Draft 0.1

Syndication Manifest

Predictable discovery for public feeds.

Syndication discovery should be simple. Ask once.

A feed reader should be able to make one request, get the publisher's syndication feed list, and stop. No full-site crawl, no HTML scraping, no guessing common feed URLs.

Clients request /.well-known/syndication before parsing pages or guessing URLs. The document lists public web feeds, publisher-declared metadata, optional collections, and external feed claims.

External listings are claims by the publication, not proof of control over another origin. Think YouTube.

Small core, large collections.

A document needs only enough information for a reader to know who published it, which feeds or collections are offered, and what media type each feed uses. Large publishers can advertise collections first and defer detailed feed lists until a reader needs them.

Simple site

{
  "version": "https://syndicationmanifest.org/spec/0.1",
  "publication": {
    "name": "Example Publication",
    "uri": "https://example.com/"
  },
  "feeds": [
    {
      "uri": "https://example.com/feed.xml",
      "mime_type": "application/rss+xml",
      "rel": ["self", "canonical"],
      "title": "All posts"
    }
  ]
}

Site with collections

{
  "version": "https://syndicationmanifest.org/spec/0.1",
  "publication": {
    "name": "Example Media",
    "uri": "https://example.com/",
    "language": "en"
  },
  "collections": [
    {
      "id": "news",
      "title": "News",
      "uri": "https://example.com/news/",
      "description": "Public feeds from the news desk.",
      "feed_count": 24,
      "feeds_uri": "https://example.com/.well-known/syndication/news"
    },
    {
      "id": "sport",
      "title": "Sport",
      "uri": "https://example.com/sport/",
      "description": "Public feeds from the sport desk.",
      "feed_count": 12,
      "feeds_uri": "https://example.com/.well-known/syndication/sport"
    }
  ]
}