Quote from: Kruw on January 11, 2025, 09:55:36 AM

Quote from: Kruw on January 24, 2024, 06:15:29 PM

- Can WabiSabi be traced?

Not unless you are the biggest whale in a coinjoin round with insufficient liquidity.

Here's an example that seems to indicate a trace vector for a whale in a round with insufficient liquidity: https://mempool.space/tx/804caad877c077c2f368b643eb9b5dbcf3f8845a977afc2ff3a3f9e0db02f5c3

At face value, you would probably assume that the 28.72536879 BTC output must belong to either the 23.24522934 BTC input or the 21.47483648 BTC input. However, it's likely that both inputs belong to the same wallet since there is only one change output, not two. This is an extremely rare edge case, and it can be mitigated by users freezing coins on the client side, coordinators issuing more vbyte credentials, or limiting the upper bound for input registration values so that amounts in the coinjoin are more uniform. Still, it's possible these inputs do not belong to the same wallet, so it's not 100% conclusive.

I'll be a broken record, but for those people who haven't heard it, hear it now. - That's probably why it might be good OPSEC to add another layer of untraceability after a CoinJoin by sending a transaction to yourself through the Lightning Network using those CoinJoined outputs.

¯\_(ツ)_/¯