Glossary

Every term used in the book, alphabetized. Each entry includes a one-sentence definition for the casual reader, a 2–4 sentence expansion for the curious reader, the chapter where the term is introduced, and TradFi analogues where they exist.

Entries are appended by the agent at draft time. Do not write entries here that aren't used in a chapter.

Template

Term

Short: One-sentence definition. No jargon.

Long: 2–4 sentences. Mechanics, not history. Why it matters to a business reader.

TradFi analogue: If one exists.

First appears: Chapter X

See also: Cross-references.

Entries

Agave

Short: The reference Solana validator client, maintained by Anza (the team that spun out of Solana Labs to focus on client development).

Long: Agave ships the default block-production scheduler — central-scheduler-greedy — which uses four worker threads, one scheduling thread, and two vote threads, prioritising transactions by the formula ((Priority Fee × CU Requested) + base fee) / (1 + CU Requested). Anza's published benchmarking showed up to 80% higher fee collection with the central scheduler versus the prior thread-pool design. The Agave codebase is forked in several directions (Jito-Solana, JitoBAM, Agave Harmonic, Rakurai), and as of March 2026 approximately 86% of Solana stake runs an Agave-based client.

First appears: Chapter 5

See also: Frankendancer, Firedancer, Validator client, Slot leader.

Aggregator

Short: A protocol that reads prices across multiple trading venues and routes a user's swap to the combination that delivers the best effective price after slippage.

Long: Aggregators replaced direct-to-pool routing as the dominant retail interface on Solana — Jupiter held roughly 93.6% share of aggregator flow in Q1 2026, and more than 70% of all Solana DEX volume now flows through aggregators rather than directly to pools. On Ethereum, the equivalent layer is fragmented across CoW Swap, UniswapX, 1inch Fusion, and direct Uniswap. The trader sends an intent; the aggregator finds the route. The pool — or the prop-AMM behind the pool — becomes inventory, not interface.

TradFi analogue: A smart order router at an institutional brokerage, except that an on-chain aggregator can split a single trade across non-affiliated venues without prearranged routing agreements.

First appears: Chapter 2

See also: Solver, Prop-AMM, AMM.

Alpenglow

Short: Solana's planned consensus-layer rewrite, approved by validator vote in August 2025 with 98.27% stake-weighted approval; testnet activation 11 May 2026; mainnet expected late Q3 / early Q4 2026. Full architectural treatment in Chapter 12.

Long: Alpenglow replaces Solana's existing TowerBFT + Proof-of-History consensus with two new components — Votor (BLS-aggregated off-chain votes) and Rotor (block propagation) — and produces two load-bearing economic effects on the validator role: vote transactions are removed from blockspace (currently approximately 70% of all Solana transactions at peak), and a Validator Admission Ticket of approximately 0.8 SOL per day is introduced as the first explicit per-epoch fixed validator cost. Helius's modelling suggests the profitable-validator stake threshold drops roughly tenfold (from approximately 4,850 SOL to approximately 450 SOL). Alpenglow widens the validator base but leaves the single-leader-per-slot ordering structure intact. The Anza team's published response to the access-vs-operational revenue gap is a two-step roadmap: Alpenglow first, then Constellation (the multi-leader proposal Brennan Watt unveiled at Solana Accelerate on 5 May 2026, targeted for Q3 2026 mainnet activation contingent on Alpenglow shipping first). Chapter 12 develops the architectural treatment.

First appears: Chapter 5 (cameo); full treatment in Chapter 12.

See also: Votor, Rotor, Validator Admission Ticket (VAT), Constellation, Validator.

AMM (automated market maker)

Short: A liquidity venue that prices any swap against a pool of paired assets using a formula, with no resting bids or asks.

Long: The dominant AMM design uses the constant-product invariant x · y = k — the product of the pool's two reserves stays fixed; the price the pool quotes is the ratio of reserves. Uniswap V2 (2020) is the canonical implementation; V3 (2021) added concentrated liquidity, allowing LPs to specify a price range within which their deposit earns fees. AMMs sidestepped the problem that on-chain order books faced — that placing and cancelling orders was too slow and expensive — by replacing the book with a curve.

