tushare

The Tushare skill is designed to convert Chinese natural language queries related to financial data into executable workflows for data retrieval, cleaning, comparison, filtering, exporting, and basic analysis. It addresses the challenge of accessing and analyzing a wide range of financial information for A-shares, indices, ETFs/funds, company financials, valuations, capital flows, announcements, news, sector trends, and macroeconomic data, all without requiring the user to know specific APIs or field names. By interpreting user intent, the skill bridges the gap between human language requests and technical data operations.

This skill offers a variety of capabilities including retrieving recent stock or index trends, checking company fundamentals, financial metrics, and valuation indicators, performing cross-company or sector comparisons, analyzing capital flows and market sentiment, and compiling announcements, news, and macroeconomic indicators. Users can also export datasets in CSV or parquet formats for further analysis or backtesting. It emphasizes generating concise research summaries instead of only presenting raw data tables. Additionally, it includes environment checks for Python version, required packages, and token availability to ensure smooth operation.

Tushare is ideal for financial analysts, researchers, and data scientists who need to quickly access and interpret Chinese financial market data. Common use cases include analyzing recent stock performance, evaluating company financials and valuations, tracking sector or thematic trends, monitoring capital movements, summarizing policy and news impact, and preparing datasets for quantitative research or backtesting. The skill is intended for research and analysis purposes and is not a substitute for investment advice, trading execution, or high-frequency trading systems.