How can I improve my score and make my writing more natural?

The skill provides specific suggestions such as varying sentence lengths, replacing buzzwords with active verbs, and breaking uniform paragraph structures to ensure better flow.

What are considered 'high-risk AI words'?

These are terms frequently overused by LLMs, such as 'delve,' 'leverage,' 'robust,' 'comprehensive,' and mechanical transitions like 'furthermore' or 'moreover.'

How does ai-check determine if text is AI-generated?

It analyzes five distinct metrics: grammar perfection, sentence uniformity, paragraph structure, frequency of high-risk AI-typical terms, and punctuation consistency.

Does a high confidence score mean the text is definitely AI?

Not necessarily. It indicates stylistic similarity to LLM output. High scores can occur in writing by non-native speakers using formal templates or authors with a very rigid, formal style.

Can I use this for non-academic writing?

Yes, while it is optimized for research and academic manuscripts, it is highly effective at identifying LLM patterns in technical documentation, formal reports, and blog posts.

AI Research Content Validator

Name: AI Research Content Validator
Author: astoreyai

byastoreyai

•

학습 및 문서화

Analyzes research manuscripts and documentation to detect AI-generated patterns and ensure authentic academic writing.

The AI-Check skill for Claude Code is a specialized quality assurance tool designed for academic and technical writers. It evaluates text against five key linguistic dimensions—grammar perfection, sentence uniformity, paragraph structure, high-risk vocabulary, and punctuation patterns—to provide a comprehensive confidence score on content authenticity. By identifying common 'AI-isms' like the overuse of words such as 'delve' or 'leverage,' this skill helps researchers, students, and documentation teams maintain a natural, human-authored voice in their publications and technical artifacts.

주요 기능

012 GitHub stars

02Detection of high-risk AI vocabulary and business jargon overused by LLMs.

03Linguistic analysis of sentence length distribution and structural uniformity.

04Automated pre-commit validation for documentation and research repos.

05Actionable improvement suggestions to restore natural flow and voice.

06Multi-dimensional confidence scoring system for text authenticity.

사용 사례

01Reviewing team contributions to ensure consistent and authentic writing styles across collaborative projects.

02Validating academic manuscripts and dissertations before submission to journals or advisors.

03Automated quality assurance for technical documentation to prevent 'robotic' sounding guides.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add astoreyai/ai_scientist ai-check

For use in Claude.ai and ChatGPT

Download Skill