TradFi analogue: No clean traditional-finance analogue. AMMs are an on-chain primitive that has no NYSE or NASDAQ counterpart.

First appears: Chapter 2

See also: CLOB, Liquidity provider, Impermanent loss, Loss-Versus-Rebalancing.

Arbitrage (on-chain)

Short: Closing a price gap between two trading venues by buying the asset cheap on one and selling it dear on the other, inside a single transaction or atomic bundle.

Long: On-chain arbitrage takes two forms. DEX–DEX arbitrage closes price differences between two on-chain pools (Raydium vs. Orca, Uniswap V3 vs. SushiSwap) and is generally treated as price-discovery work the chain pays for through validator economics. CEX–DEX arbitrage closes price differences between a centralized exchange and an on-chain pool, and is harder to measure because the centralized-exchange leg is not visible on-chain. The party that ultimately pays the arbitrageur's profit is whichever venue was slow to reprice — usually a passive liquidity provider on the on-chain side.

TradFi analogue: Statistical arbitrage between two trading venues quoting the same security, e.g. NYSE vs. NASDAQ.

First appears: Chapter 4

See also: MEV, Searcher, Liquidity Provider (LP).

AVS (Actively Validated Service)

Short: A service that runs on top of EigenLayer's restaking infrastructure, secured by restaked ETH and operated by EigenLayer's operator set. AVSs include data-availability layers (EigenDA), fast-finality services (AltLayer MACH), cross-chain messaging (Hyperlane), and DePIN (Witness Chain).

Long: The AVS is the structural innovation EigenLayer introduced to Ethereum's economy in 2024–2025: a service whose security is rented from Ethereum's staking set rather than bootstrapped from a new token's validator economy. An AVS pays restaking operators in its own token (or in ETH) in exchange for restakers committing slashable stake to the service's misbehaviour conditions. As of May 2026, approximately 1,900 EigenLayer operators secure approximately $18 billion of restaked ETH across the AVS roster; the largest single AVS by restaked stake is EigenDA, EigenLayer's own data-availability service. The "Vertical AVS" trend toward specialized verification (AI-model evaluation, DePIN proof-of-location, cross-chain messaging) is the chain's 2026 architectural answer to "what is restaked ETH actually useful for?" The structural argument the chapter develops: AVSs are the third architectural layer of Ethereum's PBS + L2 + restaking trifecta, and the economic compounding of one ETH worth of stake earning revenue from N AVSs is the structural innovation; the multi-operator-set slashing exposure it creates is the structural cost.

TradFi analogue: A clearinghouse that rents its collateral pool to multiple regulated markets simultaneously — same collateral securing multiple obligations, with cross-default exposure if any market's clearing function fails.

First appears: Chapter 10

See also: EigenLayer, Slashing, Validator.

Assistance Fund (Hyperliquid)

Short: Hyperliquid's on-chain treasury that executes autonomous HYPE buybacks from approximately 99% of validator-operated perpetual fee flow. Cumulative buybacks crossed $1 billion in the fifteen months from HYPE's November 2024 launch through February 2026.

Long: The Assistance Fund is the mechanism that converts trading fee revenue into HYPE token price support. Weekly buyback rate as of January 2026 ran approximately $1.7 million, accelerating in high-volume periods. In December 2025, the community proposed and executed a burn of approximately 37.5 million HYPE (roughly 4.17% of total supply, valued at approximately $912 million at execution prices); an additional ~13% of circulating supply was proposed for a future burn in January 2026 under the same governance pattern. The structural argument the chapter develops: Hyperliquid runs at structural breakeven by design — revenue and buybacks net out at the protocol level, and DefiLlama's published "Earnings $0" classification across all periods is the accounting reflection of that design choice rather than a P&L failure. The chain's economic value-capture mechanism is HYPE token holders, not Hyperliquid Labs equity holders.

TradFi analogue: A corporate share-buyback program, except executed autonomously on-chain against a fixed-supply token and visible per-transaction rather than per-quarter.

First appears: Chapter 5 (cameo); full treatment in Chapter 9

See also: HLP, HyperCore, HIP-3.

Backrun

Short: A transaction the searcher places so it executes immediately after a target transaction whose effect the searcher has predicted.

