[madlabman]

Adds an EIP-8148 sweep-threshold system contract based on the EIP-7002 queue and fee mechanism.

[lightclient]

Generally looks good! One thing we're doing a bit different with the EPBS is we're allowing the system address to reset the inhibitor by making a call with non-empty calldata (see: #49). I think we should add the same behavior here.

[madlabman]

Generally looks good! One thing we're doing a bit different with the EPBS is we're allowing the system address to reset the inhibitor by making a call with non-empty calldata (see: #49). I think we should add the same behavior here.

Hey @lightclient! I think it wouldn't hurt, so I've added the flow. I have a slight concern regarding the approach - whether it's a good idea to use an unspecified input as a way to disable the contract, as opposed to using a magic byte. For now, I've kept it the same way it is in the builder contract to avoid multiple approaches. If you have any input on this, it would be great to hear.

P.S. Just a nitpick: in the mentioned PR, the contract comment says the queue is drained by the call with some data, which is not entirely accurate.