Investing in today’s markets means juggling dozens of data sources — price charts, news feeds, social sentiment, earnings reports, technical indicators — all before you can decide whether to buy, hold, or sell. For individual investors and even professional traders, this information overload is the single biggest bottleneck to making timely decisions.
Enter Daily Stock Analysis (DSA), an open-source project on GitHub that has already garnered over 34,000 stars since its public release. It transforms any LLM into a personal investment research team that works 24/7, delivering consolidated daily dashboards directly to whichever communication app you prefer — Telegram, Discord, Slack, WeChat Work, Feishu, or email.
What makes DSA extraordinary isn’t just that it analyzes stocks. It’s that you can deploy the entire system for free using GitHub Actions, requiring zero server infrastructure, zero paid tools, and literally five minutes of your time.
In this comprehensive review, we’ll cover everything about Daily Stock Analysis: what it does, how it works, step-by-step setup, real output examples, supported markets and models, and how it compares to alternatives.
What Is Daily Stock Analysis?
Daily Stock Analysis is a Python-based intelligent stock analysis system powered by large language models. It monitors your watchlist across three major markets — Chinese A-shares, Hong Kong stocks, and US stocks — automatically gathering real-time quotes, K-line patterns, technical indicators, fund flows, insider announcements, breaking news, and fundamental data.
Every business day, the system runs a full analysis pipeline, synthesizes everything through an AI model, and outputs a structured Decision Dashboard containing:
- Buy / Hold / Sell ratings with numeric scores
- Price targets and trend direction
- Key risk alerts and catalyst identification
- Sentiment summaries from social media and news
- An operational checklist for each position
This dashboard gets delivered automatically to your preferred channel at a scheduled time (default: 6 PM Beijing Time).
Why This Matters for Investors
Most retail investors fall into one of two traps: either they spend hours every evening manually compiling research across dozens of tabs, or worse, they make decisions based on incomplete information because they simply don’t have time. DSA solves both problems by automating the research process end-to-end while giving you human-readable AI insights instead of raw numbers.
Core Features Breakdown
1. Multi-Market Data Aggregation
DSA pulls data from over a dozen sources simultaneously:
| Data Type | Sources | Markets Covered |
|---|---|---|
| Real-time Quotes | TickFlow, AkShare, Tushare, Pytdx, Baostock, YFinance, Longbridge | A-Shares, HK, US |
| Technical Indicators | Ta-lib integration via data providers | All three markets |
| News & Announcements | SerpAPI, Tavily, Brave Search, Bocha, SearXNG | Global coverage |
| Social Sentiment | Stock Sentiment API (Reddit, X, Polymarket) | US-focused |
| Fundamental Data | Multiple provider tiering system | A-Shares, HK, US |
The system intelligently routes queries between providers based on cost, speed, and availability — so if your primary provider times out, it automatically falls back to another without interrupting the analysis.
2. AI-Powered Decision Reports
Here’s a sample output from the system:
🎯 2026-02-08 Decision Dashboard
Analyzing 3 stocks | 🟢 Buy: 0 🟡 Watch: 2 🔴 Sell: 1
⚪ Zhongwu New Materials (000657): WATCH | Score 65 | Bullish
💭 Sentiment: Market attention on AI attributes and strong growth, positive overall
🚨 Risk Alert:
Risk 1: Major net sell-off of ¥363M on Feb 5 — watch short-term pressure
Risk 2: Chip concentration at 35.15% — distributed chips, resistance expected
✨ Catalyst:
Catalyst 1: Positioned as core HDI supplier for AI servers
Catalyst 2: Q1-Q3 2025扣non-GAAP profit up 407.52% YoY
Each report includes core conclusions, scoring, trend assessment, buy/sell price levels, risk alerts, catalyst factors, and an actionable checklist — all generated naturally by the AI rather than templated placeholders.
3. Agent Strategy Q&A
Beyond static daily reports, DSA includes a built-in chat interface where you can ask questions using any of 11 embedded trading strategies:
- Moving average crossovers (golden cross / death cross)
- Chanlun theory
- Elliott Wave theory
- Bull trend confirmation
- Volume-price analysis
- RSI/MACD divergence detection
- Support/resistance breakout patterns
- And more
Ask questions like “Should I add to my position in Tencent given the current MACD structure?” and get answers grounded in real-time data.
