How Does Scroll Work?

Bytecode-level EVM-compatible zkEVM using SNARK proofs for transaction validity

Type-2 zkEVM aiming for full EVM bytecode compatibility. Well-documented architecture but ZK circuits are complex attack surface.

Single privileged sequencer with priority for batch submission and transaction ordering

Standard centralized sequencer model. Can extract MEV and censor transactions. Force-inclusion is available but sequencer has priority ordering.

Euclid upgrade migrates proving system to OpenVM RISC-V zkVM by Axiom

Novel proving architecture using general-purpose RISC-V zkVM. Simplifies auditing via standard Rust code but introduces new, less battle-tested proving pipeline.

Canonical bridge with L1 message queue and ZK-verified finalization

Messages pass through L1 queue verified by ZK proofs. No refund mechanism if messages are censored before proof submission.

ScrollOwner contract with four distinct timelocks governed by Security Council and team multisigs

Tiered governance with different delay guarantees per action type. Security Council and team have different authority scopes.

Transaction data posted to Ethereum L1 as calldata for full data availability

Standard on-chain DA model inheriting Ethereum security. Higher cost than alt-DA solutions but stronger security guarantees.

Three-stage pipeline: commit transaction batches, generate ZK proofs, finalize on L1

Team can revert unfinalized batches and add/remove sequencers and provers while in permissioned mode.

SCR token with 19% circulating supply and major unlock scheduled for October 2026

Only 190M of 1B total supply circulating. Next major unlock could create significant sell pressure.

Sequencer can skip messages in the L1 queue and confirm specific messages directly, enabling selective censorship. Combined with no refund mechanism, user assets can become permanently stuck in gateway contracts.

Migration to novel OpenVM RISC-V proving system changes the entire proof generation pipeline. Bugs in the new prover could delay finalization or, in worst case, allow invalid state transitions to be proven.

Team multisig retains ability to revert unfinalized batches and modify sequencer/prover sets. Combined with tiered timelock, critical upgrades could bypass longer delay guarantees through lower-tier authorization.

Major token unlock in October 2026 could trigger sell pressure that reduces ecosystem confidence and TVL, creating a negative feedback loop on protocol adoption.

High L1 data availability costs squeeze sequencer margins, potentially incentivizing transaction censorship or batching delays to optimize profitability.