Structural Deslopper by Sitefire

Structural Deslopper: does your post read like AI wrote it?

Sitefire's SlopShape model detects AI content from structural features. Check your post, give the feedback prompt to your LLM, and check the rewrite here again.

0 words · version 1

Posts from 300 to 10,000 words. We read the whole post. 0 of 10,000 words

You get a score and a feedback prompt for your LLM.

Check a post to see its structural AI-likeness and get a feedback prompt for your LLM.

  • Reads human0–29%
  • Borderline30–69%
  • Reads AI70–100%

How it works

  1. 1

    Check your post

    Paste the text or enter the URL. You get the structural AI-likeness of the post in percent.

  2. 2

    Copy the prompt

    Copy the feedback prompt and give it to your LLM. It will rewrite your content to make it look more human.

  3. 3

    Check again

    Paste the rewrite into the box and check it again. Do this until the post reads human.

Where the model comes from

The model builds on the SlopShape study of human-written and AI-written blog posts.

A low score is no proof that a person wrote the post.

Use it in your agent

The deslop-structure skill runs the same check from Claude Code, Codex or Cursor and rewrites the post until it reads human.

npx skills add sitefire-ai/slop-skills

View on skills.shSource on GitHub

Structural Deslopper: does your post read like AI wrote it? - Sitefire