# daily_stock_analysis

> Daily AI reports on your stock watchlist, pushed to chat or email on a schedule

- Repository: https://github.com/ZhuLinsen/daily_stock_analysis
- Category: Business apps
- Stars: 66,898 · Forks: 55,554 · Language: Python · License: MIT
- Experience: Some experience
- Page: https://ir.adyog.com/repos/zhulinsen/daily_stock_analysis/

Checking many stocks every day means reading prices, charts and news by hand. This project automates that. You list the tickers you follow. Each day it pulls market data and news, asks an LLM to read them, and writes a report for each stock.

Each report has a conclusion, a score, a trend read, suggested buy and sell levels, risk warnings and catalysts. A summary goes to WeChat Work, Feishu, Telegram, Discord, Slack or email. It can also write a market-wide recap. Free data sources (AkShare, Baostock, YFinance) work with no setup, though the README says they can be unreliable. It suggests token-based sources for steadier data.

You can fork the repo and let GitHub Actions run it on a schedule, run it locally or in Docker, or use the built-in web workspace. The workspace adds report history, backtesting, portfolio tracking and a chat page for follow-up questions using built-in strategies. The README says it is for study and research, not investment advice.

## What it does

- **AI decision reports**: Each stock gets a conclusion, score, trend, buy and sell levels, risk alerts, catalysts and an action checklist.
- **Multi-market data**: Covers A-shares, Hong Kong, US, Japan, Korea, Taiwan and ETFs. It pulls quotes, candlesticks, indicators, news and announcements from several sources.
- **Strategy Q&A agent**: A chat page supports multi-turn questions with 15 built-in strategies, such as moving averages and wave theory. Custom strategy files are supported, and multi-agent mode is experimental.
- **Web and desktop workspace**: Run manual analyses, watch task progress, read history and full Markdown reports, backtest, track holdings and edit config. It has light and dark themes.
- **Flexible LLMs and alerts**: Works with Gemini, Claude, OpenAI-compatible services, DeepSeek, Qwen and local Ollama models. Alerts go to six channels, including email.

## What you could build

- Receive a scored report on your watchlist in Telegram or Feishu every weekday evening.
- Fork the repo and run scheduled analysis on GitHub Actions without paying for a server.
- Ask the chat agent follow-up questions about one stock using a chosen trading strategy.
- Import a stock list from a screenshot, CSV or Excel file and analyze it in bulk.
- Backtest and track a portfolio in the local web workspace.

## Who it's for

Suits retail investors and Python-comfortable students who want automated stock summaries; you need an LLM API key, a notification channel and basic familiarity with GitHub secrets or .env files.

## Get started

Easiest route: fork the repo, add secrets for an LLM API key, a notification channel and your STOCK_LIST, enable Actions and run the workflow once. To run locally instead, clone it, install requirements, copy .env.example to .env and fill it in, then run main.py.

```
git clone https://github.com/ZhuLinsen/daily_stock_analysis.git && cd daily_stock_analysis
```
```
pip install -r requirements.txt
```
```
cp .env.example .env && vim .env
```

Tags: stocks, llm, quant, notifications, github-actions