Long: Backrunning is the other half of a sandwich attack and is also the technique used to capture pure arbitrage opportunities that someone else's transaction is about to create. Unlike a frontrun, a backrun does not require the searcher to see the target's contents; it only requires the searcher to predict what the target will do to the relevant pool. Backruns are not, on their own, extractive — when a backrun is used to close an arbitrage gap, the searcher is performing a price-discovery function. When paired with a frontrun, it becomes the close-out leg of a sandwich.

TradFi analogue: An HFT desk that fills behind a known large order to capture the residual price impact.

First appears: Chapter 4

See also: Frontrun, Sandwich attack, MEV.

BAM (Block Assembly Marketplace)

Short: Jito Labs' TEE-based confidential block-construction layer for Solana, live since July 2025.

Long: BAM splits a previously-monolithic function: transaction sequencing (decision of what goes in which slot and in what order) is delegated to BAM Nodes, while execution remains with the validator. BAM Nodes run inside AMD SEV-SNP hardware enclaves that keep transaction contents encrypted until execution and produce cryptographic attestations of the ordering decision. By March 2026, approximately 28% of Solana network stake ran the Agave JitoBAM client. BAM has no direct user fees; its revenue model is a 6% tip split between Jito Labs and the Jito DAO. BAM also supports Application-Controlled Execution plugins — Jump Crypto's BAM Maker Priority plugin (April 2026) gives market makers deterministic top-of-batch execution every fifty milliseconds.

TradFi analogue: A clearinghouse that ran inside an HSM — the matching logic is sealed in tamper-evident hardware, with audit attestations replacing the requirement for everyone to trust the matching engine's operator.

First appears: Chapter 3

See also: Bundle, Block builder, TEE, Harmonic, Validator.

BAM Node

Short: The operator role under Jito's Block Assembly Marketplace — a TEE-attested sequencing service that runs inside AMD SEV-SNP enclaves and produces cryptographic attestations of the ordering decision before the validator executes the block.

Long: BAM splits the block-construction function into sequencing (delegated to the BAM Node) and execution (retained by the validator), with the BAM Node's ordering logic sealed inside hardware enclaves so that selective content-based extraction is provably difficult without the BAM Node colluding. The four launch BAM Node operators — Helius, Triton One, SOL Strategies, and Figment — are all also institutional validators or RPC providers, which means the operator set is concentrated in firms with substantial existing infrastructure businesses. By late February 2026, approximately 25% of Solana stake was connected through BAM Nodes. The open-source target is mid-Q2 2026, a prerequisite for onboarding third-party operators at scale because the TEE attestation is verifiable only against published code.

TradFi analogue: A regulated matching-engine operator with hardware-attested audit logs — the role NASDAQ's matching engine played in the 1990s, except with cryptographic rather than legal-and-regulatory attestation.

First appears: Chapter 6 (named); full role developed in Chapter 8

See also: BAM, TEE, Validator, Block builder.

Block

Short: The unit of transaction batching on most chains; transactions are grouped into blocks and validated together.

Long: Each chain has its own block-production schedule. Solana produces a new block roughly every 400 milliseconds; Ethereum L1 produces one every 12 seconds; Hyperliquid produces them on a sub-second cadence inside HyperBFT consensus. Inclusion in a block is the on-chain analogue of "the trade has matched at the venue" — the moment the trade is recorded. Finality, the moment the block becomes irrevocable, takes longer than inclusion on most chains.

TradFi analogue: The matched-trade record at an exchange before legal settlement at the clearinghouse — the trade has happened, but the cash-and-asset transfer is still pending.

First appears: Chapter 1

See also: Settlement, Finality, Validator.

Block builder

Short: The actor who assembles the set of transactions in a block, orders them, and submits the block to the validator (or proposer) for inclusion.

Long: On Ethereum since the Merge, block-building is typically outsourced to a specialist firm — Titan, beaverbuild, rsync — whose software is designed to maximize the dollar value of the block before it is sealed, then to compensate the validator with a share of that value. On Solana, the equivalent role is performed by Jito's block engine plus the leader-validator's own scheduler. The block builder is the actor that searchers most need access to: the moat of a sophisticated 2026-era searcher is the relationship with one or more builders.

