Guosen Securities ETF Filter

Problem to solve

When the ETF universe is large, engineers and researchers often need a fast way to compare short-term momentum, long-term dividend and low-volatility profiles, sector conditions, valuation, and liquidity. This skill wraps Guosen Securities' ETF data service into an agent skill, letting an LLM filter products with structured conditions and present the results directly.

How it works

The skill provides two main workflows:
- Curated rankings: request results by classId and listId, including hot sectors, T+0 short-term breakout, high dividend and low volatility, undervalued and quality, and whole-market hot ETFs.
- Custom filtering: cover basic info, trading attributes, returns, risk and volatility, market indicators, fundamentals, and trend heat, with fields such as class1, endamt, tgfandglf, isT0, sharpe1yrank, premiumrate, and temperRegion.

The key flow is to configure a valid data-service API Key, read the persisted credential, set the required environment variable, and then invoke the filtering script. The script produces intermediate xlsx or txt files. The skill is expected to parse those files and present up to 100 ETFs in a table or list, showing only essential fields such as name, code, latest price, real-time change, size, and listing age, without exposing file paths.

Scope and cautions

This skill is for ETF screening analysis and does not provide investment advice. Results depend on Guosen Securities data, API permissions, and indicator definitions, so different rankings or custom parameters may return different sets. Before use, confirm that the API Key is valid, the environment variable name is correct, and interval-style parameters such as 10,50 and multi-period formats such as profit1w:5,6 are understood. Output is limited to the first 100 matching ETFs and should include the required risk notice.