What problem does it solve? Most long-form financial articles fail on fabricated precision, subjective probability weighting, and absolute claims. This Skill produces a 1-8 part deep-dive series on a single public company (the "Understanding X" format) with a rigorous fact-checking checklist that catches cross-article number inconsistencies, double-counted holdings, and misleading valuation comparisons before publication. ## Core Features & Use Cases - Adaptive Series Structure: Scales from 1 to 8 articles based on company complexity, with fixed per-article skeletons covering moat analysis, profit engines, hidden assets, financials, management integrity, and valuation scenarios. - Fact-Check Checklist: Enforces 7 revision checks including cross-article number consistency, accounting basis labeling (Non-IFRS vs GAAP), duplicate-counting scans, and removal of probability-weighted expected returns. - Revision Workflow: Handles user feedback through a triage system (hard errors, subjective wording, granularity, unreliable third-party data) with cascading updates across all articles when a shared figure changes. - Use Case: Ask the agent to write a deep series on Tencent; it researches annual reports and sell-side notes, drafts 3 articles totaling ~60,000 words, runs a cross-article consistency scan, and only pushes to Git after your review. ## Quick Start Ask the agent to write a "Understanding [company name]" deep-dive series for publication on WeChat or similar channels.
