humanize-ai-text

What This Skill Does

The humanize-ai-text skill is a sophisticated CLI utility designed to refine AI-generated content by stripping away the characteristic "fingerprints" left by Large Language Models. By mapping text against Wikipedia’s documented patterns of AI writing, this tool identifies and neutralizes stylistic markers that are commonly flagged by academic and professional AI detectors like Turnitin, Originality.ai, and GPTZero. The skill operates by detecting 16 distinct pattern categories, ranging from critical chatbot artifacts (such as citation bugs and markdown headers) to subtle stylistic choices (like the overuse of the 'rule of three' or specific 'AI vocabulary' like 'delve' and 'tapestry'). It provides a robust pipeline for detecting issues, performing automated transformations to naturalize sentence structure, and comparing documents to ensure the final output retains original meaning while shedding synthetic patterns.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal:

clawhub install openclaw/skills/skills/moltbro/humanize-ai-text

Ensure you have Python 3.x installed, as the utility leverages native Python scripts for text processing and analysis.

Use Cases

This tool is ideal for content creators, researchers, and developers who utilize AI to draft initial versions of documents but need to ensure the final publication is indistinguishable from human writing. It is particularly useful for university students refining essays to pass institutional standards, copywriters looking to remove corporate AI 'fluff' from marketing materials, and technical writers aiming for more readable, less robotic documentation.

Example Prompts

1. "OpenClaw, scan my document at path/to/essay.txt and give me a detailed breakdown of the AI probability score and the specific paragraphs that need adjustment."
2. "Transform the AI-generated text in my-draft.txt to sound more natural. Use the aggressive mode to rewrite the heavy markdown and em-dash usage."
3. "Run a comparison between the original-content.txt and clean-output.txt to show me exactly which filler phrases and 'AI vocabulary' words were removed during the humanization process."

Tips & Limitations

While the humanize-ai-text skill is highly effective at modifying stylistic markers, it does not rewrite the underlying logic or factual assertions of your document. For the best results, use the -a (aggressive) flag on highly generic AI drafts, but always perform a manual read-through afterwards to ensure the tone remains consistent with your personal or brand voice. Note that as AI detection technology evolves, this tool is intended as a starting point for human-in-the-loop editing rather than a guarantee of bypassing all future detection heuristics.