Crowdfunding USB Security Key for Email- and Data-Encryption - Nitrokey Storage
jan at nitrokey.com
Mon Nov 23 08:54:56 CET 2015
Am 21.11.2015 12:07, schrieb Peter Lebbing:
> On 21/11/15 09:00, Jan Suhr wrote:
>> All serious findings are fixed already. Look for the "Note" at the end
>> of each issue description.
> I suppose by "serious" you mean "defined as 'Critical' in the pentest"?
> There are unfixed issues with severity "High":
> NK-01-008 OTP can be unlocked by replacing Smart Card (High)
2nd factors are usually not access protected at all e.g. may have a
display (which allows funny hacks). We introduced PIN-protection of
OTPs as an optional feature because we don't have a physical button. If
an attacker has physical access to replace the smart card, he could also
press a hypothetical button or read a hypothetical display. The PIN
isn't aiming to protect against physical attacks hence it's not in our
> NK-02-006 Micro SD and Smartcard Slots lack ejection switch (High)
An ejection switch doesn't make any sense to me. Note that ejection
switch could only be triggered if a card is ejected while the device is
powered. Furthermore any pupil would be able to use a soldering iron to
circumvent an ejection switch.
> Personally, I don't really see yet why the latter is so important;
> however, gaining the ability to issue OTP's by simply inserting my own
> OpenPGP card with my own PIN seems serious? Do I misunderstand it? Or
> it not part of the threat model because the attacker is unable to
> extract the key used for OTP generation?
Right, not part of the threat model and keys can't be extracted. Also an
ejection switch wouldn't help here because a card could be replaced
while the device is powered off which renders and ejection switch
> Anyway, thanks for all your work on the Nitrokey series! I think it's
> great you put so much effort into creating these nifty devices.
Thank you. :-)
More information about the Gnupg-users