Expired domains with Wikipedia, university and gov links

A small number of links from sites with real editorial standards is often worth more than hundreds of anonymous ones. Revised names those sites on every listing, in linked_by, and lets you filter the directory on them with source. This guide finds expired domains still linked from Wikipedia, from universities and from government bodies, and shows how to read what comes back.

It assumes you have an API key and have run the Quickstart.

How named linkers work

Every listing carries linked_by: the recognisable sites among its linkers, by display name.

"linked_by": ["Cornell University", "United Nations"]

The names come from a catalogue of 297 sources as of 23 September 2026 — encyclopedias, developer platforms, communities, news outlets, 45 universities and 20 government and intergovernmental bodies. A site that is not in the catalogue can still link to a domain; it counts toward referring_domains but is not named.

Filter on one source

source takes a display name exactly as it appears in linked_by. Case is preserved, so send Wikipedia, not wikipedia.

curl -s -G https://getrevised.com/api/v1/domains \
  -H "Authorization: Bearer $REVISED_API_KEY" \
  --data-urlencode "source=Wikipedia" \
  -d tier=open \
  -d spam=low \
  -d limit=50

--data-urlencode matters as soon as a name has a space in it: Hacker News, Cornell University, United Nations.

Filter on several

Comma-separate names to match any of them. This is how you ask for “linked from a university” or “linked from a government body”: list the institutions you care about.

import os

import requests

session = requests.Session()
session.headers["Authorization"] = f"Bearer {os.environ['REVISED_API_KEY']}"

UNIVERSITIES = ["Cornell University", "Harvard University", "Stanford University", "Yale University"]
GOVERNMENT = ["United Nations", "World Health Organization", "FCC"]

page = session.get(
    "https://getrevised.com/api/v1/domains",
    params={
        "source": ",".join(UNIVERSITIES + GOVERNMENT),
        "tier": "open",
        "spam": "low",
        "limit": 250,
    },
    timeout=30,
).json()

for row in page["data"]:
    print(row["domain"], row["referring_domains"], row["linked_by"])

requests percent-encodes the spaces for you. There is no API endpoint that lists the catalogue, and no single value that means “every university”. To see which names occur, read linked_by off a page of results, or browse the website’s .edu and .gov hubs, which group the same sources.

A name the API cannot read is refused with 400 invalid_request, not silently dropped — see Comma-separated lists.

What comes back

Open-tier listings with named academic, government or Wikipedia links, as of 23 September 2026:

Three things to notice:

• Small referring-domain counts are normal here. horriblevacuum.com has 10 to 50 referring domains, and two of them are Cornell and the UN. That is the pattern this filter exists to find.
• snapshots says whether there was a real site. 50+ archived copies is strong evidence of a site that existed and was maintained; <10 is thin, so read the archive before trusting the links.
• null age is unknown, not new. Do not add ageMin to this query unless you mean to drop every row whose age Revised could not establish.

Check the links before you act

A named link recorded in a crawl is evidence, not a guarantee.

1. Confirm the link still exists

   Find the page on the linking site and check it still links to the name. Pages on large sites are moved, rewritten and archived; a university link from 2014 may be on a page that no longer exists.

2. Check what was linked to

   A Wikipedia citation to a defunct company’s press page is a different asset from one to a reference resource. The archive shows what the page was.

3. Keep the topic

   Links from a health agency to a health resource carry their value only if what you build is related. See Have a plan for the content.

Background reading in the knowledge base: Wikipedia backlinks and .edu and .gov backlinks.

Over MCP

The MCP search_domains tool has no source argument. Ask your agent to search structurally, then read linked_by with get_domain on the survivors:

Search open-tier listings with a low spam band, sorted by referring domains, then get the full record for the top ten and tell me which have a university or government site in linked_by.

Next