Daily Stock Analysis

Daily Stock Analysis is a Python workflow for running scheduled A-share market reviews, watchlist analysis, news enrichment, and multi-channel notification summaries.

This repository is packaged as a public portfolio version. It keeps the automation structure visible, but credentials, local databases, logs, generated reports, and private runtime artifacts are intentionally excluded.

What It Does

• Builds a daily stock-watchlist review from configurable tickers.
• Pulls market data from public and optional provider APIs.
• Adds AI-assisted analysis through user-provided model credentials.
• Sends summaries through optional channels such as WeChat Work, Feishu, Telegram, email, or custom webhooks.
• Supports local runs, Docker runs, and GitHub Actions scheduling.

Repository Layout

main.py                  CLI entrypoint for daily analysis runs
config.py                Environment-driven configuration loader
analyzer.py              Analysis orchestration and AI prompt flow
llm_client.py            Gemini and OpenAI-compatible provider adapter
notification.py          Notification rendering and delivery helpers
storage.py               SQLite persistence helpers
data_provider/           Market data provider adapters
sources/                 Public demo assets
.github/workflows/       CI and scheduled workflow templates

Quick Start

py -3 -m venv .venv
.\.venv\Scripts\Activate.ps1
pip install -r requirements.txt
Copy-Item .env.example .env
python test_env.py
python main.py --help

Fill .env with your own API keys and notification endpoints before running a live analysis. The repository does not include working credentials.

Configuration

Common environment variables:

Variable	Purpose	Required
STOCK_LIST	Comma-separated watchlist, such as 600519,300750,002594	Yes
GEMINI_API_KEY	Google AI Studio key for AI analysis	One AI key required
OPENAI_API_KEY	OpenAI-compatible fallback key	Optional
OPENAI_BASE_URL	OpenAI-compatible endpoint	Optional
TUSHARE_TOKEN	Optional Tushare Pro token	Optional
TAVILY_API_KEYS	Optional news search keys	Optional
WECHAT_WEBHOOK_URL	Optional WeChat Work webhook	Optional
FEISHU_WEBHOOK_URL	Optional Feishu webhook	Optional
TELEGRAM_BOT_TOKEN / TELEGRAM_CHAT_ID	Optional Telegram channel	Optional
EMAIL_SENDER / EMAIL_PASSWORD / EMAIL_RECEIVERS	Optional email delivery	Optional

Do not commit .env, databases, logs, or generated reports.

GitHub Actions

The workflow templates under .github/workflows/ are intended as examples. To use them in your own fork:

1. Add the required secrets in GitHub Actions settings.
2. Review the scheduled run time and stock list.
3. Run the workflow manually once before enabling a schedule.

No repository secret is included in this public package.

Public Safety

This public version excludes:

• .env and local credential files
• SQLite databases and generated reports
• runtime logs
• private notification endpoints
• local machine paths

The diagnostic script reports whether credentials are configured, but it does not print token prefixes or secret values.

Status

This is a legacy automation project packaged for public review. For a cleaner current stock-research workflow kit, see Stock Analysis Plus.

License

MIT License.