Written by: 0xTodd
Taking advantage of Zircuit's $ZRC official TGE, let's talk about an interesting mechanism regarding the sequencer in this project.
Zircuit has its own L2, characterized by a solution called 'Sequencer Level Security / SLS'.
We all know that currently, the entry and exit of L2 transactions are through the official sequencer of the project party. Of course, due to the existence of ZK/challenge proofs, we generally do not worry about sequencer malfeasance.
Overall, the sequencer plays a neutral role, always executing each transaction with integrity.
So, although the sequencer is neutral, can we make an effort to have it join the good camp?
Thus, Zircuit enabled this SLS mechanism, aimed at isolating 'malicious transactions'.
Normally, how does an L2 transaction get on-chain? It's a simple four-step process:
1. User initiates a transaction and broadcasts it.
2. Transaction enters the mempool and waits.
3. The sequencer acts as a neutral party, helping to package it into a block.
4. Transaction goes on-chain.
However, based on the SLS mechanism, this step has turned into five steps:
1. User initiates a transaction and broadcasts it.
2. Transaction enters the mempool and waits.
3. The sequencer acts as the good camp, using some tools to check whether the transaction is malicious.
4. If not malicious, help package it into a block.
5. Transaction goes on-chain.
But what if there are suspected malicious transactions? Changes start from step four:
4. If suspected malicious, enter the isolation pool.
5. The isolation pool review is correct, and the sequencer continues to help package it.
Or:
4. If suspected malicious, enter the isolation pool.
5. If the isolation pool review finds it indeed to be a malicious transaction, then refuse to package it on-chain.
The standard for this SLS check for maliciousness may use some open-source libraries and let AI assist in the judgment.
In the future, there is hope to achieve some effects, for example: stolen assets may never be transferred or cross back to L1. This is still very meaningful given the currently harsh dark forest environment on the chain.
Of course, since it is checking transactions, there may inevitably be some collateral damage. However, I understand that by improving the algorithm of the isolation pool, it can minimize such issues as much as possible.
This is a double-edged sword; the blockchain emphasizes permissionless, which slightly contradicts SLS. From the perspective of normal users, such an L2 is indeed somewhat safer.
However, overall, I feel that although there is a slight impact on permissionless, the increase in safety, especially in protecting inexperienced users, is still quite worthwhile.
Attached at the end: The original paper on Zircuit's SLS mechanism: https://arxiv.org/html/2405.01819v1
