Bolt
- Name: Bolt
- URL: https://docs.boltprotocol.xyz/
- Category: out-of-protocol proposer-commitment infrastructure / Ethereum preconfirmation protocol / PBS constraints-and-proof extension
- Tags: ethereum-ecosystem
- Summary: Bolt is best understood as proposer-commitment and preconfirmation infrastructure rather than as just a validator plugin or mev-boost fork. Its docs and repository describe an out-of-protocol system that lets users request commitments from proposers, lets proposers translate those commitments into block constraints, and extends the PBS pipeline so builders attach inclusion proofs that proposers can verify before signing a header. The reusable mechanism insight is that Bolt tries to make fast confirmation and inclusion commitments compatible with PBS by turning proposer promises into a new constraints-and-proof layer, while relying on optimistic slashing-style enforcement rather than on in-protocol finality changes.
- What it does:
- Lets users request commitments such as preconfirmations from Ethereum block proposers
- Adds a commitments API for user-to-proposer interaction and a constraints API that extends builder / relay communication in the PBS pipeline
- Allows proposers or delegated entities to submit signed transaction constraints for specific slots, including top-of-block or atomic-bundle style requirements
- Requires builders to produce bids with inclusion proofs so proposers can verify that committed transactions are present before signing the block header
- Provides a sidecar / mev-boost style integration path plus contracts for proposer registration and permissionless dispute resolution around attributable faults
- Includes fallback logic where proposers can self-build if relay responses do not satisfy the committed constraints
- Key claims:
- Bolt’s introduction page says the protocol enables Ethereum block proposers to provide credible commitments about block contents and explicitly frames preconfirmations and inclusion lists as the motivating primitives
- The docs say Bolt v1 enables sub-second preconfirmations on Ethereum through proposer commitments, which makes it a useful comparison class for any system trying to sell faster confirmation without waiting for protocol-native changes
- Bolt’s docs say the system is out-of-protocol and leverages restaking for economic assurances, operating in an optimistic failure mode where proposers can breach commitments but are penalized when they do
- The repository README says Bolt is fully compatible with PBS and is implemented as a forked / extended MEV-Boost stack, which is analytically important because Bolt is trying to change the execution supply chain without replacing it wholesale
- The technical flow says proposers forward committed transactions to relays as constraints, builders subscribe to those constraints and return bids with inclusion proofs, and proposers locally verify proofs before signing; this turns “preconfirmation” into a pipeline redesign rather than a simple promise message
- The builder API docs show Bolt also standardizes delegation and revocation of constraint-submission rights, which means authority over commitments can be reassigned instead of remaining hardwired to one validator key path
- The fallback-to-self-build path matters because it reveals where ultimate execution authority still sits: proposers can reject invalid relay proofs and build directly, preserving proposer accountability at the edge of the PBS market
- The strongest comparison frame is not generic fast-confirmation UX; it is “an out-of-protocol constraints market layered on top of PBS, with proposer accountability backed by economic penalties and proof-carrying bids”
- Whitepaper: No canonical standalone Bolt whitepaper or litepaper surfaced in this pass. The clearest current sources of truth were the official docs, the repository README, and the builder-API specification pages; see
../whitepapers/bolt-primary-sources-2026-05-10.md. - Sources:
Internal linkages
-
Strongest comparison points: puffer-preconf, spire, and mev-boost.
-
Last reviewed: 2026-05-27 UTC