TradingAgentsX - Multi-Agent Intelligent Trading Analysis System

AI Stock Trading Analysis Platform Based on LangGraph, Combining Multiple Professional AI Agents for Collaborative Decision-Making

---

📖 Introduction

TradingAgentsX is an advanced multi-agent AI trading analysis system that simulates the operational model of real-world trading firms. Through LangGraph orchestration of multiple specialized AI agents (Analysts, Researchers, Traders, Risk Managers), the system can analyze stock markets from different perspectives and produce high-quality trading decisions through structured debate and collaborative processes.

💡 Tribute to the Original: This project is based on improvements and extensions of TauricResearch/TradingAgents.

🎯 Core Features

Feature	Description
🤖 Multi-Agent Collaboration	12 specialized AI agents (Analysts, Researchers, Traders, Risk Managers) work together
🌐 Multi-Model Support	OpenAI, Anthropic, Gemini, Grok, DeepSeek, Qwen and other LLM providers
🔒 Google OAuth Login	Cloud sync API settings and historical reports, supporting multi-device sync
📊 US & Taiwan Stock Support	Full support for US stocks (Yahoo Finance) and Taiwan stocks (FinMind) data
🔑 BYOK Mode	Users bring their own API keys, encrypted frontend storage, ensuring privacy
🛡️ Security Protection	Rate Limiting, Security Headers, API Key masking
📱 Responsive Design	Supports desktop and mobile browsers
🐳 Docker Deployment	One-click startup for frontend and backend services
🧠 Embeddings Model Choice	Supports sentence-transformers (local free) or OpenAI embeddings

---

🏗️ System Architecture

TradingAgentsX/
├── frontend/                   # Next.js Frontend Application
│   ├── app/                    # App Router Pages
│   │   ├── page.tsx            # Home Page
│   │   ├── layout.tsx          # Root Layout
│   │   ├── globals.css         # Global Styles
│   │   ├── analysis/           # Analysis Feature Pages
│   │   ├── history/            # History Reports Page
│   │   ├── auth/               # OAuth Callback
│   │   └── api/                # API Routes (config, auth)
│   ├── components/             # React Components
│   │   ├── AgentFlowDiagram.tsx    # Agent Flow Diagram Component
│   │   ├── PendingTaskRecovery.tsx # Task Recovery Component
│   │   ├── analysis/           # Analysis Related Components
│   │   ├── auth/               # Login Buttons
│   │   ├── layout/             # Header, Footer
│   │   ├── settings/           # API Settings Dialog
│   │   ├── shared/             # Shared Components
│   │   ├── theme/              # Theme Related Components
│   │   └── ui/                 # shadcn/ui Base Components (16)
│   ├── contexts/               # React Context (Auth State)
│   ├── hooks/                  # Custom Hooks
│   └── lib/                    # Utility Functions
│       ├── api.ts              # API Calls
│       ├── api-helpers.ts      # API Helper Functions
│       ├── crypto.ts           # Encryption Utils
│       ├── storage.ts          # Local Storage
│       ├── reports-db.ts       # IndexedDB Report Storage
│       ├── pending-task.ts     # Pending Task Management
│       ├── user-api.ts         # User API
│       ├── types.ts            # TypeScript Type Definitions
│       └── utils.ts            # General Utilities
│
├── backend/                    # FastAPI Backend Service
│   ├── __main__.py             # Application Entry Point
│   └── app/
│       ├── main.py             # FastAPI App (Middleware, Routes)
│       ├── api/                # API Routes
│       │   ├── routes.py       # Analysis API
│       │   ├── auth.py         # Google OAuth
│       │   ├── user.py         # User Data Sync
│       │   └── dependencies.py # Dependency Injection
│       ├── core/               # Core Configuration
│       ├── db/                 # PostgreSQL Database
│       ├── models/             # Pydantic Models
│       └── services/           # Business Logic
│           ├── trading_service.py  # Trading Analysis Service
│           ├── task_manager.py     # Task Manager
│           ├── pdf_generator.py    # PDF Report Generator
│           ├── price_service.py    # Stock Price Data Service
│           ├── download_service.py # Download Service
│           ├── redis_client.py     # Redis Client
│           └── auth_utils.py       # Auth Utilities
│
└── tradingagents/              # Core AI Agent Package
    ├── agents/                 # AI Agent Definitions
    │   ├── analysts/           # Analyst Team
    │   │   ├── market_analyst.py       # Market Analyst
    │   │   ├── news_analyst.py         # News Analyst
    │   │   ├── social_media_analyst.py # Social Media Analyst
    │   │   └── fundamentals_analyst.py # Fundamentals Analyst
    │   ├── researchers/        # Research Team
    │   │   ├── bull_researcher.py      # Bull Researcher
    │   │   └── bear_researcher.py      # Bear Researcher
    │   ├── trader/             # Trader
    │   │   └── trader.py               # Trader Agent
    │   ├── risk_mgmt/          # Risk Management Team
    │   │   ├── aggresive_debator.py    # Aggressive Analyst
    │   │   ├── conservative_debator.py # Conservative Analyst
    │   │   └── neutral_debator.py      # Neutral Analyst
    │   ├── managers/           # Manager Decision Makers
    │   │   ├── research_manager.py     # Research Manager
    │   │   └── risk_manager.py         # Risk Manager
    │   └── utils/              # Agent Utility Functions
    ├── dataflows/              # Data Fetching and Processing
    │   ├── interface.py        # Unified Data Interface
    │   ├── config.py           # Dataflow Configuration
    │   ├── y_finance.py        # Yahoo Finance Data
    │   ├── yfin_utils.py       # Yahoo Finance Utils
    │   ├── alpha_vantage*.py   # Alpha Vantage Series (5)
    │   ├── finmind*.py         # FinMind Taiwan Stock Data (6)
    │   ├── google.py           # Google Search
    │   ├── googlenews_utils.py # Google News Utils
    │   ├── reddit_utils.py     # Reddit Data
    │   ├── openai.py           # OpenAI Embeddings
    │   └── retry_utils.py      # Retry Utils
    ├── graph/                  # LangGraph Workflow
    │   ├── trading_graph.py    # Trading Analysis Graph
    │   ├── setup.py            # Graph Setup
    │   ├── propagation.py      # State Propagation
    │   ├── reflection.py       # Reflection Mechanism
    │   ├── conditional_logic.py    # Conditional Logic
    │   └── signal_processing.py    # Signal Processing
    ├── utils/                  # General Utils
    └── default_config.py       # Default Configuration

