Who this is for: you watched or read about the Coldcard seed flaw and thought "what if there was no single secret to steal, lose or guess?" This is the written companion to PlainCrypto Video #5. It explains an idea called threshold signing and uses one product, Frostsnap, as a case study of how it works. It is not a recommendation to buy it, and nothing here says it is safer than any other wallet.
The old model: one secret, one point of failure
A traditional hardware wallet is one device, one seed phrase, one master secret. Whoever holds those words holds the coins. That's simple, and it's why a lost backup, a stolen backup or a guessable seed can each be fatal.
The usual answer is multisig: you need, say, two out of three separate keys to approve a payment. It has worked for years. But it can be fiddly to set up, and you have to back up extra information about the wallet, not just the keys.
The new idea: threshold signing (FROST)
Threshold signing takes a different route. You choose a group of devices, say three, and a threshold, say two. Each device holds only a share of the key. According to the project's description, no device ever holds the whole thing, and the whole key never exists in one place, even when you set the wallet up.
A 2-of-3 example. Any two devices can sign; none of them ever holds the full key.
To spend, any two of the three devices sign together. The result is a single signature, so on the blockchain it looks like an ordinary payment, with ordinary fees. The project also describes being able to change the group later, adding or removing devices, without moving your coins to a new wallet.
It's also designed so your phone or computer contributes some of the randomness when the key is created. Remember the Coldcard flaw: the whole disaster came from weak randomness in one place. Mixing in more than one source is a sensible design goal, but it's a goal, not a guarantee.
How the three models compare
| Single seed | Classic multisig | Threshold (FROST) | |
|---|---|---|---|
| What protects you | One secret | Several separate keys | Shares of one key, none complete |
| On the blockchain | One signature | Several keys and signatures | One signature, looks ordinary |
| Backup effort | Write down the seed | Keys plus wallet information | Per device, in the maker's own format |
| Track record | Many years | Years | Young |
The honest trade-offs
This is the part the excitement tends to skip. These points come mostly from third-party reviews, not first-hand testing, and a young product changes fast, so confirm each one on the maker's own site before you rely on it.
| Trade-off | What was reported |
|---|---|
| It's young | The hardware launched in 2025, and the underlying FROST code was described as experimental in its early days. Fewer years of testing. |
| No secure element | Reviewers note there's no secure element chip. |
| No audit found | When the video was made I couldn't find a published independent security audit. That is absence of evidence only: search again yourself. |
| More than one device | A retailer page says at least three devices are needed, so it costs more than one ordinary wallet. |
| Bitcoin only, cable only | Reviewers report it supports Bitcoin only and connects by cable rather than QR or fully offline signing. |
| Different backups | Backups are per device, in the maker's own format, not the usual twelve or twenty-four words. One reviewer says importing a 24-word seed is possible, so it isn't completely closed off. Check the maker's documentation, and understand your backups before you put real money in. |
None of that makes it bad. It makes it a tool for people who are happy to test with a small amount first.
Four questions to ask any new wallet design
Clever design can reduce risk. Only testing and time can prove it.
Whichever wallet you choose, the last step is always the same: one wrong character in an address and the coins are gone.
Check the address before you send →Sources
Product details are from third-party reviews and may change. Education, not financial advice. I'm not an advisor.
Read first: The hardware wallet flaw that emptied thousands of wallets →