Xueqiu Data Fetching

Problem to solve

Investment research workflows often need Xueqiu data such as users, portfolios, posts, NAV, and rebalancing details. Scraping pages directly can lead to auth issues, rate limits, and inconsistent fields. @user_6eb4c6c0/xueqiu-skill wraps Xueqiu-related endpoints into callable skill actions, so agents or scripts can request data by parameters.

How the skill works

The skill sends POST JSON requests to a configured base URL, with common fields app_id and secure_key. Key capabilities include:
- Search: searchUser, searchCube, searchStatus, and searchCubeRanked find users, portfolios, and posts, with ranked filtering by gains.
- Quotes: getCubeQuote returns real-time portfolio NAV; getCubeNavHistory returns daily historical NAV.
- Details and history: getCubeInfo returns portfolio metadata, holdings, and owner info; getCubeRebalancing returns rebalancing time, cash ratio, and stock changes.
- User dimension: getUserCubes and getUserTimeline list a user's created portfolios and original posts.
A typical flow is to obtain credentials, choose the endpoint, pass parameters such as keyword, symbol, cubeId, or userId, then handle pagination, HTML descriptions, null values, and error codes.

Boundaries and notes

The skill is for data retrieval, not investment advice. Post description is HTML and should be cleaned before display. Full URLs should be built as https://xueqiu.com{target}. Some legacy portfolios or delayed sync may return null fields. Successful calls are billed; failed calls are refunded. The shared upstream limit is 10 requests per 60 seconds, and clients should pace requests around every 6 seconds. Handle 401, 402, 429, 502, and 503 by checking credentials, balance, rate limiting, or retrying later.