---

🤖 AI Agent Team

Analyst Team (4 Members)

Agent	Responsibility	Output
Market Analyst	Technical Analysis	RSI, MACD, Bollinger Bands, Support/Resistance Levels
Social Media Analyst	Sentiment Assessment	Reddit/Twitter Sentiment Indicators, Investor Confidence
News Analyst	News Analysis	Latest News Summaries, Event Impact Assessment
Fundamentals Analyst	Financial Analysis	Financial Reports, P/E, P/B, Profitability

Research Team (3 Members)

Agent	Responsibility
Bull Researcher	Bullish Argument, Upside Catalyst Analysis
Bear Researcher	Bearish Argument, Downside Risk Warnings
Research Manager	Comprehensive Decision from Bull and Bear Views

Trading & Risk Team (5 Members)

Agent	Responsibility
Trader	Integrates All Reports, Formulates Trade Plan
Aggressive Analyst	High-Risk High-Return Strategy Analysis
Conservative Analyst	Stable Conservative Strategy & Risk Control
Neutral Analyst	Neutral Balanced Strategy Assessment
Risk Manager	Comprehensive Risk Management & Final Advice

---

🚀 Quick Start

Prerequisites

• Python 3.10+
• Node.js 18.x+ or Bun 1.x+

Required API Keys

API	Purpose	Apply URL
OpenAI	GPT Models	https://platform.openai.com/api-keys
Alpha Vantage (Optional)	US Stock Fundamentals	https://www.alphavantage.co/support/#api-key
FinMind (Optional)	Taiwan Stock Data	https://finmindtrade.com/

Installation Steps

1️⃣ Clone the Project

git clone https://github.com/MarkLo127/TradingAgentsX.git
cd TradingAgentsX

2️⃣ Backend Setup

# Create virtual environment
conda create -n tradingagents python=3.13
conda activate tradingagents

# Install dependencies
pip install -e .
pip install -r backend/requirements.txt

# Configure environment variables
cp .env.example .env
# Edit .env and fill in API keys

# Start backend
python -m backend

Backend Services:

• API: http://localhost:8000
• Swagger Docs: http://localhost:8000/docs

3️⃣ Frontend Setup

# Install dependencies (run from project root)
bun install --cwd frontend

# Start development server
bun run --cwd frontend dev

Frontend Application: http://localhost:3000

---

🐳 Docker Deployment

# Configure environment variables
cp .env.example .env

# Start services
docker compose up -d --build

# View logs
docker compose logs -f

# Stop services
docker compose down -v

Service Ports:

• Backend: http://localhost:8000
• Frontend: http://localhost:3000

---

🔒 Security Features

Local Development vs Production Environment

Feature	Local Development (localhost)	Production (Railway, etc.)
Google Login	Optional	Recommended
Auto Data Cleanup	❌ No cleanup	✅ Clears when leaving if not logged in
PostgreSQL	Optional	Required
API Settings Storage	Permanently retained	Cloud sync after login
History Reports Storage	Permanently retained	Cloud sync after login

