Refines AI-generated text by removing predictable patterns and injecting natural human personality and rhythm.
The Humanizer skill transforms robotic, AI-sounding text into natural, engaging prose based on Wikipedia's comprehensive 'Signs of AI writing' research. It systematically identifies and fixes common LLM tells—such as overused vocabulary, 'rule of three' structures, and sterile analyses—while adding 'soul' through varied sentence lengths, specific emotional cues, and authentic personality. It is an essential tool for editors, technical writers, and developers who need their documentation, blog posts, or communications to sound genuine rather than algorithmic.
主要功能
01Eliminates superficial '-ing' analyses and vague weasel-word attributions
02Preserves core meaning while matching intended tone (formal, casual, or technical)
031 GitHub stars
04Injects personality and 'soul' by varying sentence rhythm and adding opinionated nuances
05Detects and removes high-frequency 'AI vocabulary' like 'delve', 'tapestry', and 'pivotal'
06Fixes structural patterns like em dash overuse and the repetitive 'rule of three'
使用场景
01Editing README files and repository descriptions to sound more authentic and less generic
02Polishing AI-drafted blog posts or articles to pass manual quality reviews
03Converting technical documentation from sterile and robotic to approachable and human-centric
