WeChat Finance Article Writer

Problem

Writing finance articles for WeChat public accounts often fails not on length, but on reasoning quality: unsourced facts, causal claims based only on timing, data piled into paragraphs, and language full of “blockbuster” or “explosive” claims. This skill turns the process into checkable constraints, requiring core judgments to come from traceable facts and explicit reasoning chains, while distinguishing confirmed facts, reasonable inferences, and speculation.

How It Works

It is built for content-creation tasks: given a finance topic, it produces a WeChat article of no more than 1,500 words, with a calm, formal, data-driven analytical style.

The main workflow includes:
- Topic review: checks the angle distribution of existing articles, assesses homogeneity and differentiation, then confirms the final angle.
- Research and framing: searches announcements, filings, reports, and reputable outlets; classifies sources from T1 to T4; confirms key data with independent sources.
- Draft writing: writes article-draft.md, requires two judgment-bearing ## subheadings, and avoids numbered headings, emojis, bullets, and tables.
- Reverse validation: rereads the full draft to catch unsupported causality, unbounded extrapolation, naked analogies, suggestive juxtapositions, and attribution overreach.
- Fact checking and polish: revises based on the audit report, then optimizes language for restrained, high-density prose.

Scope and Limitations

It fits listed-company analysis, industry trends, macroeconomic analysis, and finance-related current events, not marketing copy, trading advice, or insider information. When evidence is thin, data definitions conflict, or facts cannot be verified, the workflow asks the writer to flag uncertainty instead of filling gaps with invented conclusions.