It currently can only refer to calling contracts deployed at an address known at verification time, which doesn't make the command particularly useful. References to other contracts are usually contained in a contract's storage and read from there, so aegis_use_contract_call must be able to make the registry dependent on the storage content at least (or allow arbitrary variables?). Furthermore we must address the following errors in the current implementation:
[ ] The comment currently connects a spec to a certain contract instance, while it should instead connect the spec to the class hash, since the spec will be the same for all instances (member variables are storage variables under the hood).
[ ] The callerAddress of the metadata variable of the called contract currently is the same as the callers callerAddress, need to check if that's correct.
[ ] The contractAddress of the metadata variable of the called contract currently is the same as the callers contractAddress, which is probably incorrect.
I think one way to resolve this is to wait until there's a way to access the class hash with a syscall, then we have a value to pin the call to and that's the proper way anyway
It currently can only refer to calling contracts deployed at an address known at verification time, which doesn't make the command particularly useful. References to other contracts are usually contained in a contract's storage and read from there, so
aegis_use_contract_call
must be able to make the registry dependent on the storage content at least (or allow arbitrary variables?). Furthermore we must address the following errors in the current implementation:callerAddress
of the metadata variable of the called contract currently is the same as the callerscallerAddress
, need to check if that's correct.contractAddress
of the metadata variable of the called contract currently is the same as the callerscontractAddress
, which is probably incorrect.