Frontend Security

• API Key Encrypted Storage - Uses AES-GCM to encrypt sensitive data in localStorage
• Auto Cleanup (Production Only) - Automatically clears local data when non-logged-in users leave the page
• Safari Touch Optimization - Fixed touch event issues on iOS Safari

Backend Security

• Rate Limiting - 30 requests per minute limit
• Security Headers - X-Content-Type-Options, X-Frame-Options, etc.
• Sensitive Data Masking - API Keys are automatically masked in logs
• CORS Configuration - Restricts cross-origin request sources

Cloud Sync

• Google OAuth 2.0 - Secure third-party login
• JWT Token - Stateless authentication
• Cloud Backup - API settings and history reports sync to server

---

📱 User Guide

1. Configure API Keys

Click the "Settings" button in the upper right corner and enter your API keys.

2. Select Analysis Parameters

• Market Type: US Stock / Taiwan Listed / Taiwan OTC
• Stock Symbol: e.g., NVDA, 2330
• Analyst Team: Select the analysts you need
• Research Depth: Shallow (Fast) / Medium / Deep (Detailed)
• LLM Model: Quick Think Model + Deep Think Model

3. Execute Analysis

Click "Execute Analysis" and wait 1-5 minutes (depending on research depth).

4. View Results

• Trading Decision Summary - BUY / SELL / HOLD recommendations
• Stock Price Chart - Toggle between Line Chart / Candlestick Chart
• 12 Agent Reports - Click tabs to view detailed analysis

5. Save & Download

• Save Report - Save to local / cloud
• Download PDF Report - Export complete PDF analysis report

📄 Sample Report

View a complete PDF analysis report sample:

📥 Broadcom Inc. Analysis Report (2026-03-09)

---

🔌 API Documentation

Health Check

GET /api/health

Execute Analysis

POST /api/analyze
Content-Type: application/json

{
  "ticker": "NVDA",
  "market_type": "us",
  "analysis_date": "2024-01-15",
  "research_depth": 2,
  "analysts": ["market", "social", "news", "fundamentals"],
  "quick_think_llm": "gpt-5-mini",
  "deep_think_llm": "claude-sonnet-4-5",
  "quick_think_api_key": "sk-...",
  "deep_think_api_key": "sk-ant-...",
  "embedding_api_key": "sk-...",
  "alpha_vantage_api_key": "..."
}

Query Task Status

GET /api/task/{task_id}

Full Documentation: http://localhost:8000/docs

---

🛠️ Tech Stack

Backend

Technology	Purpose
FastAPI	Async Web Framework
LangGraph	Multi-Agent Workflow Orchestration
LangChain	LLM Application Development
ChromaDB	Vector Database (Memory System)
PostgreSQL	User Data Storage
SQLAlchemy + asyncpg	Async Database ORM
Pydantic	Data Validation

Frontend

Technology	Purpose
Next.js 16	React Full-Stack Framework
TypeScript	Static Typing
Tailwind CSS	Styling Framework
shadcn/ui	UI Component Library
Dexie.js	IndexedDB Wrapper
Recharts	Data Visualization

---

📸 Application Screenshots

Home Page

---

API Configuration Page

---

Task Configuration Page

---

Agent View Selection

12 Professional Agent Tabs, click to switch and view different agents' analysis reports:

• Analyst Team (4): Market Analyst, Social Media Analyst, News Analyst, Fundamentals Analyst
• Research Team (3): Bull Researcher, Bear Researcher, Research Manager
• Trading Team (1): Trader
• Risk Management Team (4): Aggressive Analyst, Conservative Analyst, Neutral Analyst, Risk Manager

---

Stock Price Chart (Candlestick)

---

Stock Price Chart (Line Chart)

---

Market Analyst Report

---

Social Media Analyst Report

---

News Analyst Report

---

Fundamentals Analyst Report

---

Bull Researcher Report

---

Bear Researcher Report

---

Research Manager Report

---

Trader Report

---

Aggressive Analyst Report

---

Conservative Analyst Report

---

Neutral Analyst Report

---

Risk Manager Report

---

History Reports (Login Required)

---

🙏 Acknowledgements

• TauricResearch/TradingAgents - Original Project
• LangChain - LLM Application Framework
• LangGraph - Multi-Agent Orchestration
• FastAPI - Web Framework
• Next.js - React Framework
• shadcn/ui - UI Component Library

---

📄 License

This project is licensed under the Apache 2.0 License - see the LICENSE file for details.