WeChat Article De-AI Rewrite

Problem

It addresses the rhythm problem in AI-generated WeChat public account articles, not just synonym swapping.

AI drafts often sound mechanically correct but unnatural: template-like openings, heavy parallelism, sentence-ending aphorisms, dense transition words, uniform paragraph lengths, and uplifting endings. This skill reframes the task as rewriting the draft into an article a human author would plausibly publish, rather than lightly humanizing short phrases.

How it works

• Trace diagnosis: scans for clichéd openings, forced parallel structures, stacked quotable lines, first / second / in summary style transitions, vague data, and generic conclusions.
• Author voice: builds a lightweight voice profile from account type or existing articles, including preferred openings, person, recurring phrases, and ending style.
• Structural rewrite: rewrites the opening and ending first, merges or splits paragraphs, and removes unnecessary 'inspirational' sections.
• Expression replacement: turns abstract judgments into concrete cases, numbers, scenarios, and personal observations, such as replacing broad claims with specific examples.
• Account-type adaptation: keeps facts and data for knowledge articles, adds story detail for emotional writing, adds drawbacks and price comparisons for reviews, and adds counterpoints for business or finance pieces.
• Final calibration: checks paragraph rhythm, transition density, parallelism, and whether the text keeps a consistent authorial tone.

Boundaries

It fits converting AI drafts into more natural WeChat-style articles, but it should not invent nonexistent stories, data, or author experiences. Clear source material, account positioning, and fact checks are required. Human review is still needed for industry data, quoted people, promotional intent, and platform-sensitive language.