AI Stock Short-Term Analysis Assistant

Problem

Short-term stock analysis often stops at scattered data sources: quotes, fund flows, technical indicators, dragon-tiger lists, news sentiment, and market conditions must be queried separately, then manually combined into a view. This skill turns that workflow into a concrete pipeline: given a ticker or name such as 600519, hk00700, or Apple, it resolves the market and currency, pulls structured data from westock-data, and renders an HTML report with ECharts.

Core workflow

• Data collection: covers A-shares, HK stocks, and US stocks, with basic quotes, sector or single-stock fund flow, MA/MACD/KDJ/RSI/BOLL, K-line volume, news sentiment, and market environment; A-shares also include dragon-tiger list and margin-related data.
• AI reasoning: summarizes main-force behavior, news bias, short-term strategy, and risk points, producing a bullish/bearish/neutral view plus support, resistance, and scenario-based notes.
• Report generation: assembles a 12-section JSON structure, then uses generate_report.py to render a dark terminal-style HTML report with K-line, volume, fund-flow, MACD, and RSI charts.

Boundaries and notes

• It depends on the westock-data data source, Python 3.13+, and an outputs/ directory; without the data source, collection cannot complete.
• HK stocks do not include sector fund flow or dragon-tiger lists; US stocks lack some A-share-specific data and may substitute short-selling data for certain sections.
• Commands such as news article may be unavailable on some channels, requiring a WebSearch fallback for sentiment.
• The output is a data summary and AI-assisted research draft, useful for analysis support but not a direct investment decision.