@januszgrze Not to detract from the proposal, but this is an implementation detail.. an L2 can have a multisig that could instantly steal all money in the bridge as well (and many L2s on Ethereum do have one).
@abdel I am wondering why you chose a STARK-specific TZE vs a collection of opcodes that would enable building various flavors of validity rollups? For example, we could instead add recursive covenant, merkle path verification, and prime field arithmetic opcodes. This would enable building verifiers in Script for Circle-STARKs, Halo2, and other popular proving systems. Then the L2 designer can have more options to decide what proving system to use. Personally I prefer this approach vs a STARK-specific TZE like what is proposed here.