Enshrined Proposer-Builder Separation (PBS)

Unlike off-chain PBS (e.g., mev-boost), enshrined PBS is codified directly into the blockchain's consensus rules. This eliminates reliance on out-of-protocol relays and middleware, creating a trust-minimized and cryptoeconomically secure separation between the block proposer (validator) and the block builder (specialized searcher).

The protocol enforces a two-phase commit-reveal scheme to prevent proposers from stealing valuable block contents. A builder commits to a block with a cryptographic commitment (e.g., a hash). Only after the proposer includes this commitment in a slot do they learn the block's full contents, making theft impossible.

Enshrined PBS creates a native auction for block space. Builders compete by submitting bids (a portion of the block's MEV) alongside their block commitment. The protocol's proposer payment rule automatically awards the slot to the highest valid bid, distributing profits efficiently without manual negotiation.

A core goal is to enforce credible neutrality. The protocol can mandate that winning builders include transactions meeting specific inclusion criteria (e.g., all transactions from the public mempool unless they violate consensus rules). This combats transaction censorship by dominant builders or relays.

By integrating the builder market into consensus, enshrined PBS ensures the fork choice rule (e.g., LMD-GHOST) considers the full value of blocks, not just the attestations. This prevents forking attacks where a proposer could be incentivized to reorg a block to capture its MEV, strengthening chain stability.

Ethereum's research outlines a potential enshrined PBS design featuring:

• A separate builder role with its own staking and slashing conditions.
• A list of registered builders from which proposers must select.
• A builder payment to the proposer that is burned and replaced with an equivalent issuance, making the reward protocol-native and non-dilutive.

Without PBS, validators who propose blocks have a monopoly on Maximal Extractable Value (MEV) extraction from their assigned slot. This creates incentives for validators to outsource block building to specialized searchers or builders via off-chain deals, leading to centralization of block production power in a few sophisticated entities.

The protocol's goal is to formally split the proposer (who chooses the block) from the builder (who constructs it). This ensures the proposer's role is simplified to selecting the most valuable, valid block from a competitive market of builders, rather than being the sole entity capable of constructing it.

By making MEV rewards credibly neutral and contestable through a builder market, enshrined PBS aims to reduce the economic advantage of large, sophisticated validator pools. This helps preserve the decentralization and censorship-resistance of the validator set by lowering barriers to entry for smaller stakers.

A key protocol goal is to make transaction censorship more difficult. With many builders competing, a transaction excluded by one builder can be included by another. The proposer, incentivized purely by the block's bid, is less able to discriminate against specific transactions, strengthening network neutrality.

Off-chain PBS (like mev-boost) is a temporary, trusted solution that relies on relays. Enshrined PBS seeks to bake these separation guarantees directly into the consensus protocol, eliminating trust assumptions and relay dependencies, making the system more robust and permissionless.

The design aims to create a more efficient market for block space. Builders compete on price (their bid to the proposer) and quality (including optimal MEV). This directs MEV revenue more transparently to stakers (via bids) and simplifies the validator's operational requirements.

ePBS introduces a two-phase block production process within the protocol itself. A Builder creates a full block, including transactions and ordering, and commits to it with a bid. A separate Proposer (a validator) selects the most profitable bid and commits to the block header, without seeing the full contents, ensuring they cannot censor or manipulate transactions.

• Builder: Specialized node that aggregates transactions, extracts Maximal Extractable Value (MEV), and constructs an execution payload. Their role is competitive and resource-intensive.
• Proposer: A staking validator whose role is simplified to selecting the header from the highest-bidding builder. This separation prevents proposers from needing sophisticated MEV infrastructure.

The system is driven by a commit-reveal scheme. Builders submit cryptographically committed bids for their blocks. The winning proposer only sees the header and a commitment. The full block is revealed and validated after the proposer's commitment, preventing last-second attacks and ensuring the proposer cannot steal the builder's work.

ePBS differs from the current out-of-protocol PBS used by mev-boost. mev-boost is a trusted, off-chain marketplace reliant on relays. ePBS brings these guarantees into the consensus layer, eliminating trust in third-party relays and making the separation cryptographically enforced and protocol-native.

• Decentralization: Reduces validator centralization pressures by separating MEV capture from staking.
• Censorship Resistance: Makes transaction censorship more difficult and detectable at the protocol level.
• Efficiency: Allows for specialized, competitive block building markets.
• Simplicity: Provides a cleaner, more secure long-term solution than auxiliary systems.

ePBS is a major research direction, not yet implemented. Leading proposals include:

• Two-Slot PBS: Separates the roles across two consensus slots for robustness.
• Single-Slot PBS: Aims for faster finality within one slot. Implementation is expected to follow other core upgrades like EIP-4844 (Proto-Danksharding) and Verkle Trees.

A cryptographically signed promise from a validator (proposer) to honor the outcome of the PBS auction. In enshrined PBS designs, this commitment is submitted on-chain, binding the proposer to the selected builder's block header. This is a critical trust-minimization component, as it allows the builder to safely reveal their full block with the guarantee it will be published.