TradFi analogue: An NYSE specialist's discretion over order matching, plus a market-on-close auction operator, combined.

First appears: Chapter 4 (cameo); full treatment in Chapter 5.

See also: Validator, MEV, Searcher.

Builder-direct relationship

Short: A private commercial contract between a trading firm (or a bot operator) and a specific Ethereum block builder, under which the firm routes its transactions exclusively to that builder in exchange for a margin uplift or rebate.

Long: The canonical 2026 case is Banana Gun / Titan: the bot routes all its users' transactions to Titan via private RPC, and Titan reportedly retains approximately 17.75% margin under the deal versus Beaverbuild's pre-migration ~9% and Flashbots' near-zero historical margin. Wu et al.'s peer-reviewed analysis identified 75 such arrangements accounting for approximately 71% of trading-related Ethereum builder revenue, with Banana Gun-Titan and Maestro-Beaverbuild as the two largest by estimated proposer-loss share. Builder-direct relationships are the Ethereum-side structural analogue of US equity payment-for-order-flow — same concentration shape, no regulatory disclosure regime.

TradFi analogue: The bilateral commercial arrangement between a retail broker (e.g., Robinhood) and a wholesaler (e.g., Citadel Securities), under which retail flow is routed in exchange for a per-share rebate — except that no Rule 606 disclosure equivalent exists on-chain.

First appears: Chapter 7

See also: Exclusive order flow (on-chain), Block builder, Private order flow, Payment for order flow (PFOF).

Bundle (Jito-style)

Short: A group of transactions that the protocol must execute together, in the order specified, or not at all.

Long: Bundles are the basic primitive of MEV infrastructure. A searcher submits a frontrun, the victim's swap, and a backrun as a single bundle; if any of the three would fail, the entire bundle is reverted and the searcher does not get a partial fill. On Solana, bundles are submitted through Jito's block engine, which runs an auction over them. On Ethereum, the equivalent is submitted through a builder (often via Flashbots' relay) directly into a block. Bundles are how searchers manage the risk of partial execution that would otherwise make multi-leg strategies untenable.

TradFi analogue: An all-or-none order, or a multi-leg options strategy executed as a single ticket.

First appears: Chapter 4 (cameo); full treatment in Chapter 6.

See also: Searcher, Block builder, MEV.

Central scheduler (Solana)

Short: Agave's default block-production scheduler — the component that decides which transactions enter a block and in what order when the validator is leading the slot.

Long: The scheduler ranks transactions by what the validator earns from each one relative to how much of the block's capacity that transaction consumes. In practice, transactions paying higher priority fees relative to their on-chain footprint win the auction for blockspace. The Agave scheduler has gone through two generations of defaults since 2024: the original prio-graph central scheduler (introduced as opt-in in v1.18, made default in v2.0), and the current central-scheduler-greedy (introduced as opt-in in v2.2, made default in v2.3 in mid-2025). The greedy variant skips the dependency-graph construction of the prio-graph version in favour of simpler, faster selection logic. v3.1 (early 2026) fixed a banking-worker / Proof-of-History synchronisation bug and now lets external schedulers (Rakurai, Frankendancer's pack tile) plug in via an IPC binding at <ledger-path>/scheduler_bindings.ipc rather than forking the banking stage. The current Agave default is --block-production-method central-scheduler-greedy.

First appears: Chapter 5

See also: Agave, Pack tile (Firedancer), Validator client.

CEX–DEX arbitrage

Short: An arbitrage strategy in which a searcher trades against a price gap between a centralized exchange (like Binance) and an on-chain pool (like Uniswap V3).

Long: CEX–DEX is the most consolidated and least measurable form of on-chain extraction. A peer-reviewed study published in 2025 measured $233.8 million extracted by nineteen Ethereum CEX–DEX firms over nineteen months, with the top three capturing 90% of volume. The activity is "dark" because its inputs — centralized-exchange prices and the searcher's inventory at those exchanges — are not on-chain; only the on-chain leg of each trade is visible. Passive liquidity providers in the on-chain pool are the ultimate counterparty and pay the searcher's profit through stale-price fills.

TradFi analogue: Inter-exchange latency arbitrage between two equity venues.

