What problem does it solve? Drafts of docs, READMEs, site copy, and chat replies often carry the stock LLM writing voice — em-dash sprinkling, negation-contrast phrasing, hype adjectives, and announced-honesty preambles — which reads as machine-made and undercuts reader trust. This Skill provides a standing style guide that catches those tells and enforces plain, direct prose. ## Core Features & Use Cases - Plain English base rules: Enforces short sentences (15–20 words), active voice, everyday words, and a substitution table for bureaucratic phrasing, based on the Plain English Campaign's guidelines. - LLM-voice tell detection: Lists specific machine-voice fingerprints to cut, including em-dash overuse, "not X, it's Y" constructions, colon reveals, anthropomorphized tools, and rule-of-three padding. - Reader-surface structure rules: Separates claims from proof — homepage and README lead with the benefit and link out to benchmark write-ups for full evidence. - Use Case: After drafting a new section of a README or a benchmark write-up, run the editing workflow to split long sentences, strip filler, and verify the benefit appears before the methodology. ## Quick Start Review this README draft against the plain-prose rules and rewrite any sentences that carry LLM-voice tells.