4. Zero-Cost Deployment Options
| Method | Difficulty | Cost | Best For |
|---|---|---|---|
| GitHub Actions | ★☆☆ Easy | $0 absolute | Beginners, no infra |
| Docker | ★★☆ Medium | $0 local | Private deployment |
| Local Python | ★★★ Hard | $0 + API costs | Customization |
| FastAPI Server | ★★★ Hard | $0 + API costs | Team usage |
The GitHub Actions approach is the star feature: fork the repo, configure two environment variables (your API key and stock list), enable Actions, and you’re done. No VPS needed, no cron setup, no Docker knowledge required.
5. Multi-Channel Notification
Your dashboard doesn’t live in some web portal. It arrives wherever you already check:
- Telegram Bot — push notifications instantly
- Discord Webhook — great for community traders
- Slack Bot — professional teams
- WeChat Work Robot — Chinese market focus
- Feishu Robot — enterprise users
- Email — traditional inbox delivery
6. Interactive Web Dashboard
Run python main.py --webui to launch a full interactive dashboard at http://127.0.0.1:8000 featuring:
- Configuration management UI
- Task progress monitoring
- Manual trigger on demand
- Historical report browsing
- Portfolio tracking
- Backtesting module
- Light/dark theme toggle
- Smart import (images, CSV, Excel, clipboard)
- Agent Q&A chat interface
Step-by-Step Setup Guide (GitHub Actions — 5 Minutes)
Prerequisites
- A GitHub account
- At least one LLM API key (recommended: Anspire or AIHubMix for global access; alternatives: Gemini, OpenAI, Anthropic Claude, DeepSeek, Tongyi Qianwen)
- Your stock watchlist
Step 1: Fork the Repository
Visit github.com/ZhuLinsen/daily_stock_analysis and click Fork. Consider starring the repo too.
Step 2: Configure Secrets
Navigate to Settings → Secrets and variables → Actions → New repository secret.
Required secrets:
| Secret Name | Description | Required |
|---|---|---|
ANSPIRE_API_KEYS | AI model API key (supports multiple models + search) | Recommended |
OPENAI_API_KEY | OpenAI-compatible API key | Alternative |
GEMINI_API_KEY | Google Gemini API key | Alternative |
ANTHROPIC_API_KEY | Anthropic Claude API key | Alternative |
STOCK_LIST | Your watchlist e.g., 600519,hk00700,AAPL,TSLA | Required |
Optional notification secret:
| Secret Name | Description |
|---|---|
TELEGRAM_BOT_TOKEN + TELEGRAM_CHAT_ID | Telegram notifications |
DISCORD_WEBHOOK_URL | Discord notifications |
FEISHU_WEBHOOK_URL | Feishu notifications |
WECHAT_WEBHOOK_URL | WeChat Work notifications |
EMAIL_SENDER + EMAIL_PASSWORD | Email notifications |
Step 3: Enable Actions
Go to the Actions tab and click “I understand my workflows, go ahead and enable them.”
Step 4: Run Your First Analysis
Click Actions → 每日股票分析 → Run workflow → Run workflow. Within minutes, check your chosen notification channel for your first AI-generated investment dashboard.
That’s it. You now have a free, always-running stock analysis system.
Supported AI Models
DSA supports virtually every major AI model provider:
- Anspire — Recommended. Single key provides access to global popular models plus integrated search capabilities. Includes free tier credits.
- AIHubMix — Single key for multiple models with 10% discount available
- Google Gemini — Strong multilingual support
- OpenAI (GPT-4, GPT-3.5 Turbo)
- Anthropic Claude (Sonnet, Opus)
- DeepSeek — Cost-effective option
- Tongyi Qianwen (Qwen) — Optimized for Chinese markets
- Ollama — Fully local inference for privacy-focused users
For China market analysis, models with strong Chinese language capability (Qwen, Kimi, MiniMax) produce notably better results.
