Research Area: Aggregated Validity Proofs for OP Stack

It should be possible to build an aggregated validity proof service for the OP Stack. A blockchain has inputs and outputs. The inputs to an OP Stack chain are all of the L1 data which is then mapped into execution payloads via the derivation pipeline. These execution payloads are inputs to the state transition function which ultimately produces the outputs, represented as the state root and execution logs.

If an aggregation service was scanning L1 to look for the inputs to the derivation function, it should be possible to identify the next set of inputs and then pass them to [kona](https://github.com/ethereum-optimism/kona) running in [sp1](https://github.com/succinctlabs/sp1) to generate a zkp. How would we know that its the full set of inputs? For example, a single `TransactionDeposited` event could be missing which would result in the wrong state. This may mean that the zkvm would need to do the full derivation process, ie actually observe all possible L1 data. This adds a lot of extra work for the zkvm to do, so what is the minimal amount of work that the zkvm actually has to do?

Once the proof is generated, then its possible to aggregate the proofs together. This could enable "super transactions" as noted by [Justin Drake](https://youtu.be/URCH2d1cdyg?si=1UMaP9BINbVESmfU). A super transaction is a transaction that executes across multiple rollups "synchronously", you can think of it as a bundle of regular transactions that are each individually signed for single chains but make up the full intention of the user.

No source files or tests are named. Start by reading the OP Stack derivation pipeline and the linked kona and SP1 projects, then trace how L1 data becomes execution payloads and state-transition outputs. Done should define the complete input set, the minimal zkVM work, and a viable proof-aggregation service design.

Written by the indexing model from the issue text.