22nd of 17 in Agent platforms
daily_stock_analysis
Daily LLM stock analysis for A-share, HK, US and other markets, pushed to chat apps
- Stars
- 66.4k
- License
- MIT
- Last commit
- Oct 2026
- Last release
- Sep 2026
- Language
- Python
24,779 on 25 Mar 202666,354 on 14 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Python app that pulls quotes, K-lines, news and fundamentals for a watchlist across A-share, HK, US, Japan, Korea and Taiwan markets. It has an LLM write a decision report with score, trend, buy/sell levels and risks, then pushes it to WeCom, Feishu, Telegram, Discord, Slack or email. It runs on GitHub Actions, Docker or locally, with a FastAPI web workbench for history, backtests and strategy chat.
Who it is for: Retail investors and self-hosters wanting scheduled LLM watchlist reports
Strengths
- Runs on a GitHub Actions fork with no server; default schedule is weekdays 18:00 Beijing time
- Free data sources (AkShare, Baostock, YFinance) work with zero configuration
- Supports Gemini, Claude, OpenAI-compatible APIs, DeepSeek, Qwen and local Ollama models
- Six push channels plus a web UI with history, backtest, portfolio and 15 built-in chat strategies
Weaknesses
- README is mainly Chinese; English docs are in a separate file
- Free data sources are subject to rate limits and upstream changes; stability not guaranteed
- Report quality depends on a paid news-search API key; a search service is recommended
- Output is for study only; the disclaimer excludes investment advice and liability
What it needs
- no GPU
- Docker + Compose
- Needs LLM API key or Ollama, News search API (Anspire, SerpAPI, Tavily, Bocha, Brave, MiniMax or SearXNG), Market data sources (AkShare, Baostock, YFinance; optional TickFlow, Tushare, Longbridge), Notification channel
- Models: Gemini, OpenAI-compatible, DeepSeek, Qwen, Claude
- port 8000
Head to head
Also in Agent platforms
See all 17| Rank | Project | Score |
|---|---|---|
| 1 | LangflowVisual builder for AI agents and workflows, exposed as API or MCP server | 80 out of 100 |
| 3 | deer-flowAgent harness that runs sub-agents, memory and sandboxed skills for long tasks | 75 out of 100 |
| #4 | DifyVisual builder for LLM workflows, RAG pipelines and agents, with APIs | 74 out of 100 |
| #5 | MulticaIssue board where AI coding agents take assignments like teammates | 72 out of 100 |
| #6 | AutoGPTBlock-based builder for agents that run on demand, schedule or trigger | 71 out of 100 |