Receive AI-driven stock analysis via messaging apps
Enter a stock code, and the AI synthesizes that day's news and price trends to reach conclusions like 'buy' or 'wait.' This system gathers data from A-shares, US, Japanese, Korean, and Taiwanese markets, sending analysis results at scheduled times each day. You do not need to write complex code; once configured, you can check the daily judgments in the morning or evening.
The daily 'decision dashboard'
The core of this system is the 'decision dashboard' generated every day. It does not simply report whether the price went up or down; instead, the AI synthesizes various information to suggest whether to buy, sell, or wait, providing a score for each stance. For example, if the AI judges a specific stock as 'wait,' it displays concrete reasons such as 'need to digest short-term profit-taking selling pressure' or 'outflow of funds from major shareholders.' It also organizes the latest news, earnings outlook, and risk factors related to that stock at a glance. This allows investors to pick out only the important information from the daily flood of data.
Stocks from multiple countries in one place
This tool supports A-shares (China), H-shares (Hong Kong), US, Japanese, Korean, and Taiwanese stocks and ETFs. It automatically selects data sources suited to each market's characteristics to collect stock prices, charts, technical indicators, news, disclosures, and financial information. With default settings, it can run using free data sources, so there is no initial cost. However, free sources may be unstable due to network conditions or service changes, so if you want more stable analysis, it is recommended to additionally configure paid data source tokens. This allows for more precise data and stable execution.
How to start without installation
For users who find it burdensome to manage a server directly, a deployment method using GitHub Actions is provided. By forking the repository and setting the AI model API key and notification channel (Enterprise WeChat, Telegram, Slack, etc.) address, you can complete deployment within 5 minutes. It also supports running directly on a local computer or deploying using Docker, and provides features to manually run analysis via a web browser or view past reports.
Asking the AI directly
Beyond simply receiving reports, it features an interactive function that allows you to ask the AI agent directly. You can ask about specific technical indicators or strategies, such as 'What is the moving average crossover status of this stock?' and the AI answers by referring to real-time stock price and news data. It generates answers based on 15 or more built-in strategies, including moving averages, wave theory, and event-driven approaches, and you can obtain more specific information through multiple rounds of conversation. This feature is available via the web interface, bots, or API.
What you must know before using it
This project is provided for learning and research purposes and does not replace official investment advice. AI judgments may vary depending on data and models, so final investment decisions should be left to your own judgment. Additionally, when using free data sources, service interruptions or data errors may occur, which can affect system stability. AI model API keys and data source tokens are provided at personal cost, so you must check the pricing plans of each service before use. The license is MIT, and if you modify or reuse the code, you must specify the original source.
