ZhuLinsen's Open-Source Tool Turns Any LLM Into a Daily Stock Dashboard
ZhuLinsen's open-source daily_stock_analysis brings LLM-powered A/H/US stock decision dashboards to anyone with a GitHub account, at zero infrastructure cost.
ProTakeaways
- What it is: Open-source Python tool that runs daily LLM-powered stock analysis on A/H/US/JP/KR/TW markets and pushes decision dashboards to Telegram, Slack, email, and more.
- Zero-cost deployment: Runs entirely on GitHub Actions for free; only cost is optional LLM API calls (Gemini free tier works).
- Key output: Per-stock score, buy/watch/sell signal, entry/exit levels, risk alerts, and catalysts -- delivered on a weekday schedule.
- Broad LLM support: Works with Gemini, Claude, OpenAI-compatible, DeepSeek, Qwen, and local Ollama models via a single config.
- Web workspace included: Full UI at localhost:8000 with Agent strategy chat (15 built-in strategies), backtest view, and portfolio management.
- Not investment advice: The project explicitly frames output as information aggregation, not predictive signals; free data sources can be rate-limited or unstable.
daily_stock_analysis is an open-source Python project by ZhuLinsen that wires together real-time market data, live news search, and your choice of LLM to produce a daily "decision dashboard" for any watchlist of stocks. It covers A-shares, Hong Kong, US, Japanese, Korean, and Taiwan markets, and can push results to Telegram, Discord, Slack, WeChat Work, Feishu, or email -- all running on GitHub Actions for free.
The core idea is straightforward: instead of manually checking charts and news every evening, you define a list of tickers, configure an LLM API key, and the system does the aggregation and reasoning for you on a schedule. It collects market data, news, technical indicators, sentiment signals, and fundamentals, then generates daily AI decision dashboards. Each report includes a numeric score, a buy/watch/sell signal, entry and exit price levels, identified risk alerts, and positive catalysts.

What the system actually does
The pipeline pulls from multiple data layers simultaneously. On the market-data side, free sources like AkShare, Baostock, and YFinance are used by default. On the news side, you can plug in SerpAPI, Tavily, Brave Search, or a self-hosted SearXNG instance. Social sentiment from Reddit, X, and Polymarket is optionally available for US stocks via a dedicated sentiment API.