Writing Anti Ai

Based on Wikipedia's "Signs of AI writing" guide, this detects and rewrites seven categories of AI slop: inflated symbolism ("serves as a testament"), promotional language, vague attributions ("experts believe"), superficial -ing constructions, AI vocabulary ("delve", "leverage"), negative parallelisms ("it's not just X, it's Y"), and excessive conjunctive phrases. Works in both English and Chinese. The interesting part is it goes beyond pattern removal to address rhythm and voice, because sterile rewriting is just as obviously artificial. Includes scoring rubrics and before/after examples. Honestly, this is more useful as a checklist for catching your own AI-isms than as an automated fixer, since good humanization requires judgment about when to break the rules.

Install to Claude Code

npx -y skills add galaxy-dawn/claude-scholar --skill writing-anti-ai --agent claude-code