Brothers of DAOs and vaults, really don’t put too much faith in multisig just because it exists.
Everyone thinks a 3/5 or “some fraction” multisig is solid, but there’s a fatal underlying assumption: that those few people will actually go carefully review proposals. What’s the reality? A while back I replayed on-chain a vault that got drained of 30% of its assets. Three people clicked to authorize: one was on vacation and completely disconnected, another big brother didn’t even scan the transfer details and just hit approve. The transaction went through flawlessly on-chain, no private keys leaked—purely people blindly signing without knowing.
In the end, multisig can only assemble the required number of people; it can’t restrain human laziness and attention drifting.
Recently I switched to the testnet and ran @NewtonProtocol . I found that VaultKit has changed its solution. It doesn’t bother with “is there enough signatures?”—instead it tackles “why should people have to review.” The vault’s setup compiles the logic directly onto the chain via Rego. The whitelist and per-transaction limits are hard-coded. Each time you want to move funds, the intent must first pass AVS node signature verification, and only then will it generate BLS aggregated credentials and allow it to proceed. Add in TEE isolation and a slashing/penalty mechanism tied to $NEWT itself—every step is locked down by cryptography.
Its logic is: execute only when conditions are met, instead of relying on human brains to judge right from wrong.
Of course, the cost is obvious too: zero flexibility.
Because the rules are fixed upfront, if you run into an unexpected emergency, you can’t just privately message a few people in the middle of the night to take the green channel for an urgent fund transfer. The rigid rules block you—you have to obediently go through the long, tedious rules update process. The question of whether the vault’s defense should be dead-serious deterministic certainty or flexible human intervention is, in the end, a deadlock. $BTC
Right now I’m only using small positions to test the defense boundaries of this ruleset. For large funds’ defense, when you operate in practice, would you rather trust the code’s rigid rules, or trust your brothers’ judgment? #newt $ETH
Everyone thinks a 3/5 or “some fraction” multisig is solid, but there’s a fatal underlying assumption: that those few people will actually go carefully review proposals. What’s the reality? A while back I replayed on-chain a vault that got drained of 30% of its assets. Three people clicked to authorize: one was on vacation and completely disconnected, another big brother didn’t even scan the transfer details and just hit approve. The transaction went through flawlessly on-chain, no private keys leaked—purely people blindly signing without knowing.
In the end, multisig can only assemble the required number of people; it can’t restrain human laziness and attention drifting.
Recently I switched to the testnet and ran @NewtonProtocol . I found that VaultKit has changed its solution. It doesn’t bother with “is there enough signatures?”—instead it tackles “why should people have to review.” The vault’s setup compiles the logic directly onto the chain via Rego. The whitelist and per-transaction limits are hard-coded. Each time you want to move funds, the intent must first pass AVS node signature verification, and only then will it generate BLS aggregated credentials and allow it to proceed. Add in TEE isolation and a slashing/penalty mechanism tied to $NEWT itself—every step is locked down by cryptography.
Its logic is: execute only when conditions are met, instead of relying on human brains to judge right from wrong.
Of course, the cost is obvious too: zero flexibility.
Because the rules are fixed upfront, if you run into an unexpected emergency, you can’t just privately message a few people in the middle of the night to take the green channel for an urgent fund transfer. The rigid rules block you—you have to obediently go through the long, tedious rules update process. The question of whether the vault’s defense should be dead-serious deterministic certainty or flexible human intervention is, in the end, a deadlock. $BTC
Right now I’m only using small positions to test the defense boundaries of this ruleset. For large funds’ defense, when you operate in practice, would you rather trust the code’s rigid rules, or trust your brothers’ judgment? #newt $ETH