Summary
This is a remarkably thorough bridge between natural language and China's Tushare financial data platform. Instead of forcing you to remember API names, it maps conversational requests like "how's this stock doing lately" or "which sector is strongest" into proper data workflows covering A-shares, financials, fund flows, sector rotations, and macro indicators. The author clearly spent time modeling real research tasks rather than just wrapping endpoints. It does environment checks before running queries, handles entity resolution for stock names, and defaults to sensible time windows when you don't specify. Output isn't raw tables but structured summaries with context. Worth noting it requires a Tushare token and won't give trading advice, just data retrieval and light analysis for Chinese equity research.
Install
npx skills add https://github.com/waditu-tushare/skills --skill tushare-data
