squirrelscan
GitHub

LLM Parsability

What LLM parsability means: how reliably a language model can extract a page's facts, and how squirrelscan scores it from 0 to 100 per page.

Analyzes how well LLMs can parse and understand the content

What LLM parsability means

LLM parsability is how reliably a language model can read a page and come away with the right facts: what the page is about, what it claims, and what the reader is meant to do next. It is not the same thing as human readability, and it is not a search ranking factor. It matters because AI search answers, chat assistants, and coding agents quote the pages they parsed cleanly and quietly skip the ones where the real content is tangled up with navigation, banners, and layout text. squirrelscan scores each page from 0 to 100 with a cloud model and treats 70 as the pass mark, on the scale below.

Rule IDai/llm-parsability
CategoryAgent Experience
ScopePer-page
Severityinfo
Weight3/10
DefaultEnabled

Solution

This rule evaluates how well LLMs can understand and extract information from your content, which affects AI-powered search and assistants. Improve parsability by using clear structure, explicit topic sentences, and well-organized sections. Avoid ambiguous pronouns and ensure context is clear. Use semantic HTML and structured data. Clear, well-written content for humans typically scores well for LLM parsability too.

Scoring

The rule scores content from 0-100 based on:

  • Clear semantic structure (headings, paragraphs, lists)
  • Logical content flow
  • Absence of layout text/navigation mixed with content
  • Clean text without excessive special characters
  • Well-defined sections and topics
  • Machine-readable formatting
ScoreStatusMeaning
70-100PassGood LLM parsability
40-69WarningCould be improved
0-39FailPoor parsability

Enable / disable

Disable this rule

[rules]
disable = ["ai/llm-parsability"]

Disable all AI rules

[rules]
disable = ["ai/*"]

Type to search…

↑↓ navigate↵ openesc close