This proposal defines the necessary interface that decentralized applications (dApps) must implement to sponsor a portion of the required gas for user operations utilizing a Paymaster that supports this standard. The proposal also provides a suggested code implementation that Paymasters can include in their current implementation to support dApp sponsorship. Partial sponsorship between more than one dApps may also be achieved through this proposal.
This could indeed open up exciting possibilities for tokenomics, such as offering gas discounts based on token holdings or other criteria.
Given that this ERC works alongside ERC-4337, how does the gas efficiency of transactions using this partial sponsorship method compare to standard ERC-4337 implementations? Are there any estimates on the potential gas savings for users or additional costs for dApps when implementing this sponsorship model?
Commented on the ERC.
The problems I see is that there is full trust between the paymaster and those sponsors: that is, a sponsor is tied to a specific paymaster, and the paymaster must have a “whitelist” of audited sponsors, otherwise, each of them is vulnerable to attacks - either denial of service attacks or actual funds steal.
In general, a paymaster must be very careful on external code it executes. The validation rules (ERC-7562) are there to prevent an attack by paymaster on the system (on bundlers). By calling external contracts, the paymaster exposes itself to other contracts that might breach those rules - and thus get the paymaster itself blamed for the failures.