What problem does it solve? Academic manuscripts often fail editorial triage because of weak argument structure, overclaimed results, AI-sounding phrasing, or Chinese-influenced English, and authors lack a systematic way to align prose with Nature-family journal expectations. ## Core Features & Use Cases - Axis-based fragment routing: Detects paper type, section, language, and target journal (Nature, Nature Communications, Nature Machine Intelligence, or generic) and loads only the relevant style fragments from a manifest. - Chinese-to-English academic translation: Rebuilds logic and connectives from Chinese drafts before applying English sentence and paragraph rules, preserving terminology and citation intent. - Consistency and overclaim sweeps: Scans full manuscripts for terminology, unit, and numeric drift, and flags exaggerated claims, causal overreach, and AI-like phrasing. - LaTeX layout repair: Diagnoses loose float pages, stranded headings, and oversized figures via a dedicated typesetting reference with compile-and-render verification. - Use Case: A researcher pastes a Chinese Results paragraph and asks for a Nature-style English version; the skill detects zh-to-en language, research paper type, and results section, then returns polished prose with revision notes and risk flags. ## Quick Start Use nature-polishing to rewrite this abstract into concise Nature-style English without changing any facts or citations.
