Sites in this space vary enormously, and the difference isn't cosmetic. Some are maintained catalogues; many are scraped copies that appeared after somebody else's takedown. You can usually tell them apart in about half a minute.

Check one: does the gate load first

Verified entry should happen before any explicit content renders. Not a checkbox layered over a page that has already loaded — an actual gate that comes first.

This is the cheapest signal available. A site that skips it has either not thought about compliance or has decided not to, and both answers tell you what else has been skipped.

Age verification is a platform function, run alongside accounts and moderation. It isn't something a front end can convincingly improvise, which is why we keep creation and account actions on the source platform through the production booth connection.

Check two: is the artist credited

Look at any entry. Is there a name attached, visibly, near the top? Credit is the single clearest signal of whether a site is a catalogue or a pile.

It's also the economic link between viewers and creators, since most work in this field is funded by patronage and commissions rather than by the sites showing it — the picture in our piece on who actually gets paid.

Stripped credit is characteristic of scrapes. The metadata was thrown away during copying and nobody restored it, because restoring it would take work.

There's an easy follow-up test. Click the name. On a real catalogue it leads to that artist's other entries; on a scrape it either does nothing or leads to a search that returns unrelated results, because the field was never actually populated.

Checks three and four: metadata and old links

Beside each entry, is there real metadata — title, artist, year, format, language, tags? Those six fields are what make a library usable, as we set out in the piece on metadata.

Then try an old link or an old identifier. A maintained catalogue redirects; a scrape returns nothing or, worse, something unrelated because numbers were reused. Silent failure is the giveaway: a reference that resolves to the wrong thing is far more common than one that simply breaks.

Checks five and six: advertising and payment

Redirect advertising, popups, and fake play buttons are the clearest marker of an unmaintained mirror. A site earning through those has no relationship with creators and no reason to maintain anything.

If money is involved, look at how. A named, conventional payment processor means somebody accepted terms and can be held to them. An unclear or off-platform payment path means the opposite, and payment routes are the strictest layer in this field, as covered in the piece on how limits get set.

Two smaller signals fit here too. Does the site have a working contact route and a stated takedown process? Both cost almost nothing to publish, and their absence is a reliable sign that nobody expects to answer for what is hosted.

What none of these checks tell you

They don't tell you whether the material is any good, and they don't replace the safety floor. Characters must be fictional and unmistakably adult, with no ambiguity about age, no real-person sexual depiction, and no coercion framed as romance.

A well-built site can still host something that fails those tests, which is why moderation matters more than presentation. Our standards are written out in the control room, and public tools with real account and safety controls sit at models.

Running the thirty-second check

Gate, credit, metadata, old links, ads, payment. Six things, none requiring trust in anyone's claims, all observable from the outside.

The habit is worth building because the alternative is judging by branding, and branding is the least reliable signal in a field where domains change constantly and imitators reuse names. Behaviour is checkable; a name is not.