What problem does it solve? When a stock you hold or watch suddenly spikes or crashes, you need to know within minutes whether the move reflects a real fundamental change or just noise. This Skill deploys four parallel research agents to scan company events, regulatory changes, industry dynamics, and market sentiment, then synthesizes an attribution report explaining the true cause of the price movement. ## Core Features & Use Cases - Four-Dimension Parallel Reconnaissance: Simultaneously investigates company announcements, regulatory/policy shifts, competitor and industry dynamics, and market sentiment (analyst ratings, short interest, influencer opinions). - Attribution-First Output: Produces a weighted event timeline, candidate-explanation table with confidence levels, and a verdict classifying the move as value event, sentiment fluctuation, unknown cause, or mixed. - Xueqiu Scraper Integration: Optionally pulls posts from influential investors like Duan Yongping via a Playwright-based scraper when the stock is in their coverage. - Use Case: Your holding drops 12% in three days. Run this Skill to get a 10-minute attribution report distinguishing a genuine earnings miss from sector-wide beta, plus a checklist of follow-up actions like triggering a thesis review. ## Quick Start Ask the agent to run a news pulse attribution on a company, for example: "Run finance-news-pulse on Pinduoduo for the last 14 days to explain the recent 10% drop."
