Bitcoin Forum

The issue in a few words (from the Tangem team):


The statement from the company: https://www.reddit.com/r/Tangem/comments/1hougo1/comment/m4jygh9/?context=3

Article: https://cointelegraph.com/news/tangem-security-vulnerability-fixed-private-key-leak

I have  activated ^{around  a year ago} my Tangem via SEED but never communicated with support, thus it would not impose any issue for me, nevertheless I have updated my wallet to the latest release, ^{thanks for head up}.


BTW. Just for curiosity I have  simulated ^{right now} the communication with the support ^{to see what the relevant message contains} and have found that there is an opt to edit what is going to be send to them.

I thought you were using Foundation Passport. Or probably you have many of the hardware wallets.

There was a post (https://bitcointalk.org/index.php?topic=5524807.msg64903233#msg64903233) about it on collectibles which I also posted on. I think I will prefer to use wallet on airgapped device instead. Some people might have lost their coins thinking hardware wallet are very safe.

OmegaStarScream thanks for making this understandable.

To keep  the main stash in BTC I do  use  Foundation Passport ^{batch 2}.However for on the go spending I utilize Tangem 2 which is used to top up my crypto cards ^{largely  with USDT} . I have mentioned this here (https://bitcointalk.org/index.php?topic=5441422.msg64864921#msg64864921). You may look on my posts ^{particularly this one (https://bitcointalk.org/index.php?topic=5312391.msg64087703#msg64087703)} , relevant to Tangem in dedicated thread. Tangem is very easy-to-use in the course of every day routine ^{that is why I added this wallet to my arsenal.}

Nobody should be surprised when things like this happens to closed source hardware wallets and their crap app :P
I really don't understand why anyone would use tangem products when they already have great open source alternative called Satochip.
They are both in exact credit card format but users have much more freedom and choice with all Satochip products.

I will remind everyone to STOP using all hardware wallets that are not open source.

Regarding me, I choose tangem ^{instead of Satochip} for its unique backup scheme. The other reason - they are going to issue the next generation ^{Visa approved} cards with enabled  crypto payments via  Tangem Pay . Thus the use of their current cards  permit me to get better understanding of Tangem technique. Sure both Tangem and Satochip are not recommended ^{by me} for the big BTCstash.

I'm active Tangem user and recommend it for small sum at  on-the-go payment.

@dkbit98, just of my curiosity, I wonder whether you  have the hand on experience in working  with Satochip.

Who let their programmer add code which log seed phrase or other sensitive data? Anyway, looking at their blog post[1] makes it clear Tangem company trying to downplay this security vulnerability. They use term "Potential Vulnerability" which is wrong since people can reproduce the security vulnerability.

[1] https://tangem.com/en/blog/post/tangem-resolves-log-issue/ (https://tangem.com/en/blog/post/tangem-resolves-log-issue/)

I would have said 'low risk' instead of potential.

If you didn't sent the logs to them with a support request within the time before the info was overwritten or auto purged the data never left the phone.

If you have other things on your phone that are snooping on other logs and reading / scanning / sending that info to malicious people you have many many many larger issues. Since the info was in a location that only the Tangem app should have access to.



The part that sent the keys was and still is 100% open source. So, even if the cards were open source this still would have happened.

-Dave

Nice catch. And as I have already said there is an option to see what you are sending to support and edit the message. Thus, those who sent them their log with SEED were careless people who preferred not to  take the trouble of  reading  what they are sending. I would not create a strained atmosphere relevant to Tangem.

I'm wondering about the same thing. Maybe this is unintended behavior, who knows really? They did claim it was a bug. I'd be surprised if one of their QA tester or dev noticed this in production and greenlit it since seedphrases safety is crucial for their product brand image. Then again we know companies make confusing decisions that make people distrust them like Ledger.

Edit what exactly? You mean it just opens a notepad and lets you arbitrarily modify the log file?

Assuming there wasn't some sort of vulnerability like this in the first place, who would want to do that? Most people don't read log files. It's mainly a feature for the developers and its sent automatically without modification.

Fair point, since it require action from user or the device is compromised (either due to jailbreak or use very old OS).


Being unintended is possible, since dumping data to log is a way to debug program. But QA, code reviewer, security reviewer or anyone with similar position could've noticed it and ask for change.

Tap three dots in the upper right corner of the app, tap contact support, It shows the content  of the message ^{which can be edited directly within app} plus app_logs.txt attached which can be deleted by pressing "x" at the attachment. Regarding app_logs.txt, I think  there  is a way to reach this file on my Android and analyze it . Should check this at my spare time.

UPD. Wasting no time, found the easy way how to look at app_logs.txt. At the top of message tap To field, tap Remove, isert into To-field  you own email address   and send to yourself app_logs.tx attached. ^shazam



Agreed, most people are careless.

