This strips AI-generated patterns from text and scores how human it sounds on a 0-10 scale. Built from analyzing 1,000+ AI versus human content pieces, it catches 47 specific patterns like "leverage," "in today's fast-paced world," and excessive "Moreover" transitions. You get three modes: quick for blog posts, standard for anything public (includes scoring and change logs), and deep for matching a specific writer's voice. The case study is legit: a B2B team cut bounce rate 55% after processing posts through it. Saves you from manually rewriting AI slop or spending 15 minutes crafting the perfect "sound more human" prompt every time.
npx skills add https://github.com/brianrwagner/ai-marketing-claude-code-skills --skill de-ai-ify