First appears: Chapter 4

See also: Arbitrage, Searcher, MEV.

Chronic loser (on-chain)

Short: A structurally disadvantaged on-chain trading participant who continues to participate despite the documented disadvantage. The book identifies four chronic-loser categories: retail traders unaware of toxic flow, passive LPs in informed-flow pools, slow market makers, and validators without infrastructure relationships.

Long: The chronic-loser framing is the book's editorial contribution to the question "why does the structural disadvantage persist?" The four categories differ in both loss magnitude (small per-trade × high count for retail; continuous LVR drag for LPs; displacement-not-loss-per-trade for slow MMs; opportunity cost for validators) and in the participant's informed-consent relationship to the structure (see Consent gap). The structural reason each category continues to participate is not a single mechanism; it is a category-specific stack of switching costs, sunk costs, alternative-venue unavailability, and informational gaps. The book argues that the four loser categories are not symmetric, that "just leave the table" is a behavioural injunction that ignores the structural costs of leaving for three of the four, and that the consent gap is what makes the retail loser category structurally different from the other three.

First appears: Chapter 11

See also: Consent gap, Passive LP, Loss-Versus-Rebalancing, Validator-quintile gap, Toxic flow.

Short: The asymmetric informational position of the four chronic-loser categories on-chain. The retail loser category includes a meaningful share of participants who do not know the structural mechanism producing their loss; the other three categories include participants who know the structure but lack the alternatives to exit.

Long: The consent gap is the chapter's clinical contribution to the book's adversarial thesis. The four loser categories differ in their informed-consent relationship to the structural disadvantage: retail traders bear small per-trade losses across high transaction counts with a UX that abstracts away the per-layer dollar trace; passive LPs bear continuous LVR drag where the LVR research has existed in peer-reviewed form since 2022 but is not surfaced at the deposit interface; slow market makers face displacement they understand at firm level; validators without infrastructure relationships bear opportunity costs they can measure cross-validator. The structural argument: the consent gap is largest for retail, and the regulatory frameworks documented in the book (the EU PFOF ban, the SEC's withdrawal of the Order Competition Rule, the ETF wrappers) do not address the consent gap because they regulate the asset-exposure layer rather than the trade-execution layer. The chapter does not argue that the structures are illegitimate — three of the four loser categories include participants who knew the structure when they entered it; the chapter argues that the asymmetry of informed consent across the four categories is the most important structural fact about who-loses-and-why on-chain in 2026.

First appears: Chapter 11

See also: Chronic loser (on-chain), Payment for order flow (PFOF), Private order flow, Public order flow.

CLOB (central limit order book)

Short: A trading venue where bids and asks are matched explicitly, in the order they arrive, with the spread between the best bid and the best ask defining the inside market.

Long: A CLOB is the dominant design on traditional exchanges (NYSE, NASDAQ, CME) and on centralised crypto exchanges (Binance, Coinbase). On-chain, the design failed for nearly a decade because every action — placing, cancelling, modifying an order — was a transaction the chain had to validate, and transactions were too slow and expensive for active market making. Hyperliquid is the first on-chain CLOB to find product-market fit at scale, using HyperBFT consensus with ~70-millisecond block intervals, no public mempool, and native matching inside consensus rather than via smart contract calls — though primarily for perpetuals, not spot.

TradFi analogue: The NYSE, NASDAQ, CME, Eurex — any electronic exchange.

First appears: Chapter 2

See also: AMM, Market maker, Spread, HLP.

Colocation (Solana validator context)

Short: The practice of physically placing a Solana validator in an Equinix-tier data centre near a Jito block engine — Equinix FR5 (Frankfurt) and Equinix NY5 (New York) being the two anchor sites — to achieve sub-50ms block-engine-to-validator latency.

