East Money MX Finance Data Query

Problem

Financial data queries often break down in three places: unclear entity scope, inconsistent metric wording, and results that are not table-ready. A prompt such as “query the last-year revenue of Kweichow Moutai and Wuliangye” or “what are the price changes for these stocks” can mix entity names, pronouns, and modifiers if passed directly to an API. mx-finance-data converts natural-language questions into structured query parameters and returns deliverable files.

How It Works

The skill covers stocks, sectors, indexes, shareholders, issuers, bonds, fund markets, and bond markets, and supports real-time quotes, quantitative data, and financial reporting data. Typical invocation uses:
- --query: a natural-language question that includes every entity name, such as “query the PE and market cap for A, B, and C.”
- --indicators: metric names and time scope extracted from the question, such as “last-year revenue” or “price change, PE, market cap.”

It first identifies entities. If there are ≤ 5 entities, it queries directly. If there are > 5, it uses a batch path and processes at most the first 500 valid entities. It then outputs:
- .xlsx: a multi-sheet structured workbook, with sheets by entity or metric combination.
- .md: a Markdown table version of the same content, suitable for notes and reports.

Boundaries and Notes

• It requires the EM_API_KEY environment variable. Before use, confirm the key origin, validity, scope, and revocation process; avoid hard-coding or logging the key in plain text.
• --indicators should contain metrics and time ranges only, not entity names, and should stay close to the user’s original wording.
• Pronouns such as “these stocks” or “the companies above” must be resolved to explicit entity names from context or a file.
• Too many metrics, too wide a date range, or more than 500 entities can cause limits, missing data, or batch handling; split queries by metric, date range, or entity count.