Finance Intel Briefing

Problem

Financial news is scattered across announcements, market feeds, research notes, and stock-related platforms. A single item often lacks a consistent interpretation: is it positive, negative, or neutral? Does it affect the market, an industry, or a company? Which A-share, H-share, or N-share tickers are involved? Finance Intel addresses this by turning raw headlines into structured, comparable briefs.

How It Works

The skill uses requests and beautifulsoup4 to fetch content from 12+ financial sources, storing caches and reports in a working directory. The standard flow starts by identifying keywords such as NEV, CATL, Tencent, or Fed; then it calls fetch_news.py with --keyword, --market, --source, and --limit; finally it parses sentiment, impact_level, related_stocks, and operation_advice into a brief.

• Sentiment: rules classify news as positive, negative, or neutral.
• Impact level: central banks, GDP, and interest rates map to market-level; chips, NEVs, and finance map to industry-level; otherwise company-level.
• Stock linkage: detects related A-share, H-share, and N-share names, such as 0700.HK and 9988.HK.
• Advice: outputs prompts like watch, caution, or wait and see.

Boundaries

It fits pre-market scans, industry tracking, and stock-news summaries where engineers need clean, repeatable information. It is not a replacement for research models or trading decisions. Fetching requires network access, and sentiment is keyword-based with roughly 70% to 80% accuracy, so complex contexts need manual review. operation_advice is reference-only and not investment advice.