Long: Standard 2026 Solana validator colocation runs bare-metal AMD EPYC 9354P plus 512GB RAM plus three NVMe drives (one-time CapEx approximately $8,000–$15,000), with rack and power at approximately $200–$500 per month in standard Equinix tier; premium colocation runs $400–$800 per month. The "next to the TPU port" pattern — a non-voting validator co-located in the same rack as a paid RPC, forwarding to the current leader via QUIC over LAN — produces a 5–10× latency reduction versus remote cloud routing. Solana stake concentrates in six cities (Frankfurt 19%, Amsterdam 16%, London 12%, Vilnius 6%, Tokyo 4%, Ashburn 4%) and Jito's mainnet block engines run in five of those six. The geographic concentration is what produces the chapter's signature finding: the +101% Harmonic Performance validator is, in practice, the validator co-located in FR5 or AM3 on a Jito-derivative client with a BAM-node relationship.

TradFi analogue: NYSE colocation services at Mahwah, NJ — the same product, with the same equilibrium consequences (latency-sensitive trading concentrates in the racks closest to the matching engine).

First appears: Chapter 8

See also: BAM Node, Jito tip, Validator, Validator client.

Constellation

Short: Anza's multi-leader proposal for Solana. Replaces the single-leader-per-slot structure with about sixteen concurrent proposers rotating every ~1.6 seconds, on a 50-millisecond economic tick. Designed as a preprocessor to Alpenglow; targeted for Q3 2026 mainnet activation. The architectural answer to the access-vs-operational revenue gap that Chapters 5, 7, and 8 documented.

Long: Constellation was unveiled by Anza CEO Brennan Watt at Solana Accelerate in Miami on 5 May 2026, designed by Max Resnick (Anza's lead economist) and Quint Kniep. It builds on the diagnosis Resnick and Anatoly Yakovenko published in the Path to Decentralized NASDAQ essay in May 2025: that a single leader controlling ordering inside each slot is the root cause of the +101% Harmonic Performance per-block revenue gap and of the failure of the Solana spot CLOB experiment. Constellation's mechanism: ~16 proposers assemble transactions into "pslices" on a 50ms cycle; 256 attesters sign attestations documenting what each proposer produced; the designated leader for a block can only build a valid block by including pslices attested by at least 40% of the attester set. A leader who censors a fee-competitive transaction produces a block that fails consensus. The proposal introduces a two-tier fee market — inclusion fees to validators, ordering fees burned — and a "cancels before takes" priority that gives market makers a chance to pull stale quotes before opportunistic takers can lift them. Constellation is contingent on Alpenglow shipping first; the published two-step roadmap is Alpenglow → Constellation, both targeted for Q3 2026.

TradFi analogue: The closest analogue is the post-Reg-NMS US equity market structure, where multiple exchanges quote prices simultaneously and no single venue has order-routing monopoly. Constellation imports a version of that architecture into Solana's per-slot ordering layer.

First appears: Chapter 8 (cameo in the verdict); full forward-looking treatment in Chapter 12.

See also: Alpenglow, Validator-quintile gap, Harmonic, Exclusive order flow (on-chain).

Cross-chain MEV

Short: Extractive value capture across multiple chains (typically L1-to-L2 or L2-to-L2 pairs), measured at approximately $10 million of searcher revenue across nine blockchains over twelve months by Maire et al. (2024). The L2-fragmentation equivalent of the L1 builder-direct contract — same few-firm concentration shape, one architectural level out.

Long: Maire, Sviridov, Capponi, and Wattenhofer (Flashbots + TU Munich; ACM SIGMETRICS 2024 / arXiv 2501.17335) identified 242,535 cross-chain arbitrages between September 2023 and August 2024, totaling approximately $868.64 million of trading volume and approximately $10.05 million of searcher revenue. Composition: 58.35% L1-to-L2 pairs (driven by native bridge availability); 35.67% L2-to-L2 pairs (relying on third-party bridges or pre-positioned inventory). Most trades (66.96%) use pre-positioned inventory and settle in ~9 seconds; bridge-based arbitrages take ~242 seconds. The structural finding: the top five addresses execute more than 50% of all cross-chain arbitrages, and a single address captures approximately 40% of daily post-Dencun volume — a concentration ratio comparable to the L1 builder oligopoly. Atomicity is the structural difference between L1 MEV (atomic, single-block, deterministic) and cross-chain MEV (non-atomic, multi-second to multi-minute, probabilistic). Shared sequencers (Espresso, Astria) and intent-based architectures (UniswapX cross-chain, CoW Swap) attempt to re-introduce atomicity at the sequencer or solver layer; neither has replaced bridge-based execution at scale as of May 2026.

