How the feed finder works
Paste a website's address, or type the name of a site people already follow in Feeder. The finder looks in Discover first, where Feeder keeps the feeds its readers follow, with how many people follow each one and whether it still works. If Discover doesn't know the site, Feeder visits it and looks for its feeds the way a feed reader does.
For every feed you get its address to copy into any reader, whether it works, and a button to follow it in Feeder. To see every error in a feed, open it in the RSS feed validator.
What if a site has no RSS feed?
Many sites don't link to a feed, or have none at all. Feeder can still follow them: paste the address into Feeder and it builds a feed from the page, so new posts show up like any other feed.
For AI assistants and developers
The finder answers in JSON too. Ask https://feeder.co/rss-feed-finder.json?url=theverge.com and you get each feed's address, whether it works, its Discover page and a link to follow it in Feeder.
AI assistants can also use Feeder's public MCP server at https://feeder.co/mcp/public. It has two tools, find_feed and check_feed, and needs no account. Lookups have a rate limit.