Understanding AI Discovery File Validation
Learn what AI Discovery File validation checks, what it does not prove, and what to do when validation needs review.
AI Discovery File validation checks whether the generated file is usable, safe, and consistent with available site-level evidence.
Validation is a quality check. It is not a prediction of AI visibility.
What Validation Checks
Validation may check whether:
- the file was generated;
- the file is not empty;
- included URLs appear to be public and useful;
- obvious login, account, admin, cart, checkout, search, or private URLs are not included;
- duplicate URLs were avoided;
- canonical URLs were used where available;
- a newer eligible run can create its own AI Discovery File;
- the installed file is reachable when installation checking is available.
- the output has exactly one H1 and a blockquote summary;
- eligible URLs appear in Markdown link syntax;
- file size and link count remain within practical review thresholds.
These checks improve machine readability and maintainability. They are not a ranking score and do not predict whether a particular AI system will use the file.
What Validation Does Not Prove
Validation does not prove that:
- AI systems will read the file;
- AI systems will cite your website;
- your site will rank higher;
- your site will appear in AI-generated answers;
- every important page is guaranteed to be included.
Common Validation Messages
Ready means the file exists and is available to preview, copy, or download.
Validated means required checks passed.
Needs review means one or more validation checks found something that should be reviewed before installation.
New AI Discovery File available means a newer eligible Website Rescan or Flexible Website Rescan can create its own file. Your earlier file remains unchanged.
No eligible URLs means MarketingStride did not find public URLs that passed the file's inclusion rules.
What To Do If Validation Needs Review
Start with the validation message.
If a specific URL is mentioned, check whether that URL should be public and useful. If it is a login, account, admin, cart, checkout, search, tag archive, author archive, privacy, terms, cookie, private, gated, or duplicate page, it may need to be excluded.
If a newer eligible site-level run is complete, generate that run's new AI Discovery File. This creates a separate file for the newer run and does not update or replace your earlier file.
If no eligible URLs were found, check whether the site could be crawled and whether public pages were discoverable through sitemap, navigation, canonical URL, or analyzed-page evidence.
When To Contact Support
Contact MarketingStride Support if:
- validation fails and the reason is unclear;
- the preview and download do not match;
- private or checkout-related URLs appear in the file;
- a public page you expected is missing and you cannot tell why;
- the installed `/llms.txt` file cannot be reached after installation.