A specialized text refiner that eliminates AI patterns and adapts writing for specific social platforms with natural human rhythm.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install content-humanizer-zh
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install content-humanizer-zh using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
Content Humanizer is a powerful utility designed to bridge the gap between machine-generated drafts and authentic human expression. Within the ecosystem of Openclaw Skills, this tool acts as a sophisticated editor that identifies and removes over 24 distinct AI writing markers, such as formulaic transitions and repetitive sentence structures. It ensures your output feels organic, engaging, and devoid of the common robotic smell that often characterizes large language model outputs.
Beyond simple editing, the skill understands the nuances of different digital environments. Whether you are publishing to a professional platform like Zhihu or a lively social space like Little Red Book, Content Humanizer adjusts vocabulary, emoji density, and paragraph length to match the specific cultural expectations of your audience. This makes it an indispensable tool for creators using Openclaw Skills to maintain high-quality, believable content at scale.
To integrate this capability into your Openclaw Skills workflow, ensure you have the Python environment ready and use the following commands:
# Basic text humanization
python scripts/humanize.py "Insert AI text here"
# Target a specific platform style
python scripts/humanize.py "Insert text" --platform xiaohongshu
# Check the humanization score of a text block
python scripts/humanize.py "Insert text" --score
The skill manages its transformation logic through structured reference files and scoring matrices:
| Component | Description |
|---|---|
references/platform-styles.md |
Style definitions for InStreet, Little Red Book, WeChat, and Zhihu. |
references/ai-patterns.md |
A taxonomy of 24+ AI patterns including high-frequency vocabulary and formulaic syntax. |
Quality Score |
A 50-point scale evaluating Directness, Rhythm, Trust, Authenticity, and Conciseness. |
Loading
A high-performance skill that leverages Raspberry Pi native hardware PWM interfaces to provide smooth and precise servo motor control.

Shaper is a project management skill that enables AI agents to execute the Shape Up methodology by interacting with workspaces, pitches, and hill charts via MCP.

E-SafeNet provides a secure, encrypted environment for AI agents to handle sensitive credentials and private data during automated workflows.

A specialized skill enabling AI agents to manage leads, automate email campaigns, and track customer interactions via Followup.

A global e-commerce search tool that automatically injects affiliate tracking codes into product links to generate passive commission income.

An intelligent insurance brokerage assistant designed for product discovery, lead registration, and comprehensive policy support.








































