simonlin1212 / TradingAgents-astock
A deeply localized multi-agent investment research framework that adapts TradingAgents for Chinese A-share markets with 7 specialized analyst roles and free local data sources.
星标趋势
AI 分析
项目摘要
TradingAgents-astock is a Chinese A-share market specialized fork of the TradingAgents multi-agent investment research framework. It features 7 AI analyst roles (market, social, news, fundamentals, policy, hot money, lockup monitoring) that debate and make investment decisions based on A-share-specific trading rules and free local data sources like mootdx, East Money, Sina, and Tonghuashun.
为什么值得关注
This project is notable for deeply adapting a popular multi-agent LLM framework specifically for Chinese A-share markets, adding 3 specialized analyst roles (policy, hot money, lockup monitoring) that reflect unique A-share dynamics, integrating free direct-connect data sources, and implementing A-share trading constraints (T+1, price limits, lot sizes). With 2K+ stars and active development (14 releases in 6 months), it demonstrates strong community interest in localized AI investment research tools.
优势
- Deep A-share market specialization with 7 domain-specific analyst roles including unique policy/hot money/lockup monitoring
- Free direct-connect data sources (mootdx, East Money, Sina, Tonghuashun) with zero external service dependencies
- Well-structured documentation with clear upstream comparison, architecture diagrams, and Chinese-language reports
- Active development with 14 releases in 6 months, Docker support, and working examples
- Implements real A-share trading constraints (T+1, price limits, minimum lot sizes, ST stocks)
局限性
- No CI/CD pipeline despite having tests, reducing confidence in code quality maintenance
- Built as a fork dependent on upstream TradingAgents; long-term maintenance and divergence risks unclear
- No backtesting results, performance metrics, or empirical validation of investment decisions shown
使用场景
- A-share investment research and multi-perspective stock analysis for Chinese retail investors
- Educational demonstration of multi-agent LLM systems applied to financial domains
- Quantitative trading strategy prototyping with A-share-specific signals
- Research platform for studying AI-driven market analysis and debate mechanisms