A sophisticated text editor that identifies and removes robotic AI writing patterns to restore a natural, human voice to your content.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install humanizer
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install humanizer using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Humanizer?
The Humanizer is a specialized editing skill designed to bridge the gap between sterile AI-generated drafts and high-quality human prose. Built on the technical foundations of Wikipedia's "Signs of AI writing" guide, this tool systematically identifies "slop"—the telltale markers of LLM output such as inflated symbolism, repetitive sentence structures, and overused AI vocabulary. By integrating this skill into your workflow, you ensure that your content avoids the algorithmic monotony common in modern LLMs.
Beyond mere pattern removal, the Humanizer emphasizes adding "soul" to writing. It achieves this by varying sentence rhythm, acknowledging complexity, and injecting a distinct personality or first-person perspective where appropriate. Whether you are refining a blog post or technical documentation, this Openclaw Skills integration transforms soulless reporting into engaging, authentic communication that resonates with real readers.
Humanizer Use Cases
- Refining AI-generated blog posts to sound more authentic and less like automated slop.
- Auditing technical documentation to remove vague attributions and superficial analyses.
- Transforming promotional copy by stripping away inflated symbolism and marketing fluff.
- Improving the readability of academic summaries by fixing em dash overuse and Rule of Three patterns.
- Injecting personality and a specific voice into neutral, robotic drafts.
How Humanizer Works
- The skill scans the target text or file directory using tools like Read, Grep, or Glob to identify specific AI-generated markers.
- It analyzes the content against 24 distinct linguistic criteria, including AI Vocabulary, Negative Parallelisms, and False Ranges.
- The AI editor proposes targeted rewrites that replace robotic constructions (e.g., "serves as") with direct, human-friendly language (e.g., "is").
- It applies stylistic adjustments to vary sentence length and structure, ensuring the text doesn't follow a predictable algorithmic rhythm.
- The final output is generated as a humanized version, often accompanied by an optional summary of the specific AI patterns that were corrected.
Humanizer Setup
To begin using this skill within your environment, ensure it is correctly installed and configured. This skill is part of the broader ecosystem of Openclaw Skills.
# Add the Humanizer skill to your local agent
openclaw add humanizer
# Run a scan on a specific file to detect AI patterns
openclaw run humanizer --file README.md
No complex API keys are required for the logic itself, as it utilizes the agent's internal reasoning to apply the documented Wikipedia cleanup standards.
Humanizer Data Schema & Taxonomy
The Humanizer organizes its editing process through a structured analysis of text files. It primarily interacts with Markdown and plain text formats.
| Component | Description |
|---|---|
| Input Formats | Supports .md, .txt, and various documentation source files. |
| Pattern Library | References 24+ distinct markers based on the WikiProject AI Cleanup taxonomy. |
| Tool Access | Utilizes Read, Write, and Edit tools to perform non-destructive updates. |
| Metadata | Tracks versioning (current: 2.1.1) and provides summaries of removed AI-isms. |
Humanizer Advanced Features
- Multi-agent support for auditing large repositories of documentation for AI-generated artifacts.
- Personality injection modules that shift tone from neutral reporting to opinionated, human perspectives.
- Automatic removal of collaborative communication artifacts like "I hope this helps!" or knowledge-cutoff disclaimers.
- Correction of technical formatting issues such as curly quotes, Title Case in headings, and mechanical bolding.
- Integration with Openclaw Skills workflows to automate the humanization of content pipelines.
SKILL.md
Loading
Related Openclaw Skills
Claw Shell Kali
A persistent shell execution skill that manages command-line interactions inside a dedicated tmux session for continuous terminal access.
Research Automation
An automated web research system for discovering and organizing the latest developments in peptides, longevity science, and biohacking protocols.
Paper Fetcher
A dedicated research tool that automatically retrieves and organizes academic papers from Sci-Hub using DOI identifiers.
Image Breaker
A sophisticated tool for converting visual and web-based documents into structured, tagged Markdown notes that automatically sync to your Obsidian vault.
LinkedIn Automation
A browser-based AI skill for managing LinkedIn professional networking, messaging, and profile research through automated relay or isolated sessions.
E-commerce Price Comparison
A sophisticated tool for crawling and comparing product prices across major Chinese e-commerce platforms like JD, Taobao, and Pinduoduo.
Featured*
