The RSS Directory Verifier
The RSS Directory lets publishers manage the listing of their feed. Before we hand a listing to someone, they have to show that the feed is theirs: they put a token we give them on their website, in their feed or in their domain's DNS, and the RSS Directory Verifier checks that it's there. This page explains what the verifier requests, when, and how to reach us if it causes you trouble.
The verifier is not a crawler. It doesn't follow links, doesn't index anything, and only checks addresses that belong to a feed a publisher has claimed in our publisher console.
How it identifies itself
Every request the verifier sends carries this user agent:
RSSDirectory-Verifier/1.0 (+https://rssdirectory.org/verifier.html)
What it requests
The publisher chooses one of three ways to show the feed is theirs. The addresses come from the feed as our directory knows it, never from what the publisher types in:
- A meta tag on the homepage. One
GETrequest for the homepage the feed links to, looking for<meta name="rssdirectory-verify" content="…">in its<head>. - A tag in the feed. One
GETrequest for the feed, looking for<podcast:txt purpose="rssdirectory">on the feed itself (or its equivalent in a JSON feed). - A DNS record. A lookup of the TXT records of the feed's domain, and of the domain above it, through public DNS resolvers. Your web server sees nothing of it.
Only the first megabyte of a homepage and the first five megabytes of a feed are read. The verifier doesn't load images, scripts or stylesheets, doesn't run JavaScript, and doesn't keep the pages it reads: we only store whether the token was found and the HTTP status of the answer.
It follows one redirect, and only within the same site (from http:// to https://, or to the www. address, for example). A redirect to another site, or a second redirect, ends the check.
When it checks
Nothing happens until the publisher tells us the token is in place, or submits their claim. Then the verifier checks:
- right away,
- every minute for five minutes,
- every three minutes until an hour after the start,
- every 45 minutes until 48 hours after the start.
That's at most 87 requests over two days, and it stops as soon as the token is found. The same address is requested at most once a minute, however many publishers claim it, and requests to one host are made one after the other, never in parallel.
When it stops
The verifier keeps checking after a 2xx answer without the token, a 404 Not Found, a 5xx server error, or no answer at all: those are the usual states of a page that is being changed.
Any other answer stops it at once, for example 401, 403 Forbidden, 410 Gone or 429 Too Many Requests
, and so does a redirect to another site. We tell the publisher why, so they can fix it. The publisher can also cancel the checks at any time. When they restart the check, the verifier makes one request right away and then continues on the schedule above, without the checks every minute. A publisher can restart it at most every three minutes, and after the 48 hours each restart is a single request.
The verifier doesn't read robots.txt: it only requests the few addresses a publisher asked us to check, and only as often as described here.
Contact
If the verifier causes you trouble, or you want to know who claimed your feed, write to . Please include the address it requested and roughly when.