Removes AI-writing patterns and LLM-isms to restore a natural, human voice to your prose.
De-Slop is a specialized Claude Code skill designed to identify and eliminate the characteristic 'tells' of AI-generated text. It uses a structured three-tier diagnostic process—categorizing issues as Red (immediate flags), Yellow (statistical signals), and Green (structural patterns)—to address everything from obvious chatbot leakage to subtle monotonous sentence structures. By applying principles of plain language, varied sentence length, and specific examples, this skill helps users transform robotic AI drafts into authentic, engaging content suitable for professional publication.
Características Principales
01Automatic removal of chatbot leakage and grandiose filler phrases
02Synonym de-cycling to prevent repetitive noun usage
034 GitHub stars
04Structural rewriting to break 'relentless balance' and uniform register
05Three-tier diagnostic reporting to identify AI writing artifacts
06Plain language substitution for overused AI words like 'delve' and 'leverage'
Casos de Uso
01Polishing AI-generated documentation to sound more professional and direct
02Humanizing automated summaries and technical README files
03Reviewing blog posts or marketing copy to remove detectable AI patterns
