Guotai Haitong Lingxi Financial Data Search

Problem

Financial data queries often involve scattered indicators, awkward mapping from natural language to field names, and repeated requests for batch comparisons. This skill targets A-share, sector, and index scenarios, turning natural-language questions into executable financial data retrieval. It helps engineers and agents obtain quotes, fundamentals, and technical data within a workflow without manually assembling API parameters.

Capabilities and Workflow

The skill calls the financial-search tool through the Guotai Haitong Lingxi MCP gateway, using query as the main parameter. Supported data includes:

• Real-time quotes: price change, volume, turnover, and other intraday data.
• Fundamentals: revenue, net profit, gross margin, ROE, debt ratio, and more.
• Derived metrics: P/E, P/B, PEG, turnover rate, and other valuation indicators.
• Technical and fund-flow data: MACD, candlestick patterns, support and resistance levels, and main capital flow.
• Batch queries: retrieve multiple tickers or multiple metrics in one request.

Key steps include:

1. Authorization check: an API key is required, stored in the shared gtht-entry.json file.
2. Intent mapping: translate user requests into the query parameter and clarify tickers or sorting rules when needed.
3. Result validation: inspect the returned date field, since “today” may correspond to the previous trading date; display any query conditions in the response.

Boundaries and Caveats

This skill is scoped to basic A-share financial data. It does not cover non-A-share assets, complex research modeling, or unofficial data sources. Failures may be caused by expired API keys, missing permissions, gateway outages, or invalid ticker codes. Cross-platform execution assumes a node environment; Windows PowerShell should use ; instead of && as a command separator. If a request exceeds the service scope, guide users to Guotai Haitong Junhong or the Guotai Haitong Lingxi app.