A specialized editing skill that identifies and removes robotic patterns from Chinese AI-generated text to ensure it sounds natural and authentic.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install humanizer-zh-op7418
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 humanizer-zh-op7418 using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
Humanizer-zh is a developer-centric tool designed to combat the statistical average tone of Large Language Models. By implementing research from the WikiProject AI Cleanup, this skill systematically strips away common AI signatures—such as exaggerated symbolism, hollow promotional language, and formulaic sentence structures—specifically for the Chinese language. Using this skill within Openclaw Skills allows teams to maintain high-quality, personality-driven content that bypasses the mechanical feel of raw AI output.
The skill doesn't just delete bad patterns; it actively works to vary sentence rhythm and inject subjective perspective where appropriate. This ensures that the final text feels like it was written by a human with a clear point of view, rather than a neutral machine repeating the same 2023-era vocabulary lists.
To integrate this skill into your workflow via Openclaw Skills, ensure your agent has access to the standard file manipulation tools.
# Standard setup within an agent environment
openclaw install humanizer-zh
# To process a specific file
openclaw run humanizer-zh --path ./docs/ai-draft.md
The skill processes text through a structured evaluation framework, often returning a quality scoring table as metadata. This allows users of Openclaw Skills to benchmark their content quality.
| Dimension | Evaluation Criteria | Max Score |
|---|---|---|
| Directness | Measures factual clarity vs. circular declarations | 10 |
| Rhythm | Checks for variation in sentence structure and length | 10 |
| Trust | Evaluates if the text respects reader intelligence | 10 |
| Authenticity | Scores how natural the tone sounds when read aloud | 10 |
| Refinement | Identifies the absence of redundant AI filler | 10 |
Loading
A comprehensive development toolkit for building, deploying, and debugging full-stack decentralized applications on the Jovay Layer 2 blockchain.

A sophisticated five-layer memory architecture for AI agents that enables episodic recall, semantic compression, and intelligent information forgetting.

A high-speed information capture tool that monitors multiple platforms for low-price goods, real estate, and bidding opportunities with instant alerts.

An AI-powered passenger assistant for Hellobike Hitch that manages price comparison, ride booking, and real-time trip tracking.

A professional-grade skill for managing DeepInfra model routing, provider authentication, and cost-aware fallback chains.

A specialized tool that generates professional PowerPoint presentations from natural language prompts using Anthropics-inspired design principles.








































