Verified project record

hsliuping/TradingAgents-CN

tradingagents-cn is a multi-agent LLM-based research assistant focused on the Chinese A-share stock market, providing debate-style analysis, natural-language stock screening, paper trading and study courses. It is designed as a learning and research aid rather than an investment advisory tool.

Project overview

It applies a multi-agent debate structure with multiple analyst roles and optimistic/cautious researcher perspectives to A-share research, and pairs it with domestic data sources (Tushare, AKShare, BaoStock) and a Chinese-language experience, which is hard for individual learners to orchestrate on their own.

Project type: AI Agent · Learning Resources
Use cases: Data Analysis · Learning & Education
Deployment: Refer to project documentation
License: License pending

Best for

Individual learners and researchers of the Chinese A-share market who want multi-perspective, debate-style analysis and paper trading as a study aid.
Users comfortable setting up a Python web application, including self-installing MongoDB and Redis and configuring API keys.

Key capabilities

Multi-agent debate-style analysis for a single stock, available at 5 depth levels with multiple scenario viewpoints.
Create, publish and manage custom agents — Pro edition only; availability of this capability is unclear.

Limitations and risks

Supports only A-share market research and produces no trading instructions such as target prices, stop-loss or buy/sell timing.
The project does not ship its own database; users must install MongoDB and Redis themselves.
All AI output is for learning reference only and does not constitute investment advice; users must judge independently.
The application layer (app/, frontend/, core/) is proprietary: redistribution and commercial use are prohibited without a commercial license. Review the license terms before adopting or redistributing.

Getting started

Install MongoDB 7.0+ and Redis 6+ yourself (they are not bundled); clone the repository; create a Python 3.11 venv and run pip install -r requirements.txt; copy .env.example to .env and fill in database credentials and JWT_SECRET; start the backend with uvicorn app.main:app; in frontend/ run npm install then npm run dev; finally log in as admin and configure an LLM API key in system settings.

Evidence and sources

README: multi-agent stock research assistant for the Chinese A-share market. It orchestrates LLM-based analyst agents for debate-style research, natural-language stock screening, paper tr…
README
README
README
README

Source updated at: 2026-09-23T01:31:01Z

Open original GitHub project