Humanize AI Writing
Audits and rewrites content to eliminate machine-generated patterns and 'AI-isms' for more natural, human-sounding text.
This skill provides a comprehensive framework for identifying and removing common AI writing tropes, such as overused vocabulary, excessive formatting, and predictable sentence structures. By operating in both detection and rewrite modes, it helps content creators, researchers, and developers ensure their prose feels authentic and engaging. It uses a tiered flagging system to minimize false positives while systematically improving readability and removing the hallmarks of large language model output, ensuring your content resonates with human readers.
Key Features
01Tiered vocabulary filtering targeting specific AI-only words like 'delve', 'tapestry', and 'realm'.
02Automated formatting cleanup including em-dash reduction and bolding constraints.
03Detailed diff summaries that explain exactly what was changed and why.
04Dual-mode operation: 'Detect' for auditing and 'Rewrite' for automated humanization.
051,851 GitHub stars
06Structural analysis to remove hollow intensifiers, hedging, and generic transition phrases.
Use Cases
01Cleaning up technical documentation to remove repetitive AI-generated fillers.
02Refining AI-generated blog posts or social media drafts to sound more authentic.
03Auditing academic or social science research papers for professional, human-led prose.