Checks > Addresses
The addresses this site publishes. All of them are entry points to one service, so the account, balance and open orders are identical behind each.
nexusb2l7fmqnefwphyy7m5zjhlkytlbo7qbb5lu5dlczr3azgii2gyd.onionnexusma2iqgauqqvjcgds4ckv5xbf272tkfagq4epojjhsgleqpwxiqd.onionnexusabcd6tyfhdwilyitaqiri6tisj2v2hueyjuj6qkvd6azvi5tuqd.onionRun C-07 before you type anything. Open in Tor Browser only, then compare the onion printed on the login screen against your browser address bar. A mismatch means the page is a copy and the tab should be closed.
Four items on this site govern how an address is obtained and used. They are listed here rather than explained again, because each has its own page with pass criteria and a stated cost of failing.
| Check | Passes when | Weight |
|---|---|---|
| C-06 | The address was copied from a saved source rather than typed or recalled | high |
| C-09 | The source was held before this session began, not found during it | high |
| C-07 | The onion on the login screen matches the browser bar, compared in full | critical |
| C-18 | The saved set was compared against a current source within the last month | standard |
The distinction matters more than it sounds. An address saved directly becomes wrong the day it is retired, and a retired address is indistinguishable from a hostile one when you open it months later. A saved page that publishes the current set survives every rotation, which is why C-18 asks about the source rather than the strings.
This is also why a bookmark beats memory. Fifty six characters is beyond reliable recall, and partial recall is the worst state to be in, because a nearly correct address is precisely what a lookalike is registered to catch.
Nothing on its own. Getting an address from a poor source is only dangerous in combination with skipping the login screen comparison, which is why C-07 carries critical weight and the others do not. The address items reduce how often you need that comparison to save you. The comparison is what saves you when they have already failed.
That layering is deliberate. A single control that must never fail is a fragile design. Four cheap controls where the last one is decisive is a better one, and it is why this set is graded rather than presented as equally important.