airgap-vault VS keystone-se-firmware

I did get a pretty quick reply by posting this issue on the airgap github. Their implementation of BIP-39 passphrases is at the address level. This is unfortunate because a BIP39 passphrase generates a unique private key with its own set of addresses. I'm going to let this percolate for a bit...

Fast: https://github.com/airgap-it/airgap-vault (Get old android, flash lineageOS for good measure then airgap)

You might get a kick out of the airgap vault project: https://github.com/airgap-it/airgap-vault

The project has apps for mobile, which make turning an old phone into an hardware wallet pretty easy, as long as your phone can support it (more on that in a bit). If you're technically minded you can build the project yourself and run it as a web app: https://github.com/airgap-it/airgap-vault (Note, it's a JS project. I had issues building with npm but no problems building with yarn which the dockerfile also uses). This will allow you to turn an old desktop or laptop into hardware wallet. I wanted to try on a raspberry pi, but I don't have a camera and have read about some issues getting chromium to recognize pi cameras.

The project is completely open source and free. Our app is also reproducible, however, due to some recent issues where builds were randomly not reproducible, the status on walletscrutiny is currently "not reproducible" (if you read the analysis, it will state that the latest version is in fact reproducible, but until mentioned issue is addressed, it will stay "not reproducible"). We're working with walletscrutiny to resolve this and it looks like we have resolved the issue with the latest release, but it will take some time to verify it. (see https://github.com/airgap-it/airgap-vault/issues/127).

Thanks for giving our wallet a shoutout. A few months ago someone asked us a similar question regarding the "popularity" of our wallet, you can read my response here: https://github.com/airgap-it/airgap-vault/issues/74. In summary, I think it's a combination between people assuming that a HW device must be more secure, and a lack of marketing on our side.

Are you planning to use AirGap as a co-signer in a Bitcoin Multisig setup? This is currently not possible, but we are working on bringing our Bitcoin implementation up to speed and support the latest standards. You can read more about it here: https://github.com/airgap-it/airgap-vault/issues/73

Keystone's firmware of gen2 is already open source, and gen3 I think is being released as 100% open source soon. So you can check everything, and build it yourself. Otherwise you can double-check the checksum of the downloaded binary before you put it into your Keystone.

that's because it's verifiably not true - the chip they're using isn't documented (huge red flag when mentioning using a SE) and their "source code" still contain binaries blob for critical functionalities such as https://github.com/KeystoneHQ/keystone-se-firmware/blob/bitcoin/source/driver/ecdsa_v2.lib

A: The answer is NO. To make sure there's no access of the firmware, we have made our firmware open-sourced for the whole community to examine that we didn’t have any backdoors or misconduct from ourselves. The firmwares are released by Keystone official to ensure that no malicious third party can get into your device. We also use QR code data transmission which is transparent and verifiable to prove that there’s no secrets inside the data.

We use a EAL 5+ bank-grade Secure Element. We cannot reveal the secure element to public because of the IP of the manufacturer. But to be as transparent as possible, we have open-sourced the Secure Element firmware and the datasheet (We are the first hardware wallet company who did so). If you are hoping to further verify the operations of the Secure Element, we can send you a development board under NDA. You can find more infos in this blog.

The signature nonce generation is closed-source in Keystone: https://github.com/KeystoneHQ/keystone-se-firmware/issues/8

open source firmware