First appears: Chapter 10

See also: MEV, Searcher, L2 sequencer, Bridge.

EigenLayer

Short: Ethereum's restaking protocol. Approximately $18 billion of restaked ETH across approximately 4.6 million ETH at approximately 1,900 active operators in mid-2026. Approximately 94% market share of the restaking ecosystem.

Long: EigenLayer's structural innovation is to allow Ethereum validators to "restake" their staked ETH — pledging additional slashing exposure beyond the base PoS slashing — to secure arbitrary Actively Validated Services (AVSs). Each AVS pays operators in its own token or in ETH in exchange for the operators' restaked stake committing to the AVS's misbehaviour conditions. Mainnet slashing went live 17 April 2025, making EigenLayer "complete" in the founders' published framing. ELIP-006 Redistributable Slashing introduces operator-set-specific governance of slashed-fund destination (burn vs redistribute). Named AVSs include EigenDA (data availability; the largest single AVS by stake), AltLayer MACH (fast finality for OP Mainnet and Arbitrum One), Hyperlane (cross-chain messaging), and Witness Chain (DePIN proof-of-location). The structural argument the chapter develops: EigenLayer is the third architectural layer of Ethereum's PBS + L2 + restaking trifecta — and the same operator concentration that appears at the L1 staking layer (Lido, Coinbase Cloud, P2P.org, RockX) appears at the restaking layer.

TradFi analogue: A re-hypothecation market for collateral — same dollar of margin securing multiple positions, with cross-default exposure if any position fails.

First appears: Chapter 10

See also: AVS, Slashing, Validator, Lido.

ePBS (enshrined PBS)

Short: The proposed Ethereum upgrade (EIP-7732) to bring proposer-builder separation into the protocol layer rather than the out-of-protocol MEV-Boost relay layer.

Long: ePBS was selected as the headliner for Ethereum's Glamsterdam upgrade, originally targeted at the first half of 2026. As of the Ethereum Foundation's April 2026 Checkpoint #9 update, Glamsterdam has not shipped: implementation has proved trickier than anticipated, with partial blocks and two-party consensus coordination as documented bottlenecks. No firm replacement target has been published. The structural argument: when PBS lives inside the protocol, the proposer's choice of builder becomes a protocol-defined market rather than an out-of-protocol auction, and the structural concentration around Titan / BuilderNet / Quasar could in principle be replaced by a wider competitive surface.

First appears: Chapter 5

See also: PBS, Block builder, MEV-Boost, Proposer.

Exclusive order flow (on-chain)

Short: A private commercial contract under which one party — a builder, RPC provider, wallet, or validator — receives a specific slice of transaction flow on preferential terms not available to other actors, bypassing the chain's nominally-permissionless competitive surface.

Long: On-chain exclusive order flow takes three structural shapes in 2026. Direct exclusive arrangements between bot operators and builders (Banana Gun / Titan; Maestro / Beaverbuild) dominate the Ethereum builder layer — Pahari and Canidio's November 2025 measurement found exclusive transactions account for 77.2%–84% of total fees paid in winning Ethereum blocks. Pipeline rebate relationships between RPC providers and wallets (Helius's 50/50 split with Phantom, Backpack, Solflare and others) dominate the Solana side. Validator-side fragmented surfaces — the post-June-2024 grey-market private mempools whose status the Solana Foundation describes as "ongoing enforcement" — operate beneath both. The structurally important fact: none of these arrangements are formally disclosed to any regulator, and the chains' founding documents did not anticipate the form.

TradFi analogue: Payment for order flow in US equity markets, banned in the EU effective 30 June 2026 and retained in the US after the SEC's 12 June 2025 withdrawal of the Order Competition Rule — same concentration shape, opposite regulatory direction across the two jurisdictions, and no regulator at all for the on-chain version.

First appears: Chapter 7

See also: Builder-direct relationship, Pipeline rebate, Validator-quintile gap, Private order flow, Public order flow.

ERC-7683 (cross-chain intent standard)

Short: The cross-chain intent standard introduced by Uniswap Labs and Across (ratified early 2025). By April 2026 approximately 88% of Across's volume was settled via ERC-7683 orders; implemented by Across, UniswapX, CoW Protocol, and Eco.