What the actual fuck. That code was clearly not reviewed (or they don't have a proper process in place).

Going to repost something I have posted here and reddit and github and other places over time.

So more or less quoting myself.


I can go on with dozens of other examples if you want.


So the code could have been reviewed. But people missed it.

If 1000s of people reviewing the above projects over years and years missed some (after the fact) totally obvious issues like these then a smaller company missing something like this is GOING TO HAPPEN.

Or to put it another way.

OPEN SOURCE IS NOT MORE SECURE. OPEN SOURCE ALLOWS PEOPLE TO SEE WHAT IS HAPPENING. AND POSSIBLY FIND MISTAKES THAT OTHERS HAVE MADE. BUT UNLESS THE PEOPLE LOOKING AT IT SEE THE MISTAKE AND REPORT IT THEN IT'S NO BETTER THEN CLOSED SOURCE.

Ending rant.

-Dave

OK, Tangem is apparently a display-less smartcard. I admit, I don't know much about this device.

Now, what did the developers not understand, that a private key or seed of a hardware wallet should never leave the device? I wonder how such a basic paradigm of hardware wallets could be intentionally broken? (Don't ask about the Ledger crap...)

I can imagine that for support or debugging purposes such a "device" needs to provide a bit more extensive logs as there's no status lights or display or whatever. But still I don't get why a log is created on a "hot" online device with the most precious secrets of the wallet exposed.

For the longest time you could NOT get a seed from a Tangem card.
People kept freaking out about recovering it in case something happened to all their cards so Tangem gave in to them and allowed you to see your seed on your phone when you created your wallet.
Now the app has always loged a fair amount of data, none of it security compromising.

When they re-wrote the app to display your seed it was understandably logged for testing / debugging.

And then, someone screwed up and did not take the logging out when they pushed it to production.

So if you did send your logs out to them for support before they were overwritten they got a copy of your seed.

Which proves that people love to talk about open source but nobody reads / understands a lot of it even when they use it.

-Dave

Actually, this proves the importance of being fully open source.

If the app wasn't fully open source, the logging might not have been spotted.  And if someone at Tangem realized it was there, they could have used it maliciously.

But since the app is fully open source, the logging code was found, leading to it being removed.

Even if the average Joe Shmoe doesn't understand why open source matters, being open source helps to keep them safe.

Being open source matters.  I'll never trust my Bitcoin to closed source code.

Yes I worked with them on a project for creating special edition designed Satochip cards.
This was announced, posted publicly in bitcointalk forum, and I think most of the cards sold very quickly.
I stand by that Satochip is 10 times better than Tangem in every way.

This is NOT firmware and I was talking about that.
They app is super crap, and it's not important it if is claimed to be partially open source when it can't be compiled.
They are deceivers and amateurs, so anyone choosing to trust them is playing Russian roulette.

Open Source or not, my advice is that less known Hardware Wallets should not be the choice for storing significant value.  If a crucial mistake is done in the coding, it is one thing for Trezor and another for Tangem.  Trezor has how many users versus Tangem?  Which I presume also increases the likelihood that somebody finds the problem in the code faster for the Trezor Source Code than for the Tangem.

Stick to either Trezor or, if further paranoid, let Hardware Wallets be maybe only the portable Hardware Wallet for when you travel and keep the stash on an Airgapped Computer.  That solves about 100 percent of these type of problems.

I have tangem card and I did use seed phrase option, I imported my ledger seed into tangem, thinking it's safe, thankfully didn't message their support so I assume I must be safe? I mean my funds are still there.

I just came to know about this issue yesterday and ranted on their discord for about two hours. I was disappointed with vulnerability itself — how can you record seed? Even literal hot wallets don't do that, and with their communications.

I am done with them and moving to trezor. Reminder for me to stick with tried, tested and reputed HW providers.

If you didn't communicate with their support during one week after your imported you SEED your stash should be safe ^{You may check this via Tangem mobile app.}  providing Ledger SEED did not leaked in a different way .
P.S. I wouldn't use old SEEDs generated by Ledger to initiate any new wallet.

I checked tangem app, there is no in-app notification to me as they said you would get if you were affected.

Regarding importing seed, my old ledger display is impossible to work with so imported seed directly into tangem. But yes, now gonna create new one with trezor.

I can relate with the fact that it is possible that most open-source projects are weakly scrutinized if at all by anybody since the job on validating the security of those projects aren't easy and it may be without a reward and most experienced white-hat hackers would not be willing to do it for free and black-hat hackers would gladly do it to take advantage.

If there are mouth watering bounties attached to reviewing these security limitations  and every bug found has a big reward, it would go a long way to make the open-source pattern serve its full purpose as participants would want to take the prize. I feel those that offer good rewards would get most of their bugs identified on time before it causes a real trouble for their users.

How do you PROVE it?

I'll never understand why people put their Bitcoin at risk by using closed source wallets, especially when there are outstanding fully open source alternatives.

