detect-ai

Detect AI-written text via HumanizerAI and report overall score.

1|Updated Jan 24, 2026

Search Tags:#text analysis #editorial review #content moderation #ai detection #authorship verification #HumanizerAI #submission screening

Ready to run

npx skills add https://github.com/tawf-labs/zkt-hackathon --skill detect-ai-tawf-labs

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: detect-ai
Source: https://github.com/tawf-labs/zkt-hackathon/tree/main/.agents/skills/detect-ai
Command: npx skills add https://github.com/tawf-labs/zkt-hackathon --skill detect-ai-tawf-labs

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires WebFetch.

What problem does it solve?

It helps you assess whether a piece of text is likely AI-generated so you can make better decisions before publishing, submitting, or sharing content.

Core Features & Use Cases

AI-likeness scoring: Returns an overall score (0-100) plus supporting metrics like perplexity and readability.
Verdict classification: Provides a verdict (e.g., "ai") alongside the numerical score for quick interpretation.
Workflow fit: Useful for editors, instructors, and reviewers who need lightweight pre-publication screening for authored content.

Quick Start

Invoke detect-ai with the text you want to check and read the overall score to decide whether the content likely needs human review.

Frequently Asked Questions about detect-ai

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I check if text was written by AI?▼

To check if text was written by AI, submit the content for an AI-likeness score. The detection service evaluates the text and returns a quantitative score from 0 to 100, plus a verdict for quick interpretation.

What metrics are used for AI content detection?▼

AI content detection uses an overall score alongside supporting metrics like perplexity and readability. These metrics help editors and reviewers assess the likelihood that a piece of text is AI-generated.

Can I use AI detection for pre-publication editorial review?▼

Yes, you can use AI detection for pre-publication editorial review. It serves as a lightweight screening tool for editors, instructors, and reviewers to decide whether authored content requires human review before publishing or submitting.

How do I get an AI-likeness score for submission screening?▼

To get an AI-likeness score for submission screening, invoke the detection process with the target text. The system sends the text to the detection endpoint and reports the overall score, verdict, and associated metrics.

Does AI text detection work for general prose and messages?▼

AI text detection works for general prose and messages, applying directly to content review scenarios such as pre-publication checks, academic auditing, and editorial QA to identify AI-generated text accurately.

What is the best way to interpret an AI-likeness score?▼

The best way to interpret an AI-likeness score is to review the overall 0-100 score alongside the provided verdict classification, such as "ai", to quickly determine whether the content likely needs human review.