Help center
The desk is small, so the help center is short and direct. Start with these how-tos, and if the answer isn't here, go to the FAQ or write us.
How to run an appraisal
Type a name — bare or with a TLD like example.com — into the analyzer on the appraisal page. It normalises the input, strips any extension, and runs the machine-checkable gates: the phone test, length and form, banned tails, slur screen, the anti-slop and flex-word screens, and the TLD policy. You get a 0–100 score, a tier, and the full check list.
How to generate names
Type into the name generatorand it proposes made-up words that still read as real. Each one is scored for how naturally it reads and how close it sits to words people already know. You get a shortlist, not a verdict — the judgement calls are still yours.
How to read the score
The appraisal score is the mechanical half only: 100 minus failures and warnings across the auto-evaluable gates. It is a floor, not a verdict. Checks marked human judgement are deliberately left out of the number — 0h first read, the invoice test, and tri-civilisation landing cannot be scored honestly by code.
Why availability must be re-checked at the registry
Every availability list is a snapshot. Open seams close in days — names are registered while a hunt is still reading its own results. The desk's own verdict is that a candidate is only available once the registry says so at the moment you act, same-day.
What is public, and what is not
The desk, the doctrine and the market pages are public and indexable — they are the product's argument, and an argument nobody can find is not one. The workspace, sign-in and sign-up carry noindex, nofollow, noarchive and are absent from the sitemap, because they render nothing without an account.
How to read a bulk table
Submit several names — comma or space separated — and the analyzer routes to the bulk analysis view, where each name is appraised side by side. Compare scores and failing gates across a whole shortlist rather than one name at a time.