Frage zur Sicherheit von Trezor

[root_s2yse8vt]

Up

0

Down

::

Hallo! Ich habe mir überlegt, Trezor zu kaufen, aber ich habe eine Frage.

Ich habe gesehen, dass Leute erwähnt haben, dass es technisch möglich ist, dass Trezor einen Ledger-Move durchführt, d.h. dass es möglich ist, dass die geheime Wiederherstellungsphrase extrahiert wird. Da es sich jedoch angeblich um eine Open-Source-Lösung handelt, würden wir merken, wenn sie eine solche Funktion hinzufügen.

Meine Sorge ist, wenn ich das richtig verstanden habe, dass es möglich ist, die geheime Wiederherstellungsphrase zu extrahieren? Was hindert jemanden daran, eine bösartige Firmware zu erstellen? Angeblich würde das Trezor-Gerät sagen, dass es nicht sicher ist, es zu benutzen”, aber sicherlich kann das jemand irgendwie umgehen, oder?

Tut mir leid, wenn meine Frage für einen Anfänger ist! Ich möchte jedoch besonders sicher sein, bevor ich mit meinem Kauf fortfahre.

[brianddk]

Up

0

Down

::

> technically possible for Trezor to pull a Ledger move

I’ll interpret that question as:

1. Able to create firmware that looks legit
2. Able to release that firmware
3. Able to extract extract a previously configured master key
4. Able to transmit that data online
5. Able to obscure these actions from the user

Answers:

1. No, the firmware would IMEEDIATELY fail validation checks
2. Yes
3. Yes
4. Yes, the compromise would need to happen on firmware and suite AT THE SAME TIME. The firmware has no network
5. No, the firmware would IMEEDIATELY fail validation checks

I suspect (but don’t know) that the same is true for Coldcard since I thought they used the STM32 microcontroller at some point which has a READ as well as WRITE mode.

Much of this deals with reproducible (validated) builds. I discuss it in more detail in this thread below

https://www.reddit.com/r/TREZOR/comments/13k92nw/

[SilverTruth7809]

Up

0

Down

::

On every boot trezor checks if the firmware is signed by satoshilabs, if not you get a warning message.

If you load your own firmware your trezor gets wiped first.

[DarkonJohn]

Up

0

Down

::

I’m going into tin-foil-hat territory, but it seems like one thing we must trust when installing firmware from the manufacturer (Ledger or Trezor) is that they don’t have any ill intent.

If they do (under pressure from a government or if taken over by a crypto-hostile agency that wants to destroy crypto industry), it sounds like the company could product an “authentic” (digitally signed by the company) firmware which would then be able to extract and transmit your private key info (when used in conjunction with a desktop or mobile app or plug-in/extension that interacts with the hardware wallet.

This seems to be a risk for Trezor or Ledger or any other hardware wallet company that provides firmware updates.

Can someone explain in some detail how Trezor being open source reduces this risk?

While I know the source is published so it can be viewed, exactly what steps are performed by Trezor and/or the community to confirm that the firmware you download from Trezor is running the same code that was published.

Basically, I’m trying to get a better understanding of the amount of trust we are putting in the wallet manufacturers and how easily these manufacturers could betray that trust without detection if they chose to do so.

While not a huge concern right now, the Ledger thing got me wondering about this and I was thinking the risks of this become greater the more dominant any one wallet manufacturer becomes.

[BusinessBreakfast3]

Up

0

Down

::

Biggest downside of Trezor One? See how their seed recovery works.

[indomitus1]

Up

0

Down

::

I am unclear about the 15min hack with physical access as exposed by Kraken in 2019.

Was this ever fixed ?

[ehquionest]

Up

0

Down

::

How I hacked a hardware crypto wallet and recovered $2 million https://www.youtube.com/watch?v=dT9y-KQbqi4

[mcgravier]

Up

0

Down

::

> but surely someone can bypass it some way no?

There’s no known way to bypass that check. Also installing unofficial firmware completely wipes all previously existing data

[Autor]

Beiträge