mmtickerlab

mmtickerlab is an Agent skill library for secondary-market (A-share/HK/US stock) investment research, quantitative analysis, and simulated trading, suitable for agent environments that support Agent Skills, such as Claude Code and OpenClaw. The repository is organized as a skill-package matrix and contains 10 standardized skills: trading-sop serves as the unified entry point responsible for intent recognition and SOP routing; ticker-pipeline supports multi-Agent parallel investment research and risk-control decisions; market provides market quotes, candlesticks, technical indicators, capital flows, Dragon-Tiger List, financial data, and commodity futures data; market-intel outputs market intelligence briefs; asset-analysis performs fundamental and multi-model valuation analysis; risk-guard executes quantitative risk-control gates and has veto power; industry-research provides industry supply-chain insight; first-board-overnight targets the A-share first-board overnight strategy; sim-trade implements simulated trading with T+1 rules and five-level order-book matching; plan-review generates three-stage reviews before the market open, at midday, and after the close. The project emphasizes data determinism and completeness. When key data is missing, it triggers Fail-Fast blocking. HK/US stock data is supplied by yfinance and Sina direct-connect sources, while A-share-specific data is carried by AKShare. It includes a lightweight built-in HTTP dashboard service (default port 19876, implemented purely with the standard library), which can archive and visually display various research reports and provides interactive indicator charts such as dynamic candlesticks and moving averages, trading volume, MACD, and RSI. Installation supports one-click installation of the full set or individual skills via ClawHub, and the repository can also be cloned directly into the agent skills directory. The project is open-sourced under the MIT license.