AI finding
Most of the page text is missing from the HTML source
AI Warning Severity: High
Part of the audit The content exists only after JavaScript runs. Finding id: ai.mainTextOnlyRendered.
What it costs you
A crawler that does not run JavaScript reads a page that is mostly empty. It has little to quote and may not use the page at all.
What to do
Render the main content on the server, so the first response holds the text.
What it checks
- Main content missing until scripts run
How to fix it
Render the main content on the server
- Check which pages are built in the browser: the ones this rule lists usually share a template.
- Turn on server-side rendering or prerendering for that template in the framework the site uses.
- Where that is not possible, put at least the main text in the HTML and let the script add the rest.
How to check: curl the URL and search the output for a sentence from the middle of the page. Then run this rule again.
A real case
An SEO consultant tested a site that builds all its content with JavaScript: ChatGPT said it could not read the pages, Perplexity found no content on any address tested, and Claude got pages without visible content. Comparison pages that did not depend on JavaScript were read by all three.
Thresholds
| Name | Default | Meaning |
|---|---|---|
min_source_text_share | 0.7 | Share of the rendered text (0 to 1) that must already be in the HTML source. |
How to verify by hand
Logs showing AI bots fetch the HTML but not the script and data requests.
What this relies on
Most AI crawlers do not run JavaScript, so they see the source version.