L1 Auditor
Overview

L1 Auditor

Suzaku L1 Auditor is a tool that helps L1 teams audit the security of their network through the lens of decentralization.

Go to auditor.suzaku.network (opens in a new tab) now to audit any Avalanche L1.

L1 Stages

L1s are classified into 4 stages based on their decentralization progress:

  • Stage 0: Insufficient decentralization with negative impact on bridged TVL security, liveness and censorship resistance
  • Stage 1: Minimal decentralization that eliminates single points of failure for bridged TVL and liveness
  • Stage 2: Good decentralization that ensures bridged TVL security, liveness and censorship resistance
  • Stage 3: TBD

Those stages are subject to change. They are currently based on a proposal made by the Suzaku team in an ongoing discussion in the ACP (Avalanche Community Proposal) forum: Nuttymoon comment in discussion #249 (opens in a new tab).

Decentralization Metrics

The following metrics are used to determine the stage of each L1:

Stage 1

Validator set management is not controlled by a single EOA

Category: Bridged TVL Security

A breach of a single EOA could lead to validator set changes and potential loss of funds bridged to the L1. This can be enforced by setting up multisignature wallet ownership or governance-based security.

Methodology used

We analyze onchain data to determine the ultimate ownership of the ValidatorManager contract of the L1.

Each operator controls strictly less than 67% of the L1 total weight

Category: Bridged TVL Security

A single operator's breach could lead to a loss of all funds bridged to the L1.

Methodology used
  • For non-Suzaku-secured L1s
    We look at the ownership of each validator on the P-Chain: if a single P-Chain address controls multiple validators, we know that they belong to the same operator. But if validators are controlled by different P-Chain addresses, we cannot be sure that they belong to different operators.
  • For Suzaku-secured L1s
    We use the operator registry of the protocol. We can attribute ownership of all validators with certainty.

The L1 has at least 5 validators

Category: Liveness

Below this threshold, any single validator failure would halt the chain.

Methodology used

We analyze onchain data to determine the number of validators on the L1.

Stage 2

No more than 20% of the L1 total weight is controlled by a single operator

Category: Liveness

A single operator failure would halt the chain.

Methodology used
  • For non-Suzaku-secured L1s
    We look at the ownership of each validator on the P-Chain: if a single P-Chain address controls multiple validators, we know that they belong to the same operator. But if validators are controlled by different P-Chain addresses, we cannot be sure that they belong to different operators.
  • For Suzaku-secured L1s
    We use the operator registry of the protocol. We can attribute ownership of all validators with certainty.

No more than 20% of the L1 total weight is hosted on a single data center

Category: Liveness

Concentration in a single data center creates a critical infrastructure failure point.

Methodology used

We use the IP addresses of the validators to determine the data center they are hosted in.

All validators are hosted in OFAC-compliant jurisdictions

Category: Liveness

Validators in sanctioned jurisdictions expose the L1 to compliance risk and potential forced censorship of transactions.

Methodology used

We use the IP addresses of the validators to determine the jurisdiction they are hosted in.

The L1 has multiple RPC endpoints operated by different entities

Category: Censorship Resistance

Not implemented yet

A single operator's outage could lead to censorship of user transactions.

Validator ownership of more than 67% of the L1 total weight is verified at the social layer

Category: Verifiability

Not implemented yet

Stage 3

The L1 Auditor doesn't yet support this stage.