What problem does it solve? AI-generated and AI-assisted prose carries recognizable statistical fingerprints β negative parallelism ("not X but Y"), em-dash overuse, rule-of-three lists, puffery vocabulary, and uniform sentence cadence β that make text read as generic machine output. This Skill mechanically detects those tells and rewrites the text into its target genre without reintroducing new ones. ## Core Features & Use Cases - Mechanical tell detection: Runs runnable grep -Ein regex patterns from a fixed catalog covering negative parallelism, puffery, hedging, false ranges, formatting, and cadence β never relying on subjective judgment. - Meaning-based rewriting: Triages every finding (delete strawman negations, earn real contrasts, cut empty claims) instead of paraphrasing the pattern into a disguised equivalent. - Verify loop and register check: Re-scans every rewrite until zero pattern hits (capped at 4 passes), then validates the result against genre profiles for academic, tweet, reddit, LinkedIn, email, blog, marketing, and docs. - Use Case: You drafted a LinkedIn post with an AI assistant and it reads like every other LinkedIn post. Invoke the skill to get a findings table (categories, counts, worst examples), a rewritten version in your voice, and a change log of what was fixed. ## Quick Start De-slop this draft and rewrite it as a reddit comment, then show me the findings table and change log.