Tangem... I've never heard of this wallet. This serves as an example to show that it's not enough to be a hardware wallet, it needs to be open source too to avoid such discrepancies as this. It seems like an amateur mistake, it's so absurd that even hot wallets are more secure than this type of HW.

The seed is not a problem if I can use passphrases.  ;)

But, and this point has been made the software that had the issue IS OPEN SOURCE.
The card itself is closed source, but the software that runs it (and that had the issue) it's right here: https://github.com/tangem

But open source is soooo much better:   https://www.splunk.com/en_us/blog/security/my-cups-runneth-over-with-cves.html

-Dave

I missed this thread initially, saw it got necrobumped (sort of) for some reason and also saw your previous post about open-source software still being vulnerable to being exploited despite the fact that the code is out there for all to review.

Can you just clarify something here?  Are you arguing that open-source software isn't better than closed-source because oversights have been made?  That runs contrary to everything I've ever read here and also to my own *tech ignorant* opinion, and I guess I'd ask if you agree or disagree that, all things being equal, it would be better to use a HW wallet or its software that has open-source code than using a competing alternative that has closed-source code.  Even if you don't know that the open-source code has been extensively reviewed (or reviewed at all), I just can't see how any HW wallet that doesn't divulge its code is any better in light of the argument you made.

I ask this question respectfully.

Also, I have a Tangem wallet around here somewhere that I bought as a collectible thingee.  Never used it, never wanted to, and whatever their state is now there's no way in hell I'm ever going to.

No, what I am saying is that people have the mentality of open source being "more secure" and that closed source is "less secure"
But, as has been shown both have glaring gaping holes in them at times.

Open source means others can see what is going on.
That's all. And let me point out that code with 10's of thousands of eyes on it FOR DECADES can have GAPING SECURITY HOLES FOR DECADES

https://en.wikipedia.org/wiki/Shellshock_(software_bug)

September 1989 to September 2014

So, yes in terms of wallets open source is PROBABLY better.
But, as I pointed in another thread


And that's the problem.

You can put out an open source hardware wallet with bad code that links to a software wallet with bad code and people will still buy and use it. Because open source is "better"

But, if you make a closed source one that is 100% secure, people automatically think it's bad.

It's not black and white, it's one big mess of grey. And people have to get used to living in the grey.

Because if bash, a piece of software that was on just about all *nix systems forever had a vulnerability that was there for 25 years, how well do you think any piece of crypto wallet software run by a much smaller segment of the population is going to be reviewed for vulnerabilities?

-Dave

You're generally right about the vulnerabilities that might be present in both open-source and closed-source software. However, there is still a difference between the two. Closed-source code may contain both intentional and unintentional backdoors while open-source code is less likely to include intentional backdoors, simply because developers know that their presence is likely to be discovered by the community, ^{sooner or later}.

I think it's a bit of terminology.
Open source should be considered SAFER since you (or someone) can check to see what is happening. If you (or someone) understands the code.
This would prevent the send all coins to Dave issue.

But, you can't really say it's more *secure*.

A bit pedantic on my part but when dealing with customers I have to point things like this out.

In the end discussing it here probably does not matter as much, but when talking to non tech people it's probably better to make the distinction between safer and secure.

-Dave




 

I think it's important that people understand the consequences properly. I'm with you that open-source doesn't automatically mean, it's more secure. But it's likely a better starting point than closed-source.

But with open-source AND reproducible builds you know exactly what source code yielded the executable. In my opinion only when these two conditions are met, you can then audit the source code and check how things are actually done and executed.

Not having reproducible builds means, I can't be sure what's in the executable. Is something missing or has something been added? It's possible that the open-source code doesn't show all the truth. Not being reproducible is probably less worse than closed-source, but in my opinion not very much, because it would be difficult to prove that the executable doesn't contain something that's not in the open-source code.

Of course, you need to have people who are capable of a code audit and who do it or have done it, unless you can do it yourself which is likely rarely the case either by lack of skills or lack of time or both.

The problem with relying on reproducible  is that there have been several discussions over the years where people have built / compiled from source but it was deemed not reproducible because they were using a totally different generation of software to build it.

What I would like to see too is better / more detailed instructions in the wallets github, or wherever in terms of build instructions too.

Even tell us the compiler version and os and everything else that was used to make the build.

But all you usually get is clone git and type make run and that is if you are lucky.

-Dave

Hm, I don't do such reproducability checks myself, would maybe only do it for the firmware build of my hardware wallet or so, occasionally.

I sort of assumed that the required exact build stack is part of the build instructions, maybe in the form of a Docker container build where you specify exactly the environment what's used to build from the sources. Because, how would reproducability otherwise work. You must be able to produce an exact bit-copy of the executable that's usually offered for download, too.

Yeah, been there, seen that, too.