RSS/Atom Feed Validator

Enter a site URL or a feed URL

• Site URL (e.g. https://example.com/): if the response is an HTML page, we read up to 5 feed declarations (<link rel="alternate" type="application/rss+xml"> or application/atom+xml) from it. If the page declares none, we try 6 common paths instead: /feed/, /feed, /rss.xml, /atom.xml, /index.xml and /feed.xml. Each candidate is checked by its first few KB, and the first one that starts like RSS or Atom is validated in full.
• Feed URL (e.g. https://example.com/feed/, or https://example.com/?feed=rss2 — query strings are kept): if the response starts with <rss>, <feed> or <rdf:RDF>, it is validated directly and no discovery runs.
• URLs containing a username or password (user:pass@…) are rejected. Do not enter tokens or other secrets either — the fetch is a plain public GET with no cookies or credentials.

Three outcomes to tell apart

• A result card without “Discovered feeds”: the URL you entered is itself an RSS or Atom feed. The badge reads “Syntax errors” when at least one finding is an Error — that covers invalid dates and missing required elements as well as broken XML.
• A result card with “Discovered feeds” above the findings: the URL returned HTML and the feed was found from it. If you expected that URL to be the feed, it is serving a web page instead (for example, a feed address that redirects to the home page).
• An error message and no result card: no feed was validated. “No feed could be found” means the URL returned HTML but no candidate was RSS or Atom. “Neither a feed nor HTML” covers other content such as plain text, and a JSON Feed is named as such. An error status or a failed connection means we could not fetch it: 404 usually points to a wrong URL, while 403 or 429 means the site refused or rate-limited our server — not that the feed is missing. We do not retry or work around it.

What RSS and Atom are checked for, and how to read the findings

• RSS 2.0: <channel> needs a title, link and description (a missing one is reported as missing_required with channel/title and so on). Each <item> should have a title, an absolute <link>, a <guid> and a <pubDate> in RFC 822 form — Mon, 12 Oct 2026 09:00:00 +0900.
• Atom 1.0: the root <feed> needs the Atom namespace plus a title, id and updated (feed/title, feed/id, feed/updated). Each <entry> should have a title, an absolute <link rel="alternate" href>, an <id> and an <updated> or <published> date in RFC 3339 form — 2026-10-12T09:00:00+09:00.
• For both formats we also look at the rel="self" link, newest-first order, and whether the declared encoding matches the bytes.

Example: this RSS feed has an unreadable date, a missing link, a relative link, an Atom-style date and no closing </rss>.

<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
  <title>Example Blog</title>
  <link>https://example.com/</link>
  <description>Latest posts</description>
  <atom:link rel="self" href="https://example.com/feed/"/>
  <item>
    <title>Second post</title>
    <guid>https://example.com/2/</guid>
    <pubDate>15/10/2026</pubDate>
  </item>
  <item>
    <title>First post</title>
    <link>/1/</link>
    <guid>https://example.com/1/</guid>
    <pubDate>2026-10-12T09:00:00+09:00</pubDate>
  </item>
</channel>

The validator reports these five findings, in this order (Errors first, then by line):

• Error date_invalid line 8

  The date cannot be parsed, so readers cannot order the items.

  15/10/2026
• Error unclosed_tag line 13

  An element is never closed. Readers abort parsing here, so items may not appear at all.

  <rss>
• Warning item_missing_link line 8

  The item has no <link>.

• Warning item_relative_link line 13

  The link is relative. Feeds are read on other domains, so links must be absolute.

  /1/
• Warning date_wrong_format line 13

  The date format does not match the spec (RSS 2.0 wants RFC 822 — "Fri, 18 Sep 2026 09:00:00 +0900"; Atom wants RFC 3339 — "2026-09-18T09:00:00+09:00"). Many readers cope, but matching the spec is safer.

  2026-10-12T09:00:00+09:00

• date_invalid (Error): 15/10/2026 cannot be read as a date at all, so readers cannot place the item. The line number points to the item that holds the date.
• unclosed_tag (Error): the element named in the finding — here <rss> — is never closed; add </rss> at the end. For this finding the line number can point past the opening tag, so go by the element name.
• item_missing_link / item_relative_link (Warning): the first item has no <link>, and the second uses /1/. Feeds are read on other domains, so write the full https://example.com/1/. We only look at what the feed says; we do not open the article to see whether it exists.
• date_wrong_format (Warning): the value is a valid date, but in Atom form. RSS 2.0 expects Mon, 12 Oct 2026 09:00:00 +0900.
• Up to five findings of the same kind are listed; further ones are counted under the first. Feeds larger than 512KB are validated up to that point only, and a tag left open by the cut is not reported.

What this validator does not check

• Content added by JavaScript. We read the server's response as it arrives and run no scripts. If a feed link is inserted into the page by JavaScript, discovery will not see it — enter the feed URL directly.
• The article pages. Item links are not opened or compared with the feed, and the item text is not checked against the published article.
• Delivery. We do not check whether feed readers, podcast directories or RSS-to-email services pick the feed up, or whether email sent from it arrives. A clean result does not guarantee acceptance by any service or any change in search rankings.
• Over time. Each run validates one feed once. Nothing is scheduled or stored, so it is not a monitor — run it again after you change the feed.