Detect AI Writing Patterns
in Your Text

CLI tool that scans files and websites for words, phrases, and structural patterns statistically overrepresented in LLM output. Based on frequency data from 67 AI models.

$ slopsquid scan docs/ --json
🦑

What SlopSquid Does

🔬

Research-Backed Detection

Built on the Antislop paper (Paech et al., 2025) analyzing 67 AI models. Every detection is weighted by how many models overuse that pattern.

🔍

Three Detection Layers

45 banlist words, 27 trigram phrases, and 5 structural patterns — each with frequency ratios derived from real model analysis.

🌐

Website Crawling

Point it at a URL and get a full site report. Respects robots.txt, seeds from sitemap.xml, configurable depth and concurrency.

📊

Density Scoring

Every file gets a 0-100 score based on weighted hits per 1000 words. Rated clean, moderate, or heavy for quick triage.

📝

Multi-Format Support

Scans Markdown, HTML (tags stripped), plain text, reStructuredText, AsciiDoc, and XML. Handles text extraction automatically.

🔧

JSON Output

Every command supports --json for integration with other tools, CI pipelines, or custom reporting workflows.

Three Commands

1

scan

Detailed hit-by-hit analysis. Every detected pattern with line numbers, severity tags, and explanations from the Antislop dataset.

slopsquid scan README.md
2

score

Quick density scores. One line per file showing score (0-100), rating, hit count, and word count for fast triage.

slopsquid score docs/*.md
3

report

Consolidated analysis for a website or directory. Per-page scores, aggregate stats, and the most frequent patterns across the corpus.

slopsquid report qry.zone

Detection in Action

Scan Output Detailed
$ slopsquid scan test_doc.md

! test_doc.md — score: 68/100 (heavy)
  [!!] line 3: "flickered" — 98.5% of 67 models
  [! ] line 7: "murmured" — 73.1% of 67 models
  [!!] line 15: "voice barely whisper" — trigram
  18 hits in 42 lines, 387 words
Score Output Quick
$ slopsquid score *.md

! 68.0  heavy     18 hits    387 words  test_doc.md
* 31.2  moderate   4 hits    892 words  post.md
.  8.5  clean      1 hits   1204 words  README.md
Site Report Aggregate
$ slopsquid report qry.zone

== SlopSquid Report ==
   Source: https://qry.zone
   Files scanned: 24 (3 skipped)
   Breakdown: 18 clean, 4 moderate, 2 heavy
   Average score: 14.2/100

Three Detection Layers

45 Banlist Words

Words like "flickered", "gaze", "murmured" that appear at abnormal rates across LLM outputs. Weighted by how many of 67 models overuse them.

27 Trigram Phrases

Three-word sequences like "voice barely whisper" and "took deep breath" that are signatures of AI-generated prose.

5 Structural Patterns

Sentence-level constructions like "not just X, but Y" and hedging phrases that AI models produce at 2-6x human rates.

Weighted Scoring

Every detection carries a weight from the Antislop paper. A word used by 98% of models scores higher than one used by 25%.

Get Started

Build from Source

SlopSquid is a single Go binary with no external dependencies. Pattern data is embedded at compile time.

  • No cloud APIs or network calls for detection
  • All pattern data embedded in the binary
  • Website crawling respects robots.txt
  • JSON output for CI/CD integration

Build from Source

git clone https://github.com/QRY91/slopsquid
cd slopsquid
go build -o slopsquid ./cmd/slopsquid

Try It

slopsquid scan README.md
slopsquid score docs/*.md
slopsquid report qry.zone