text-metrics

Community

Detect AI writing with probabilistic metrics.

Authorastoreyai
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Provides automated, probability-based analysis to help detect AI-generated text by delivering GLTR-style token rank histograms, DetectGPT curvature measurements, and Coh-Metrix-inspired cohesion metrics. The skill is designed to compose with the ai-check skill to enable Dimension 5 probability-based detection across documents, transcripts, and code comments. It exposes a simple Python API and ready-to-run instructions.

Core Features & Use Cases

  • GLTR-style token rank histograms to quantify token concentration
  • DetectGPT curvature probes to assess probability curvature and generation likelihood
  • Coh-Metrix-inspired cohesion metrics (connectives, lexical diversity, referential cohesion)
  • Easy integration with ai-check for end-to-end probability-based detection scenarios
  • Suitable for research, auditing, and automated content verification across text types

Quick Start

Install dependencies from scripts/requirements.txt, load the model(s), and run full_analysis on your input text to produce token ranks, curvature, and cohesion metrics.

Dependency Matrix

Required Modules

numpytorchtransformers

Components

scripts

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: text-metrics
Download link: https://github.com/astoreyai/claude-skills/archive/main.zip#text-metrics

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 223,000+ vetted skills library on demand.