humanizer-zh

Rewrites Chinese text to remove AI writing patterns and restore natural human style.

1|Updated Mar 25, 2026
One-click install
npx skills add https://github.com/qfwc258/trae-skills --skill humanizer-zh-qfwc258
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: humanizer-zh
Source: https://github.com/qfwc258/trae-skills/tree/main/zh-humanizer
Command: npx skills add https://github.com/qfwc258/trae-skills --skill humanizer-zh-qfwc258

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? AI-generated Chinese text often contains recognizable machine patterns—buzzwords like 赋能 and 闭环, template structures, empty conclusions, and inflated claims—that make content feel generic and untrustworthy. This Skill systematically detects and rewrites these traces so text reads as if written by a real person. ## Core Features & Use Cases - 21-rule detection library: Identifies meaning inflation, fake authority citations, pseudo-deep verbs, advertising tone, template paragraphs, high-frequency AI words, dash and bold overuse, emoji residue, filler phrases, and hollow endings. - Structured rewrite workflow: Follows a fixed pipeline of reading, marking AI traces, rewriting, rhythm optimization, and naturalness checks. - Standardized output: Delivers the rewritten text, a list of remaining AI traces, a final corrected version, and optional change notes. - Use Case: Paste an AI-drafted WeChat article or product announcement full of phrases like 标志着新时代 and 全方位赋能, and receive a concrete, direct version that preserves all facts while sounding naturally written. ## Quick Start Rewrite this Chinese article to remove AI writing traces and make it sound like a real person wrote it.

Frequently Asked Questions about humanizer-zh

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I remove AI writing traces from Chinese text?

Paste the AI-generated Chinese text and ask for a rewrite. The Skill applies 21 detection rules covering buzzwords, template structures, and filler phrases, then outputs a rewritten version plus a list of any remaining AI traces.

What AI writing patterns does this detect in Chinese content?

It detects meaning inflation, fake authority citations, pseudo-deep verbs like 赋能, advertising tone, three-part template structures, high-frequency AI words like 闭环 and 抓手, dash and bold overuse, emoji residue, and hollow endings.

Does rewriting change the original meaning or facts?

No. The core principles require preserving the original meaning and facts while prohibiting the addition of new information. Only expression, structure, and rhythm are modified to improve naturalness and information density.

Can this humanizer handle English text or only Chinese?

The rule library is designed primarily for Chinese-language contexts, with detection patterns targeting Chinese-specific AI vocabulary and rhetorical habits. English text is not the intended scope.

What output format does the rewrite produce?

By default it produces four parts: the rewritten text, a brief list of remaining AI traces if any exist, a final corrected version, and optional change notes explaining the modifications.