/evaluators

AEO · aeo-llms-txt-quality

llms.txt Quality Evaluator

Checks whether a site publishes an llms.txt file that gives AI agents a compact, truthful map of important pages, product facts, and documentation.

What it checks

AI agents increasingly look for machine-readable entry points before crawling deep pages. A clear llms.txt reduces ambiguity and helps models find canonical product, pricing, docs, and methodology pages.

Pass criteria

  • The file is available at /llms.txt with a 200 response.
  • It links to canonical product, docs, pricing, methodology, and support pages.
  • It is concise enough for agents to parse without duplicating the whole site.
  • Claims match the visible public website and pricing copy.

Example

# Example SaaS

> AI search optimization platform for marketing teams.

## Core pages
- Product: https://example.com/
- Pricing: https://example.com/pricing
- Docs: https://example.com/docs
- Methodology: https://example.com/methodology

Common failures

  • - Missing /llms.txt entirely.
  • - Linking only to marketing pages and not methodology or docs.
  • - Publishing stale pricing or exaggerated claims.

Related surfaces

FAQ

Does llms.txt guarantee ChatGPT or Claude citations?
No. It improves discoverability and crawl clarity, but citations still depend on source authority, content quality, crawl access, and the user query.
Should llms.txt contain full article text?
Use /llms.txt as a structured index. If you expose full content, place it in a separate /llms-full.txt file and keep it consistent with canonical pages.