TradingAgents-AShare

Why we included this project

Teams building LLM-driven investment research will find this a complete, runnable reference for applying a multi-agent debate architecture to a real market. The system runs analyst, researcher, and risk-control agents that argue bull and bear cases for A-share stocks and converge on a structured verdict with direction, confidence, target price, and stop-loss. Around the debate engine sits a web dashboard, natural-language queries, watchlists with scheduled analysis, and position tracking, so it works as a daily tool rather than a demo. Since it is built on LangGraph and supports multiple model providers, it also works as a template for adapting the TradingAgents pattern to other markets or asset classes.