well-known

From IndieWeb

This article is a stub. You can help the IndieWeb wiki by expanding it with relevant information.


Well-known refers to designating a common URL on domains for data to be located typically automatically by some software (like a browser or a search engine), and is generally an antipattern to be avoided, because it breaks the portability of content across directories and systems; see follow your nose instead.

"Well-known" are prescribed URLs/paths, whereas there are also emergent common page URLs published by IndieWeb sites like about, contact, etc. For those, see:

There are a couple of common patterns or categories of "well-known" URLs:

  1. a root level file
  2. stuff inside a root level "/.well-known/" directory, standardized by https://www.rfc-editor.org/rfc/rfc8615

Criticism

See also the .well-known: Criticism section.

Root Level

emergent pages

Main article: page#Examples

There are a small handful of emergent root level pages that some community members have published:

See those individual pages for specific IndieWeb Examples etc.

There is also a search engine that is indexing specifically a few of these pages:

See pages for more.

historical for browsers and bots

Historically there are several root level paths that are used by browser and search engines, and it's unlikely that we’ll ever be rid of them:

something.txt

There has been a disturbing pattern of new proposals in form of something.txt over the years, and in short, they're pretty much all bad ideas. (if this list gets too big, create a separate page)

ai.txt

ai.txt

ads.txt

Reference: Wikipedia:Ads.txt

humans.txt

humans.txt

human.json

human.json is a variant of this concept, but in JSON format, with additional protocols to assert human authorship and vouch for the humanity of others.

  • 2026-03-13: some discussion in #indieweb-dev about an HTML version using rel values like XFN. rel=met is probably the strongest vouching for someone. It is unclear if anything less has much value beyond individuals’ own statements, e.g. an AI policy.

llms.txt

llmstxt.org is a proposal for a side file called /llms.txt "to provide information to help LLMs use a website at inference time" but also has secondary recommendations, such as suggesting authors provide a markdown version of all pages on a site, seems to constitute the sidefile-antipattern, and despite having "txt" in the name, is supposed to be written in Markdown.

  • There may be utility in this, but it seems all over the place in its goals -- Joe Crawford

An experiment showed no AI traffic to llms.txt:

Update 2026-09-10 llmstxt.org now proposes a "v2" (version 2) which includes link rel relationships to Markdown equivalents for all pages. More details at llmstxt.org

indieweb.txt

indieweb.txt A file proposed to show off the IndieWeb tactics employed by sites that has seen little use since its inception.

trust.txt

API endpoints

Apparently these root files are a convention on some sites:

Per:

inside a root well-known directory

Main article: .well-known

See Also

Retrieved from "https://indieweb.org/wiki/index.php?title=well-known&oldid=113950"