Stock Analysis for Tongdaxin

Problem it solves

In equity research workflows, data often comes from separate sources: quotes, financials, disclosures, broker reports, and macro series. Manually mapping securities to codes, markets, periods, adjustment flags, and tool parameters is easy to get wrong. @jiangyun/tongdaxin-stock-analysis turns Tongdaxin capabilities into one MCP service, tdx-finance_qclaw, so an agent can call structured tools when a user asks for quotes, K-lines, fundamentals, screening, reports, notices, news, or macro data.

How it works and what it covers

The skill is driven through mcporter and uses the HTTP MCP server rather than direct curl calls. It exposes four main capability groups:
- Quotes and K-lines: tdx_quotes fetches real-time prices, order book depth, turnover, and related data for A-shares, HK stocks, US stocks, indices, and sectors; tdx_kline retrieves 5-minute, 15-minute, 30-minute, hourly, daily, weekly, and monthly bars, with adjustment options.
- Fundamentals and financials: tdx_api_data covers preset datasets such as income statements, balance sheets, cash flow, share capital, shareholder changes, institutional holdings, dragon-tiger lists, research ratings, industry comparisons, and valuation history; tdx_indicator_select answers indicator queries like P/E and P/B in natural language.
- Screening and lookup: tdx_lookup_stock resolves names to security codes; tdx_screener accepts natural-language filters such as “limit-up,” “MACD golden cross,” or “main inflow.”
- Research and macro: wenda_notice_query, wenda_report_query, and wenda_news_query query company notices, broker reports, and news; wenda_macro_query queries GDP, CPI, social financing, rates, exchange rates, and trade data.

Before business calls, the skill initializes by obtaining a token, checking whether mcporter already has tdx-finance_qclaw configured, and creating or refreshing the configuration if needed. If calls return 401/403, it restarts the token and configuration flow.

Boundaries to keep in mind

This skill is best suited for querying, comparing, and summarizing structured market data. It is not a source of investment advice. Returned market data can be delayed, so timestamps should be surfaced when results are used. Some higher-tier data services, such as Level-2 quotes or real-time money flow, may not be available through the current tool set. For individual stock analysis, outputs should still be framed as informational only.