Analysis updated 2026-05-18
Rewrite AI-generated blog posts, emails, or documentation to sound more natural and conversational.
Match your personal writing voice by providing a sample so rewrites reflect your style instead of generic prose.
Remove corporate jargon and AI-isms like false significance, vague attributions, and overused em dashes from generated text.
| blader/humanizer | handlebars-lang/handlebars.js | mysqljs/mysql | |
|---|---|---|---|
| Stars | 18,629 | 18,631 | 18,631 |
| Language | — | JavaScript | JavaScript |
| Setup difficulty | easy | easy | easy |
| Complexity | 2/5 | 2/5 | 2/5 |
| Audience | vibe coder | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Humanizer is a skill, a reusable instruction set, for AI coding tools like Claude Code and OpenCode. Its purpose is to take text that sounds like it was written by an AI and rewrite it to sound more natural and human. Once installed, you invoke it with a slash command and paste the text you want rewritten. The skill is based on a detailed catalog of 29 patterns that AI-generated writing commonly exhibits, things like significance inflation (calling routine events "pivotal moments"), vague attributions ("experts believe"), the overuse of em dashes and bold text, sycophantic openings, filler phrases, and an over-reliance on the "rule of three." The skill goes through these patterns and removes or rewrites them. It includes a second pass specifically to catch AI-isms that survived the first rewrite. An optional voice calibration feature lets you provide a sample of your own writing first, so the rewrite matches your personal style rather than producing generic clean prose. Installation requires cloning the repository into the skills directory of your AI tool.
A reusable instruction set for AI coding tools that rewrites AI-generated text to sound more natural and human, using a catalog of 29 common AI writing patterns.
Use freely for any purpose including commercial, as long as you keep the copyright notice.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly vibe coder.
This repo across BitVibe Labs
Verify against the repo before relying on details.