If anyone fancies having a quick look at a cryptosystem I shamelessly prodded a llm to work out with me, I made a proof of knowlege scheme for my game (a monero fork, so all data exchanged goes via the chain, all public, no private data exchnage). I would be grateful if someone could check whether it looks solid. I coded it and it works, I just want to know if it looks exploitable. The

stakes aren't high at all, it is meant to be a riddle system for the scripting engine. 36 lines of text, the crypto scheme itself is a dozen lines: paste.debian.net/hidden/e27f9150

Actual code if interested: git.townforge.net/townforge/townfor…/src/branch/cc-tmp/src/cc/cc_pok.cc (and .h)

<jpk68:matrix.org> Is there any rough date for when the Trail of Bits audit will be released publicly?

<atomfried:matrix.org> eprint.iacr.org/2026/1202

<atomfried:matrix.org> maybe that is interesting for range proofs

moneromooo: The proof of knowledge is basically a standard Schnorr signature of the nonce N. Technically, it proves the knowledge of the secret scalar x but not the knowledge of the riddle solution S. You need the mapping from S to x to be invertible (block cipher, not a hash).

Oh that's quite enough for me to assume you know S if you know a hash of S + nonce.

(assuming I'm not missing your point)

Thanks for looking.