In the world of hardware crypto wallets, security is sacrosanct. However, a recent public experiment conducted by renowned researcher Cole has made me and the entire crypto community seriously rethink the boundaries of this security. The test results, during which a Coldcard Mk3 was compromised, proved to be both alarming and instructive.

The researcher went to extreme lengths to test the device's resilience under real-world conditions. He generated a secret phrase on a Coldcard Mk3 running firmware 4.1.8 and created five different wallets based on it, funding each with an identical amount of 10,800 satoshis. The goal was simple: to find out how quickly attackers could discover and drain these addresses if the protection proved insufficient.

Methodology and Vulnerable Points

As part of the experiment, various protection configurations were tested: from a standard wallet without a password to variants with a BIP39 passphrase (one, two, and three words) and even one with an arbitrary account number. All addresses used the single bc1q standard and a shared key derivation path. Notably, as it turned out, the issue potentially affects not only the Mk3 but also newer models — the Mk4, Mk5, and Q.

Technical difficulties also arose during the setup process: direct export of public keys to the Sparrow program caused a crash, rendering the device unresponsive to commands. The researcher had to resort to saving CSV files to an SD card, which further underscores the unpredictability of working even with legitimate software.

Staggering Result

The most alarming outcome was the instantaneous draining of the base wallet without a password. The funds were moved to a new, unknown address almost the moment the mempool opened. This means hackers use automated bots that scan the blockchain for weak or compromised keys and react instantly to any transaction.

The remaining four wallets, including those protected by a passphrase and even the one with a non-zero account number, remained untouched. However, the very fact that the secret phrase is initially considered fully vulnerable calls into question the reliability of the entire Coldcard ecosystem in the eyes of many users. Monitoring of these addresses continues, and it reminds us: security is not a static state but a constant arms race.

My expert opinion: This experiment is a vivid confirmation that a hardware wallet is only part of the protective perimeter. Having a BIP39 passphrase (the 25th word) is not just a recommendation but a critically important barrier against automated attacks. Relying on "standard" protection without additional measures is like playing Russian roulette with your own assets. It is high time the industry reconsidered default security standards for hardware devices.