Building a vault using blinded co-signers

Congratulations on the announcement, and the prototype. I loved your work on Elf trace as well.

I strongly believe in blind co-signers to emulate functionalities like these on bitcoin.

My only criticism, is that bootstrapping trust of co-signers is a tough work regardless of what they are signing, so maybe we should try to get as much leverage from them as possible, and I suggest that the most leverage would be to emulate the few Bitcoin script’s op_codes that are necessary to enable Elf-trace and other fraud proof application.

My reasoning is that doing so allow us to:

  1. Build a protocol that doesn’t necessarily depend on Risc0, maybe a TEE is sufficient, if not maybe TEE + a slashable bond on another chain is good enough (if you keep the logic simple, reimplementation in Solidity/Cairo can be verifiable).
  2. Make the execution of these opcodes visible on-chain, proving (or disproving) the demand for these features, as well as hopefully defeating arguments along the line of MEV.
  3. Track record of honesty and liveness of co-signers

That being said, I very much want to see this project getting a lot of adoption and usage, because adoption is really the hard part for these systems, not the details of the protocol, but that is also why I hope that bootstrapping work would also help the ecosystem have good data about what opcodes deserve to be in a soft fork next.

2 Likes