The index

aka.tophhie.cloud/index lists the short links that have been published. It answers with JSON rather than a web page, so it is as useful to a script as to a person.

curl -sS https://aka.tophhie.cloud/index

{
  "links_count": 33,
  "root_url": "https://aka.tophhie.cloud",
  "links": [
    {
      "title": "Azure DevOps",
      "shortname": "devops",
      "redirect_url": "https://dev.azure.com/tophhie",
      "short_url": "https://aka.tophhie.cloud/devops"
    }
  ]
}

Each entry gives you the human-readable title, the shortname itself, the redirect_url it points at, and a ready-made short_url you can copy. Entries come back sorted by title.

Looking one up

Add a shortname query to filter to a single link:

curl -sS "https://aka.tophhie.cloud/index?shortname=devops"

The response has the same shape, with links_count reflecting what matched. A name that does not exist, or is not published, returns a count of zero and an empty list rather than a 404. That is the difference between asking the index a question and following a link: the index tells you nothing was found, where the redirect itself gives you a 404.

The index is not everything

The index lists published links only. A link can be perfectly functional and deliberately absent from it, because being listed is a separate decision from working. See Link states.

So treat the index as a directory of what has been advertised, not an inventory. If somebody sends you a short link that does not appear in the index, that is normal and does not mean the link is suspect.

Using it in a script

The response is stable and cheap to parse, so it is a reasonable thing to build on:

curl -sS https://aka.tophhie.cloud/index \
  | python3 -c "import sys,json;[print(l['shortname'],'->',l['redirect_url']) for l in json.load(sys.stdin)['links']]"

Two things to keep in mind. There is a rate limit, so poll it sparingly rather than on every page load. And the list changes when somebody edits a link, so cache it with an expiry rather than treating a copy as permanent.