Stock Catalyst Hunter

The Problem It Solves

When analyzing capital markets, you face突发新闻，政策变动，国际冲突或社交媒体传闻。Your immediate need is to answer two core questions: Is this information real? How does it map to A-shares? Traditional approaches either rely on static concept stock lists (which are outdated) or require manual searching across multiple financial platforms and synthesizing the logic, which is inefficient and risks missing critical transmission chains.

How It Works

The skill employs a structured analysis pipeline to transform raw information into a list of stocks annotated with probability and time horizons. Its core capabilities include:

• Real-time Information Verification and Aggregation: It automatically searches major financial platforms like Cailian Press, Jin10, and East Money to cross-verify information and extract the current market narrative.
• Multi-layer Causal Reasoning: It goes beyond the first-order reaction (Event -> Sector -> Leading Stocks). It defaults to a 2-3 layer deep inquiry:
  • Direct Impact Layer: What product, industry, region, or company is directly hit by the event?
  • Transmission Impact Layer: Does it trigger chain reactions in price, cost, supply-demand, policy, or orders?
  • A-Share Mapping Layer: Which part of the logic is the market likely to trade first (fundamental logic, historical pattern, or sentiment-driven theme)?
• Structured Output: The final result is presented clearly in a Markdown table format, including the stock, correlation probability, time horizon (e.g., 1-3 days, Swing), and core analytical rationale.

Key Execution Steps

The skill follows a default execution flow:
1. Information Extraction: If the input is an image, it first extracts key information like event subject, action, and time based on a template.
2. Real-time Search: It conducts multiple rounds of searches across platforms using the event keywords.
3. Deep Analysis: It applies its built-in reasoning rules (e.g., differentiating the transmission paths of supply disruptions versus geopolitical sanctions), avoiding superficial associations.
4. Stock Mapping and Filtering: It screens stocks based on search evidence and defaults to filtering out ST stocks, illiquid securities, etc.

Scope and Limitations

• It is not a static database. All conclusions are based on real-time search and reasoning. Accuracy depends on the availability and timeliness of public information sources.
• The output is a research framework, not investment advice. Probability and time horizons are designed to assist in judging event relevance and market trading tempo, not as returns promises.
• Requires a clear event input. The skill is designed for the "event-to-A-share mapping" scenario. It works best when provided with specific news, rumors, or image content.
• Searches may be constrained. If major financial platforms are inaccessible or lack sufficient information, the skill activates a degradation strategy (e.g., switching platforms, rewriting queries), which may correspondingly lower the confidence level of its conclusions.