Skip to content

Fidingks/Mao-Quant

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 

Repository files navigation

MaoQuant Skills

Validate any A-share trading idea in one sentence.

简体中文 | English


You have a trading idea. MaoQuant generates a complete backtest report with equity curves, proving whether your strategy works -- programmatically.

Quick Start

npx skills add Fidingks/Mao-Quant

Then tell your AI assistant:

/backtest ema-crossover SH600000

Done. Full report with charts, metrics, and trade log.

What You Can Do

Just ask like this... What you get
"Can I make money on Moutai with a moving average strategy?" Full backtest with equity curve and report
"Is CATL good for short-term trading? Try KDJ" KDJ strategy backtest, every trade marked on chart
"Does buying on MACD golden cross actually work? Test it on Ping An" MACD backtest with win rate, profit factor
"Find me stocks with PE below 15 and high volume" Full-market scan, filtered stock list
"How much drawdown if I trade Bollinger Band bounces?" Bollinger band backtest, max drawdown highlighted

Talk to it like you'd talk to a quant-savvy friend. MaoQuant handles the rest.

Built-in Strategies

Strategy Logic Best For
EMA Crossover Fast/slow EMA golden cross / death cross Trending markets
RSI Overbought / oversold reversal Range-bound markets
MACD DIF / DEA crossover Medium-term trends
KDJ Stochastic extreme values Short-term swings
Bollinger Bands Price touching upper/lower bands Mean reversion

Data Sources

Dual data engine -- choose what fits:

Engine Coverage Setup
FaceCat API A-shares, daily bars Zero config, works out of the box
TDX (TongDaXin) Full A-share, daily/1min/5min Requires TDX client with local data

Built-in data works immediately. No API key needed.

A-Share Rules (Auto-Enforced)

No configuration needed. MaoQuant enforces these automatically:

  • T+1: Buy today, earliest sell is tomorrow
  • Price Limits: Main board +/-10%, ChiNext/STAR +/-20%, BSE +/-30%
  • Lot Sizing: Minimum 100 shares per trade
  • Stamp Tax: 0.1% on sell side
  • Commission: 0.025% both sides, minimum 5 CNY

Architecture

MaoQuant follows the AI Skill Manifest specification. The skill system is fully self-describing:

skills/
  SKILL.md              # Root manifest with capabilities, contracts, environment
  backtest/SKILL.md     # Backtest skill (user-invocable)
  scan/SKILL.md         # Screening skill (user-invocable)
  data/SKILL.md         # Data engine reference
  catquant-expert/      # Knowledge base + 6 rule files
catquant/               # Python engine (backtest, indicators, charts, data)

Key design decisions:

  • BarSeries container: get_history() returns a BarSeries whose repr shows only a summary -- raw K-line data never leaks into AI context
  • Selftest: python -m catquant.selftest validates the entire environment in 10 seconds
  • Contracts: T+1, price limits, fees, and lot sizing are enforced by the engine, not by prompts

Environment

pip install -r requirements.txt
cp .env.sample .env     # Edit FaceCat_URL and TDX_DIR if needed
python -m catquant.selftest

Supported Clients

OpenClaw, Claude Code, Cursor, Windsurf, Copilot, Cline, OpenCode, Trae, and 40+ AI coding clients.

Our Team

Built by the FaceCat Quantitative Research Team. Members from: DZH (LongRuan), East Money, Soochow Securities, GF Securities, Donghai Securities, Shanxi Securities, Xiangcai Securities, Huatai Securities, Hengtai Futures, Deutsche Bank.

Full Service

We offer:

  • Full-market data API -- A-share real-time and historical data with proprietary analytics
  • Custom strategy development -- Bespoke backtesting solutions for your trading ideas

Contact us: https://www.jjmfc.com


Built by FaceCat Quantitative Research Team

About

MaoQuant - 面向AI Agent的A股数据+回测+分析技能包(Sills&MCP),内置A股交易规则(T+1、涨跌停、整手交易)与策略模板。

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages