Meerkat Markdown for AI Visibility collects URLs using WordPress queries and generates an llms.txt file (short for “Large Language Models”) in the website’s root directory — optionally providing dynamic .md files — to help AI-based systems better understand and make use of the site’s content. The llms.txt is stored in the root directory. You can configure the number of URLs, set a minimum word count, and exclude specific categories or tags. File generation can be triggered manually or scheduled to run automatically.

Features

Gathers URLs with WordPress queries; the wp-sitemap.xml URL is only added to llms.txt as metadata.
Creates llms.txt with [Title • host/path](URL): content entries.
Add a custom site description to the file header.
Limit export to 100 entries or process all URLs.
Posts or pages with fewer than 15 words are skipped: they do not count toward
the 100-entry limit and are omitted from llms.txt.
Choose 30, 50, 100 words or the full text per entry.
Exclude categories and tags.
Sort llms.txt entries by popularity or most recent first.
Optionally serve dynamic .md files for published posts, pages and public custom post types. Responses include canonical, noindex and MIME-sniffing protection headers.
Generates a valid XML sitemap for Markdown URLs without declaring llms.txt as a sitemap or blocking it in robots.txt.
Manual generation via the Generate Files Now button.
Timezone-aware WP-Cron schedule (hourly, daily, weekly, monthly or disabled) with configurable local run time and daylight-saving-safe rescheduling.
Save settings with the Save Changes button.

License

This plugin is open source software licensed under the GPLv2 (or later). See LICENSE.txt for license information.