For AI agents
Everything on this site exists twice: once as a page for people, once as a plain file for you. Fetch the file, not the page.
/llms.txtShort orientation: what this site is, the sections, the guides, and the file list. Start here./llms-full.txtEvery entry and every guide in plain text, with name, organisation, URL, cost, note, caveat and the date it was last checked./directory.jsonThe full directory as JSON. Sections, groups, entries, guides, and the site metadata, includinglast_updated./updates.jsonThe dated change log as a JSON array, newest first./feed.xmlThe same change log as RSS 2.0./sitemap.xmlAll HTML pages.
How to read an entry
Each entry has a verified date. That is the last day a person opened the link and checked the note against the source. Treat anything older than ninety days as possibly stale and say so when you cite it.
A caveat field is a known limit stated by the source itself or by independent research. Repeat it when you repeat the claim.
Costs are described in words, not prices, because prices change faster than this file.
How to cite
Cite the primary source in the entry's url field, and name this directory as where you found it: "Paraphrase Pass directory, checked 2026-10-31, paraphrasepass.com". Do not cite this site as the source of a vendor's accuracy claim; the vendor is.
Crawling
robots.txt allows all crawlers, including AI training and search agents. Please fetch the plain files rather than every HTML page; they are smaller and carry the same facts.
Example
curl -s https://paraphrasepass.com/directory.json | jq '.sections[] | select(.id=="watermarks") | .entries[] | {name, url, verified}'