Daily Stock Analysis: The Complete Guide to AI-Powered Stock Intelligence
What if you could get a professional Gemini AI decision dashboard for your stock watchlist — every single day — delivered to WeChat Work, Feishu, Telegram, or your email inbox, at zero cost? Daily Stock Analysis does exactly that. With 17,000+ GitHub stars, it's an LLM-driven intelligent analyzer for A-shares, Hong Kong stocks, and US stocks that runs automatically via GitHub Actions.
Daily Stock Analysis on GitHub
What Is Daily Stock Analysis?
An open-source AI-powered stock analysis system that combines multi-source market data, real-time news, and LLM intelligence to generate daily decision dashboards for your personal watchlist. Zero cost, fully automated, scheduled execution.
- Language: Python
- License: MIT
- Stars: 17,000+ ⭐
- Forks: 17,800+
- Releases: 5
- Markets: A-shares (China), H-shares (Hong Kong), US stocks
Core Features
🤖 AI Decision Dashboard
Every day, the system generates a comprehensive decision dashboard for each stock in your watchlist:
- Market overview and trend analysis
- Individual stock analysis with buy/sell signals
- News sentiment integration
- Technical indicator analysis
- Risk assessment
📊 Multi-Source Market Data
- YFinance for US stock historical data and real-time quotes (consistent dividend adjustment)
- Multiple Chinese market data providers for A-shares and H-shares
- Real-time news aggregation
📱 Multi-Channel Push Notifications
Daily analysis results pushed automatically to:
- WeChat Work (企业微信)
- Feishu (飞书/Lark)
- Telegram
🆓 Zero Cost Deployment
- GitHub Actions — Recommended. Runs entirely free on GitHub's infrastructure
- Docker — Self-hosted deployment
- Local — Direct Python execution
🖥️ Web Interface
Full-featured web UI with:
- Configuration management
- Task monitoring
- Manual analysis triggers
- Optional password protection (
ADMIN_AUTH_ENABLED=true)
📸 Vision AI: Add Stocks from Screenshots
Drag or select screenshots from your stock app (positions page, watchlist). The system uses Vision AI (Gemini/Claude/GPT) to automatically recognize stock codes and add them to your watchlist.
POST /api/v1/stocks/extract-from-image
→ { "codes": ["600519", "300750", ...] }
🤖 Agent Strategy Mode
Enable AGENT_MODE=true and access /chat for multi-turn strategy Q&A:
- 11 Built-in Strategies: MA crossover (均线金叉), Chan theory (缠论), Wave theory (波浪理论), Bullish trend (多头趋势), and more
- Natural Language Queries: Ask "Analyze 600519 using Chan theory" — the Agent calls real-time data, K-line, technical indicators, and news tools
- Streaming Progress: Watch the AI's reasoning path in real-time (data fetch → technical analysis → news search → conclusion)
- Multi-turn Conversations: Context-aware follow-up questions with persistent session history
- Bot Commands:
/ask <code> [strategy] - Custom Strategies: Add YAML files to
strategies/directory — no coding required
⚙️ Multi-LLM Support
Via LiteLLM integration — supports Gemini, OpenAI, Claude, and other LLM providers. Also supports AIHubMix for Chinese users.
Quick Start
Option 1: GitHub Actions (Recommended — Free)
- Fork the repository
- Set GitHub Secrets (API keys, stock codes, push channels)
- Enable GitHub Actions
- Done — runs daily on schedule
Option 2: Docker
docker-compose up -d
Option 3: Local
pip install -r requirements.txt
python main.py --webui # Web UI + scheduled analysis
python main.py --webui-only # Web UI only
Daily Stock Analysis vs Alternatives
Category: This tool is an LLM-powered stock analysis & notification system.
| Feature | Daily Stock Analysis | FinGPT | TradingAgents |
|---|---|---|---|
| Focus | Daily watchlist intelligence + push | Financial LLM training | Multi-agent trading simulation |
| Stars | 17K ⭐ | 18.8K ⭐ | 31.6K ⭐ |
| License | MIT | MIT | Apache 2.0 |
| Language | Python | Jupyter Notebook | Python |
| Markets | ✅ A/H/US stocks | US stocks | US stocks |
| Daily Automation | ✅ GitHub Actions (free) | ❌ | ❌ |
| Push Notifications | ✅ WeChat/Feishu/Telegram/Email | ❌ | ❌ |
| Decision Dashboard | ✅ Visual dashboard per stock | ❌ | ❌ |
| Web UI | ✅ Full config + monitoring | ❌ | ❌ |
| Agent Strategy Q&A | ✅ 11 strategies, multi-turn chat | ❌ | ❌ |
| Vision AI (Screenshots) | ✅ Add stocks from images | ❌ | ❌ |
| Custom Strategies (YAML) | ✅ No-code | ❌ | ❌ |
| Multi-LLM (LiteLLM) | ✅ Gemini/OpenAI/Claude | ✅ Fine-tuned models | ✅ Multi-provider |
| Docker Deployment | ✅ | ❌ | ❌ |
| Zero Cost | ✅ GitHub Actions | ❌ GPU required | ❌ API costs |
| LoRA Fine-Tuning | ❌ | ✅ Financial LLM training | ❌ |
| Sentiment Analysis | ✅ via news | ✅ Specialized NLP | ✅ Dedicated agent |
| Trading Execution | ❌ Analysis only | ❌ Research only | ✅ Simulated execution |
| Multi-Agent Simulation | ❌ | ❌ | ✅ Trading firm simulation |
| Risk Management Agent | ❌ | ❌ | ✅ Dedicated risk agent |
| Backtesting | ❌ | ✅ | ✅ |
| Academic Paper | ❌ | ✅ Published research | ✅ arXiv paper |
| Team | ZhuLinsen | AI4Finance Foundation | Tauric Research |
When to choose Daily Stock Analysis: You want a zero-cost, daily automated stock intelligence system that monitors your personal watchlist across A/H/US markets, generates visual decision dashboards, pushes results to WeChat Work/Feishu/Telegram/Email, and offers Agent-based strategy Q&A with 11 built-in strategies. Perfect for individual investors who want AI insights without the complexity.
When to choose FinGPT: You want to train or fine-tune your own financial LLM using LoRA. Academic-grade with published research, specialized in sentiment analysis, robo-advisory, and market forecasting. Better for researchers and quantitative analysts building custom models.
When to choose TradingAgents: You want a multi-agent trading simulation that mimics a real trading firm with specialized agents (fundamental analyst, sentiment analyst, technical analyst, trader, risk manager). Academic paper on arXiv. Better for research, backtesting, and simulated trading.
Configuration
| Setting | Description |
|---|---|
| Stock codes | Your watchlist (A-shares: 600519, H-shares: 00700.HK, US: AAPL) |
| LLM API keys | Gemini/OpenAI/Claude API key |
| Push channels | WeChat Work webhook, Feishu webhook, Telegram bot, Email SMTP |
| Schedule | Cron expression for GitHub Actions |
| Agent mode | AGENT_MODE=true for strategy Q&A |
| Web auth | ADMIN_AUTH_ENABLED=true for password protection |
Conclusion
Daily Stock Analysis democratizes AI-powered stock intelligence. No expensive Bloomberg terminal. No complex quantitative infrastructure. Just fork a repo, set your watchlist, configure your push channel, and get daily AI decision dashboards — all at zero cost via GitHub Actions. The Agent mode with 11 built-in strategies brings institutional-grade analysis tools to individual investors.
Explore Daily Stock Analysis on GitHub