Humanize Chinese Text
Detects and rewrites AI-generated Chinese text to improve naturalness, reduce AIGC detection rates, and adapt content for specific social media or academic styles.
This skill provides a specialized workflow for identifying and eliminating the synthetic 'AI flavor' often found in large language model outputs in Chinese. It systematically detects rigid structures, mechanical connectors, and repetitive sentence rhythms before applying targeted rewrites that preserve factual integrity while enhancing readability. Whether you are preparing academic papers for CNKI/VIP checks or adapting content for platforms like Zhihu or Xiaohongshu, this skill offers practical strategies to make Chinese text sound authentic, varied, and human-authored.
Key Features
0140,674 GitHub stars
02Targeted AIGC reduction for academic papers (CNKI, VIP, and Wanfang styles)
03Multi-platform style conversion including Zhihu, Xiaohongshu, and WeChat
04Advanced pattern detection for common AI-writing markers in Chinese
05Minimal-pass rewriting strategy to avoid unnecessary content drift
06Preservation of discipline-specific terminology and factual claims
Use Cases
01Converting generic summaries into platform-specific content for social media engagement
02Reducing AIGC signals in academic theses and research papers to meet submission requirements
03Polishing AI-generated marketing copy to sound more natural and engaging for Chinese audiences