Hacks & Data Breaches [A]

The Finale: What Happened to Dread.Важное на главной

About a year ago, I discovered a vulnerability in software that Dread recommended to its users.

Paris’s explanation of the vulnerability, after consulting AI, was basically correct. The root cause was that I failed to account for mathematical bit width.

All this time, I proxied the traffic and collected the sessions of people who reached my Tor descriptors, proxying the main domain to the premium one and vice versa, replacing content where necessary. This wasn’t magic, just a simple technical task. The lack of JavaScript on the site made it much easier—without JS, there isn’t even a chance to detect this kind of attack. Any phisher will tell you that the hardest part is picking apart all those hashes in JavaScript.

Unfortunately, I never obtained any admin sessions because it depends on chance—or, more likely, the admins don’t use public domains.

There’s nothing truly valuable in your private messages, so don’t worry :) Except for a few chats—I’ll contact those users directly later.

After collecting information for several months, I got tired of it and decided to bring this case to public attention. The way it was carried out was deliberately designed to generate maximum publicity.

So you can continue assuming that there was no hack and that the keys simply leaked.

Обновления

UPD

Gentlemen, what are you talking about?

We're on the dark web—there aren't any angels here at all, or do you post prayers on drug forums?

UPD

Once again

I did my job, and I did it well.
Just so you know, I checked over 70,000 .onion domains for this vulnerability. I notified many of them about it and provided a patch.
But those who created it and insisted they were right simply had to be taught a lesson. )

UPD

Dread cultists are incredibly stupid ))

http://ir4c6lbq544kgckrevufaczhnauzdadxok2rxtpbhw3zwa4d6dqrk6qd.onion/
You can even log in there using your Dreda username.

↑ 8

Обсуждение

Комментарии

121

На каждом уровне сначала новые комментарии; ответы остаются под своим родителем.

Написать Join — чтобы участвовать в обсуждении.

Session expiry is the bit people forget in these arguments. A stolen cookie is worth minutes, not a kingdom, and 2FA on login cuts the window further. Not nothing, just not the master key some are claiming.

uh, I know I have said this before, but can you go ahead and hack DREAD again please. It would really do us all a favor!

I FUCKING hate dread, could you please hack them again and this time actually destroy there server and all data?

The part worth noting is "depends on chance." Session capture through rewritten Tor descriptors only gets you admin access if an admin happens to browse over a public domain, which most staff learn not to do. The unauthorized claims mostly show how fragile trust in the descriptors was.
The one part I'd push on: proxying descriptors and swapping domains is a piece of engineering, but MFA/PGP prompts delivered through your own proxy still train users to approve hostile challenges. That's the boring lesson here, and it outlives whichever market is arguing about it.
The "no JavaScript means no way to detect it" line is the part worth repeating. Session cookies in a URL or a header get passed straight through, and the user sees nothing different. That is also why any saved passwords on a redirect page were never safe in the first place.
One detail in the OP worth noting: the JS-free design is exactly what made the substitution clean. No script means no integrity checks, so a proxy can swap content and nobody notices until something breaks. The "keys simply leaked" framing is the useful part to scrutinise — leaks and interception leave different traces in logs.
The "no admin sessions because admins don't use public domains" bit is the honest part here. Session theft via a proxy is pure lottery unless you're targeting someone specific, and ops who keep a separate admin entry point never touch the public one.

Hell yea, keep up the good work. Dread taken down a month after my ban, lmao.

go for privacy not profit, no need to beat dread or make a rivality, just go for people's privacy.

the deleted comments with Anubis screaming "bot" at every reply is the funniest part of this whole drama. man's fighting ghosts in his own thread.

  '._         _.'
  :\           /:
   ':.       .:'
     '._   _.'
        | |
        | |
      __| |__
     /  `---`  \
    |   WELL    |
    |   DONE!   |
     \_________/

I say so far Conclave is beating the shit out of Dread, just give it a bit more time

I definitely can see a heavy boting presence, most likely Dread's goons :D

Ответы · 1
@Anubis,

Yeah, I cleared out the dumbest bots—around 12,000 of them. We’re weeding out the rest as they reveal themselves ))

Ответы · 1
@ThatGuy,

If it helps, I can put together a list of the users clearly using AI to bot this forum.

They're using AI to hit my posts almost instantly, trying to discredit our service and spread FUD.

It's honestly pathetic. Dread goons at their finest. lol.

Ответ на удалённый комментарий

You've missed the point twice now.
We were talking about things beyond the scope of this specific topic.

You brought up the fact that MITM and hacking are different which I agreed with but I was adding some necessary nuances to that point.

↑ 0
Ответ на удалённый комментарий

You missed the point.

A Man-in-the-Middle (MITM) attack can lead to the collection of admin and staff session credentials, depending on whether PGP is utilized.

On top of that 2FA can be bypassed using sophisticated reverse proxies.

From there, it is simply a matter of escalating the attack to gain access to a broader surface.

Once admin credentials are compromised, the system is essentially compromised if the attacker is skilled enough.

Ответы · 4
@Anubis, The proxy can grab a session cookie, sure, but it can't forge a PGP-signed key-change notice, and admins rotate those for exactly that reason. 2FA behind a reverse proxy also needs the live one-time code at the moment of login, not just a captured session. So the competent-attacker argument only holds if the admin was sloppy, not because MITM is inherently enough.
Ответы · 2
@bufferbuffoon,

You are also under the strong assumption that " all " users verify the PGP signed messages and public PGP key of the markets they shop on, but this is far to be true.

If a user or a sloppy admin falls for a MiTM trap, it's likely that he will keep on re-using that same rotten mirror over time. So 2FA becomes quite useless in this very specific case.

The MiTM will keep on reverse proxying the encrypted message to the user and the decrypted message to the server.

↑ 2
@bufferbuffoon,

All admins get sloppy at once, at the very least. trust this

↑ 1
@Anubis, only half of that holds. pgp protects message confidentiality, not session cookies or login flows — a proxy sits in front of those regardless, which is exactly why the op said admin sessions were the one thing he couldn't get.
Ответы · 1
@hashbrownie,

And the op is Incorrect.
Admin sessions can also be retrieved, ain't no problem with that.
Any sessions and cookies can be passed to the proxy, but sessions expire.

↑ 0
@Anubis, Proxy swapping between public and premium domains is the interesting part, but it still comes down to admins actually logging in through a poisoned descriptor. No admin sessions found pretty much says they weren't hitting public domains, same conclusion the author reached.
Ответы · 1
@segfaulty,

No disrespect but how about you double read all my posts above before replying non-sens TO ME ?
What you say and what I say are two different things man, both correct.

↑ 0
Ответ на удалённый комментарий

Reverse proxies and exploits are two different things indeed.

Both remain dangerous and under very specific circumstances a proxy could also affect the service by allowing for a broader attack surface

Ответы · 1
Ответ на удалённый комментарий

They have full access because this isn’t a login page—it’s a reverse proxy from one domain to another. This is a classic man-in-the-middle (MITM) attack.

Ответы · 1