/03-市场增强分析器.py:恐慌贪婪指数modules/04-实时行情分析器.py:量比/换手率/量价关系modules/05-筹码分布分析器.py:集中度/获利盘/形态识别modules/06-邮件通知器.py:SMTP邮件推送modules/07-通知中心.py:多渠道统一接口
API密钥
可选(邮件通知):
- SMTP邮箱配置
外部依赖
- Python 3.7+
- akshare
- efinance
- pandas
- numpy
- smtplib(邮件通知,可选)
配置说明
安装方法
cd 12-量化交易V2.2完整版
bash 一键部署脚本.sh
配置文件
编辑 config/V2.2配置文件.yaml:
# 数据源配置
data_source:
primary: akshare
fallback: efinance
# 信号阈值
thresholds:
strong_market: 8.5
normal_market: 9.0
weak_market: 9.5
# 邮件通知
email:
enabled: true
smtp_server: "smtp.qq.com"
smtp_port: 587
sender: "[email protected]"
password: "your-password"
使用示例
场景1:分析单只股票
用户:"分析贵州茅台"
AI:
from core.分析流水线 import analyze_stock
result = analyze_stock("600519")
# 返回:
# - 8因子信号评分 (0-10分)
# - 技术指标分析
# - 实时行情分析
# - 筹码分布分析
# - 买入/卖出建议
场景2:市场扫描
用户:"今天有哪些买入信号?"
AI:
from modules.市场扫描器 import scan_market
signals = scan_market(market="A股", threshold=9.0)
# 返回:信号评分 >= 9.0 的股票列表
# 包含股票代码、名称、信号评分、建议
场景3:大盘环境判断
用户:"大盘环境如何?"
AI:
from modules.大盘环境分析器 import MarketAnalyzer
analyzer = MarketAnalyzer()
env = analyzer.analyze_market()
# 返回:
# - 市场环境:强势/震荡/弱势
# - 涨跌家数统计
# - 指数分析
# - 动态阈值建议
场景4:运行回测
用户:"回测这个策略"
AI:
from tests.完整回测系统 import run_backtest
results = run_backtest(
start_date="2023-01-01",
end_date="2024-12-31",
initial_capital=100000
)
# 返回:
# - 总收益率
# - 年化收益率
# - 最大回撤
# - 夏普比率
# - 交易明细
场景5:实时监控
用户:"监控我的自选股"
AI:
from modules.实时监控器 import monitor_stocks
stocks = ["600519", "000858", "002475"]
monitor_stocks(stocks, interval=300)
# 每5分钟检查一次
# 有信号时发送邮件通知
8因子信号系统
因子权重分配:
• 动量指标 (Momentum) - 25%
• 均线偏离 (MA Deviation) - 20%
• 成交量 (Volume) - 15%
• 波动率 (Volatility) - 10%
• MACD - 10%
• 布林带 (Bollinger) - 10%
• 实时量比 (Volume Ratio) - 5% [V2.2新增]
• 筹码集中度 (Chip) - 5% [V2.2新增]
动态买入阈值
强势市场: 信号 ≥ 8.5分
震荡市场: 信号 ≥ 9.0分
弱势市场: 信号 ≥ 9.5分
系统特色
1. 多数据源容错
- 主数据源:AKShare
- 备用源:Efinance
- 自动故障切换
- 熔断机制(3次失败=5分钟冷却)
2. 实时行情分析
- 量比5级分析(极度缩量~巨量)
- 换手率5级分析(死寂~高换手)
- 量价关系分析
- 信号评分(0-10分)
3. 筹码分布分析
- 价格分布计算
- 90%/70%集中度
- 获利盘比例
- 形态识别(单峰/双峰/多峰/吸筹/派发)
- 支撑压力位
4. 大盘环境判断
- 市场环境分类(强势/震荡/弱势)
- 动态阈值调整
- 涨跌家数统计
- 指数分析
5. 邮件推送通知
- 买入信号推送
- 卖出信号推送
- 风险警报推送
- 每日汇总报告
6. 通知中心
- 多渠道整合(邮件/微信/飞书/Telegram/钉钉)
- 优先级管理(低/普通/高/紧急)
- 失败重试机制
- 消息队列
- 频率控制
故障排除
问题1:数据获取失败
解决:
- 检查网络连接
- 系统会自动切换到备用数据源
- 如仍失败,等待5分钟后重试(熔断机制)
问题2:邮件通知失败
解决:
- 检查SMTP配置
- 确认邮箱授权码正确
- 检查网络连接
- 查看日志获取详细错误
问题3:回测结果异常
解决:
- 检查数据完整性
- 确认日期范围正确
- 验证初始资金设置
- 查看交易明细定位问题
文档阅读顺序
新手入门
docs/V2.2完整部署与使用指南.md- 了解如何部署和使用docs/V2.2快速实施指南.md- 了解分阶段实施方案tests/系统测试脚本.py- 运行测试验证系统tests/完整回测系统.py- 运行回测查看效果
进阶使用
docs/V2.2集成架构设计.md- 深入了解系统架构core/分析流水线.py- 学习如何使用分析流水线modules/- 各个模块的源代码
注意事项
- 投资风险:本系统仅供参考,不构成投资建议
- 数据延迟:可能有15-20分钟延迟
- 零成本运行:使用免费数据源,无需API费用
- 邮件配置:邮件通知为可选功能
- 合规使用:遵守相关法律法规,不得用于非法用途
安全使用建议
This repo looks like a legitimate quant trading package, but take these precautions before installing or running it:
- Inspect install.sh (the included installer) before running; do not run it as root or on your primary machine until reviewed. The SKILL.md references a different deploy script name — verify actual script behavior.
