I'm developing free and open source personal secrets manager under GPLv3, yes the ui part completely vibecoded, but the cryptography algorithm and vault design is the original idea. So, I need some help, Is there anyone here who knows cryptography? I’ve already checked with all the models I have access to, and they all pointed out the obvious flaws, but every single one says the algorithm is good. In short, it’s E2EE without disclosure, signed and containing a hash chain. Of course this product is an early stage, so I’d like help now to avoid having to break the vault’s format later.
Cryptographers won't even bother to answer anything that sounds like this, they get too many such things from independent developers, and historically such things are not even interesting cryptography.
If you want their attention, first see if you (not an LLM, you) can independently re-discover known flaws in existing cryptography systems.
Then, when you can do that, see if you can find a novel flaw nobody else has found yet in something live.
Then cryptographers might start to get interested.
I'm developing free and open source personal secrets manager under GPLv3, yes the ui part completely vibecoded, but the cryptography algorithm and vault design is the original idea. So, I need some help, Is there anyone here who knows cryptography? I’ve already checked with all the models I have access to, and they all pointed out the obvious flaws, but every single one says the algorithm is good. In short, it’s E2EE without disclosure, signed and containing a hash chain. Of course this product is an early stage, so I’d like help now to avoid having to break the vault’s format later.
Cryptographers won't even bother to answer anything that sounds like this, they get too many such things from independent developers, and historically such things are not even interesting cryptography.
If you want their attention, first see if you (not an LLM, you) can independently re-discover known flaws in existing cryptography systems.
Then, when you can do that, see if you can find a novel flaw nobody else has found yet in something live.
Then cryptographers might start to get interested.
Thanks for reply, yes I already done meat analyze.
Article: https://confidex.me/cryptography GH: github.com/confidex-me/confidex