Hey guys, I got referred to this problem and wonder if what we’re building at Peanut would help here. I think it would be useful for 1-of-n safes, but I’d like input if it’s helpful in a more general way.
It seems like the main problem is that Safes are not guaranteed to have the same address across chains, and that mishaps like the Wintermute mishap are best avoided.
At Peanut we’re developing the Smart Cheque, an abstraction that lets you e.g. create address-agnostic & chain-agnostic transfers. The way this would work is you create an initial secret-protected deposit on chain A. Then, from chain B, the target gnosis safe initiates a transaction to claim the deposit. This would bridge the funds over Axelar (or another AMB), and they’d then arrive with no possibility of getting the address wrong.
The advantages of this:
- can’t get receiving address wrong
Disadvantages:
- 2 step process (1. create deposit & 2. claim deposit)
- if it’s a simple single-secret protected deposit, the person with knowledge of this secret could “steal” the funds - this is problematic for Safes with more than 1-of-n trust assumptions
Thoughts? In principle you could have multi-sig protected deposits, but I wonder how cumbersome that would be.