- Review config/V2.2配置文件.yaml for placeholders like ${TUSHARE_TOKEN} and ${VISION_API_KEY} and understand which features require them. The registry metadata lists no required env vars even though the code expects some — treat any API keys or SMTP credentials as sensitive.
- Do not paste real SMTP credentials or API keys into the example YAML on a public or shared machine; use test accounts or sandboxed credentials first.
- Audit the code (especially modules that perform network calls: data fetchers, notification modules, and any install/setup scripts) for unexpected network endpoints or hardcoded URLs before granting network access.
- Run the code initially in an isolated environment (container or VM) and with limited network access to observe its behavior (what external hosts it contacts, what files it writes).
- If you need automated notifications (WeChat/Feishu/Telegram), expect to provide additional service credentials; verify those integrations require exactly what you provide.
If you want, I can (a) summarize the contents of install.sh and the top-level modules, (b) search the code for network endpoints or sensitive-sink patterns (HTTP post, socket, subprocess exec), or (c) point out exactly where env vars are referenced in code so you know which secrets to manage.
功能分析
Type: OpenClaw Skill
Name: 12-v2-2
Version: 1.0.0
The skill bundle provides a comprehensive quantitative trading system for stock analysis, signal generation, and backtesting. It includes modules for multi-source data management (AKShare, Efinance), market environment analysis, and a notification system supporting email and various webhooks (Telegram, WeChat, Feishu). The code is well-structured, follows standard financial application patterns, and lacks any indicators of malicious intent, data exfiltration, or unauthorized execution. Sensitive configurations like SMTP credentials and API keys are handled via user-provided placeholders in 'config/V2.2配置文件.yaml'.
能力评估
Purpose & Capability
Name, description, docs and code align to a quant trading system (data fetchers, signal calculation, backtest, email notifications). The functionality requested by the skill (AKShare/efinance data, SMTP optional, optional AI vision API) is coherent with that purpose. However, the skill metadata declares no required env vars while the configuration and docs reference environment variables (TUSHARE_TOKEN, VISION_API_KEY) and SMTP credentials — a mismatch between declared requirements and what the code/config expect.
Instruction Scope
SKILL.md gives concrete runtime instructions (edit YAML config, run a one‑click deploy script, run Python backtests and pipeline). The instructions stay within the trading domain (fetch data, compute signals, send notifications). Oddities: the SKILL.md tells users to run 'bash 一键部署脚本.sh' while the repo contains install.sh (name mismatch), and some docs reference copying files from /tmp which may not reflect the packaged layout. No instructions request reading unrelated system secrets or exfiltrating data, but the pipeline will use networked data sources and optional AI APIs as expected.
Install Mechanism
There is no remote install spec in registry metadata (no downloads from external URLs). The repository includes an install script (install.sh) and many code files; that means any install will run local code only. No evidence of downloading arbitrary code from untrusted URLs in the provided metadata. Still, review the install.sh content before running since it will execute on the host.
Credentials
The registry metadata declares no required environment variables or primary credential, but the config files and docs reference sensitive items: ${TUSHARE_TOKEN}, ${VISION_API_KEY}, and SMTP credentials. The skill will function (with degraded features) without those, but asking for SMTP credentials or API keys at runtime is required for notification/vision features. The absence of declared env requirements in the metadata is an incoherence and increases risk because users may accidentally provide secrets without noticing which are necessary.
Persistence & Privilege
The skill does not request 'always: true' and does not declare system-wide privileges. It contains code that writes logs and backtest DB files under local paths (logs/, data/), which is normal for this application. Autonomous invocation (disable-model-invocation=false) is the platform default; combined with the above concerns, you should be cautious but there is no explicit privilege escalation requested.
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 12-v2-2 - 安装完成后,直接呼叫该 Skill 的名称或使用
/12-v2-2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
quant-trading v2.2.0
- 提供完整的8因子量化交易与信号系统,包含多数据源容错(AKShare+Efinance)、实时行情与筹码分布分析。
- 新增实时量比与筹码集中度作为信号因子,信号系统升级为8因子评分。
- 支持自动市场环境判断,并据此动态调整买入阈值。
- 内置邮件与多渠道通知推送,可发送买卖信号及风险警报。
- 一键部署与详细配置说明,文档完善,支持回测和实时监控自选股。
- 增强了数据获取故障熔断与自动重试能力,提升系统稳定性。
元数据
常见问题
12 量化交易V2.2完整版 是什么?
AI量化交易系统,8因子信号系统、多数据源容错、实时行情分析. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 291 次。
如何安装 12 量化交易V2.2完整版?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install 12-v2-2」即可一键安装,无需额外配置。
12 量化交易V2.2完整版 是免费的吗?
是的,12 量化交易V2.2完整版 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
12 量化交易V2.2完整版 支持哪些平台?
12 量化交易V2.2完整版 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 12 量化交易V2.2完整版?
由 nidhov01(@nidhov01)开发并维护,当前版本 v1.0.0。
推荐 Skills