Comparison With Alternatives
| Feature | Daily Stock Analysis | TradingView Alerts | Koyfin | FinMind |
|---|---|---|---|---|
| AI-generated reports | ✅ Full dashboard | ❌ Basic alerts | Partial | ❌ |
| A-share support | ✅ Native | Limited | Limited | Yes |
| Multi-market (HK+US) | ✅ All three | ✅ Excellent | ✅ Good | Partial |
| Zero-cost setup | ✅ GitHub Actions | Paid plans only | Free tier limited | Self-host |
| Auto-push notifications | ✅ 6 channels | Email/web only | Email only | Self-built |
| Social sentiment | ✅ Reddit/X/Polymarket | ❌ | Limited | Partial |
| Strategy Q&A | ✅ 11 built-in | ❌ | ❌ | ❌ |
| Multi-model routing | ✅ Automatic fallback | N/A | N/A | Single model |
| Web dashboard | ✅ Built-in | ✅ Excellent | ✅ Good | Basic |
DSA uniquely combines deep Asian market expertise with modern AI capabilities and truly zero-cost deployment — something proprietary platforms simply cannot match.
Real-World Use Cases
Active Day Trader
Set up a morning briefing that runs at 8 AM Beijing Time, covering overnight US moves, pre-market HK futures, and A-share sector rotation signals. Receive the compiled brief on Telegram before opening your broker terminal.
Swing Trader
Configure a weekly deep-dive report analyzing your top 20 candidates with technical patterns, fundamental health checks, and relative strength ranking across sectors.
Passive Investor
Create a monthly portfolio review script that scans your holdings against broader market indices, flags dividend changes, earnings surprises, and material corporate actions — all summarized by AI.
Financial Content Creator
Use the agent Q&A interface to generate talking points and data-backed arguments for investment content on YouTube, Twitter/X, or Substack.
Limitations to Be Aware Of
While DSA is impressive, no tool is perfect:
- Data source quality varies — Free tiers of data providers may have rate limits. Paid data subscriptions improve reliability significantly.
- LLM analysis is probabilistic — AI generates opinions, not guarantees. Treat recommendations as starting points for your own due diligence.
- Chinese-language focused defaults — While fully functional for global markets, some configuration documentation is primarily in Chinese.
- GitHub Actions runtime limit — Free tier has a 6-hour monthly compute cap. One daily run stays well within limits, but intensive multi-day batch jobs may hit caps.
- Not a trading execution platform — DSA provides analysis and alerts but does not execute trades on your behalf.
Getting Started Today
Daily Stock Analysis represents one of the best free tools in the open-source fintech space. Its combination of broad market coverage, AI-powered synthesis, flexible notification channels, and effortless deployment makes it accessible to anyone — from complete beginners to experienced quant traders.
The most important decision isn’t technical. It’s picking your first stocks to watch and choosing an AI model you trust. Everything else is genuinely just five minutes of configuration.
Common Pitfalls and How to Avoid Them
Having deployed this system for dozens of users, there are several recurring issues worth noting:
API Key Configuration Errors
The most frequent error is misnaming environment variable keys in GitHub Actions. The repo expects exact matches like ANSPIRE_API_KEYS (note the plural). A single character mismatch will cause the workflow to fail silently with no clear error message in the dashboard. Always double-check spelling before running your first workflow.
Stock Symbol Format Confusion
A-shares need full codes (e.g., 600519 not Moutai), Hong Kong stocks use hk prefix (e.g., hk00700), and US stocks use ticker symbols (e.g., AAPL). Mixing up these formats is another common gotcha — the system won’t always produce an error, but the analysis results may be incomplete or wrong.
Notification Channel Setup Complexity
Telegram bot setup requires two separate steps: creating a bot via BotFather and then obtaining your chat ID from @userinfobot. Discord webhook is simpler but you must place it in a specific channel. If notifications don’t arrive within 30 minutes of triggering the workflow, check both the secrets names and the target URLs/tokens for typos.
Rate Limit Throttling
Free data providers have daily query limits. If your stock list exceeds 50 instruments, you may start seeing timeout warnings during peak hours. Consider splitting into batches or upgrading to paid tiers like TickFlow for production use.
Resources: