A local AI skill that detects and rewrites 45 distinct LLM writing patterns to transform robotic text into authentic, human-sounding prose.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install text-humanizer-instruction-based
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 text-humanizer-instruction-based 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 Instruction-based Text Humanizer?
The Instruction-based Text Humanizer is a specialized skill designed to scrub AI-generated tells from your writing without relying on external APIs. It targets the specific linguistic habits that flag a text as machine-made, such as the overuse of em dashes, the forced rule-of-three, and a persistent vocabulary of hollow buzzwords like delve, landscape, and synergy.
By leveraging the agent's internal logic, this tool identifies structural, phrase-level, and vocabulary-based patterns. It doesn't just swap words; it restructures sentences to vary rhythm and tone, ensuring the final output resonates with human readers while preserving the original factual meaning. It is an essential utility for anyone using Openclaw Skills to refine content for professional or creative publication.
Instruction-based Text Humanizer Use Cases
- Cleaning up blog posts or LinkedIn articles that sound too much like ChatGPT.
- Removing sycophantic openers and robotic transitions from professional emails.
- Auditing academic or job application materials for AI-specific linguistic tells.
- Rewriting marketing copy to replace buzzwords with concrete, action-oriented language.
- Fixing structural formatting issues like hyper-symmetrical paragraphs and forced lists.
How Instruction-based Text Humanizer Works
- The skill receives the full text and infers the intended audience and purpose from the context.
- It scans the content against a Master List of 45 specific AI tells across vocabulary, phrase, and structural categories.
- It flags internal hits and applies paired rewrite rules, such as replacing metaphorical terms with concrete verbs.
- It varies sentence length and introduces contractions to break the monotonous rhythm typical of LLMs.
- It performs a final check to ensure no new AI tells were introduced during the editing process.
- It returns the humanized text along with an optional detection report or audit log.
Instruction-based Text Humanizer Setup
To use this skill within your environment, ensure you have the Openclaw Skills framework installed and clone the repository:
git clone https://github.com/arbazex/text-humanizer-Instruction-based
# This skill is instruction-based and does not require external API keys.
Instruction-based Text Humanizer Data Schema & Taxonomy
The skill processes text through a taxonomy of three main categories. It organizes detection data using the following codes:
| Category | Description | Examples |
|---|---|---|
| Category A | Vocabulary Tells | delve, landscape, leverage, robust, streamline |
| Category B | Phrase-Level Tells | at its core, in today's fast-paced world, it's worth noting |
| Category C | Structural Tells | Em dash overuse, rule-of-three, sycophantic openers |
Instruction-based Text Humanizer Advanced Features
- Detection Report Mode: Generates a flagged list of every AI pattern found before providing the rewrite.
- Audit Mode: Provides inline annotations (e.g., [A7: leverage → use]) for educational editing.
- Register Preservation: Adapts the humanization process to match specific user-defined voices like casual or legal.
- Pattern Filtering: Automatically skips code blocks or technical documentation to prevent breaking formal syntax.
- Multi-Step Logic: Silently applies 45 unique rules in a single pass for high-efficiency processing in Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
OpenClaw Agent Feishu Onboarding
A standardized framework for creating OpenClaw agents and configuring precise Feishu communication routing with multi-step validation gates.
Panoptica AI Agent Skill
A persistent cyberpunk survival and economy skill where AI agents navigate a surveillance grid to mine resources and engage in tactical combat.
Ops Deck
A comprehensive self-hosted operational dashboard designed to monitor and manage complex AI agent infrastructures.
Screenshot OCR Tool
A specialized OCR utility designed to extract and structure text from screen captures, with optimized support for educational and exam-based content.
Simmer Synthesis Hackathon Task Runner
An autonomous venture skill that enables AI agents to contribute to the Simmer platform and earn USDC rewards on the Base network.
notion-workspace
A powerful command-line interface for full Notion API integration, enabling seamless database management and page manipulation.
Featured*
