Thanks for putting together this proposal. It will definitely help with UX and adoption. A few questions:
-
How does the proposal plan to handle fee payment? Intents, in its generic form, may include many types of operations. Some of them can be simply automation tasks that the end users would like to achieve. Such tasks may not have much MEV value to extract and solvers likely may not have incentive to process these intents.
-
How do you plan to mitigate the potential state discrepancy between simulation and execution? In ERC-4337, it applied fairly restrictive standards in terms of storage access, which I think, would ban a ton of operations. In fact, if a smart contract wallet user wants the solver to open up a Uniswap V3 position on his/her behalf, it would not be possible using ERC-4337, because the position open would entail the access of non-owner storage slot.