DSA

Overview

DSA is a research system that combines multi-source market data, news and event context, technical indicators, fundamentals, and LLM reasoning to produce watchlist decision reports and daily market reviews. It supports A-share, Hong Kong, and US markets, offers strategy Q&A for individual symbols, and can deliver reports through several notification channels.

Key Features

• Cross-market quotes for A-share, Hong Kong, US stocks, and common ETFs through multiple data providers and fallback paths.
• News, filings, search results, and social sentiment aggregated to provide context for earnings, themes, flows, and risk events.
• LLM decision reports that turn quotes, news, technicals, and fundamentals into structured conclusions with score, trend, advice, risk alerts, and a checklist.
• Agent strategy Q&A covering moving averages, Chan theory, wave theory, trend views, and follow-up questions from Web, CLI, or Bot entry points.
• Reports stored in SQLite for direction accuracy, simulated returns, portfolio review, and market-review discipline checks.
• Multi-channel delivery of watchlist reports and market reviews through WeChat Work, Feishu, Telegram, Discord, Slack, email, and more.

Use Cases

• Daily watchlist research: generate per-symbol summaries, scores, trends, risk alerts, catalysts, and action checklists.
• Market review: summarize indices, market breadth, sector strength, market temperature, and trading-day status by market.
• On-demand strategy Q&A: ask about a symbol and combine quotes, indicators, news, and built-in strategies into a research view.
• Scheduled delivery: send watchlist and market reports to team chat, email, or other notification channels.

Getting Started

• Use GitHub Actions: fork the repository, configure Secrets for at least one AI key and one notification channel, set STOCK_LIST, enable Actions, and run the Daily Stock Analysis workflow.
• Run locally: clone the repository, create a Python virtual environment, install requirements, copy .env.example to .env, and run python main.py --stocks AAPL,MSFT,NVDA.
• Deploy with Docker: use the local Compose build or the official Docker Hub or GHCR images after setting STOCK_LIST, an AI key, and a notification channel in .env.

Deployment & Requirements

• Python 3.10+ for local runs.
• At least one AI model API key, such as Gemini, OpenAI-compatible, Anthropic, or an Ollama local model for local or Docker deployments.
• STOCK_LIST configured with watchlist symbols.
• At least one notification channel or local inspection path.
• For Docker, use docker/docker-compose.yml or the official images.

Before You Adopt

• License: MIT. Review its terms before using, modifying, or distributing the project.
• Built-in free quote sources can be affected by upstream rate limits, API changes, and network fluctuations, so stability is not guaranteed.
• News-source configuration materially affects the quality of sentiment, announcements, events, and catalyst analysis; at least one search service is recommended.
• The project describes itself as a research system rather than investment advice.