Long: ERC-7683 standardises the message format and settlement semantics for cross-chain intents — the user signs an order that specifies the destination chain and the acceptable settlement parameters, and a solver / relayer competes to fulfil it across the cross-chain bridge layer. Wallet support includes Safe, Argent, Rabby, and MetaMask; major L2 support extends across Arbitrum, Optimism, Polygon, and Base. The Open Intents Framework — a cross-protocol consortium working on ERC-7683-adjacent tooling — launched February 2025 with 30+ teams. Across migrated its production solver network to ERC-7683 in Q3 2025; approximately 88% of its volume now uses ERC-7683 orders by April 2026. The structural argument the book develops: ERC-7683 adoption is the cleanest single piece of evidence that the cross-chain intent pattern is winning the architectural argument; the relayer-concentration data (Across top-5 relayers handle ~70% of volume) is the cleanest single piece of evidence that the firms operating that pattern are concentrating.

First appears: Chapter 12

See also: Intent, Cross-chain MEV, L2 sequencer.

Finality

Short: The moment a trade or transaction becomes irrevocable — when no one can roll it back.

Long: On a centralised exchange, finality from the user's perspective is instant: the trade appears in the account and is treated as done. The legal settlement at a clearinghouse (T+1 for US equities since May 2024) may complete later, but the user does not see it. On most chains, finality is probabilistic and takes a chain-specific amount of time: Solana finalises blocks in approximately 12.8 seconds (with the Alpenglow upgrade targeting ~150 milliseconds once it ships); Ethereum L1 finalises in approximately 12.8 minutes via Casper FFG; Hyperliquid produces consensus-level finality in roughly 200 milliseconds. Ethereum L2s offer fast sequencer confirmation but require a seven-day fraud-proof window for canonical settlement back to L1.

TradFi analogue: The completion of a clearinghouse settlement cycle (T+1 for US equities post-2024, T+2 prior).

First appears: Chapter 1

See also: Settlement, Block, Validator.

Firedancer

Short: Jump Crypto's full C/C++ rewrite of the Solana validator client. Mainnet adoption was approximately 2% of stake in March 2026.

Long: Firedancer is intended to be the long-term performance upgrade path for Solana validators — a complete rewrite of the entire validator stack in C/C++, rather than the Rust-based Agave reference client. Adoption on mainnet has been slower than expected; most production-focused operators have preferred the Frankendancer hybrid (which uses Firedancer's networking layer but Agave's consensus and execution) for risk reasons. The pack tile in both Firedancer and Frankendancer supports user-selectable scheduler modes — perf and balanced — that materially affect block-construction behaviour and revenue capture.

First appears: Chapter 5

See also: Frankendancer, Agave, Pack tile (Firedancer), Validator client.

Flashbots Protect

Short: The canonical Ethereum private-routing RPC service; users change their wallet's RPC URL to Flashbots' endpoint and their transactions are delivered as private bundles directly to block builders, bypassing the public mempool.

Long: Flashbots Protect has served approximately 2.1 million Ethereum accounts and shielded approximately $43 billion in cumulative DEX volume since launch; third-party benchmarks place its share at roughly 7% of Ethereum's monthly transaction count. The mechanism: a transaction submitted through Protect is bundled by Flashbots and forwarded to block builders without ever entering the public mempool; if the transaction creates a backrunning opportunity that a searcher captures, a share of the searcher's payment is refunded back to the original submitter. Protect is the simplest of the major Ethereum private-routing options — a one-time wallet configuration change with no further user action required.

First appears: Chapter 3

See also: Private order flow, MEV-Blocker, MEV-Share, MEV-Boost, Bundle.

Flashnet

Short: Flashbots' February 2026 design for a network-anonymized mempool — a submission-layer privacy primitive that hides who submitted a transaction even from the recipient builder.

Long: Flashnet is positioned as the submission-layer half of a two-part answer to the visibility-versus-censorship trade-off, complementary to FOCIL (which addresses censorship resistance at the consensus layer). The argument: privacy at submission plus enforced inclusion at consensus addresses both extraction and censorship risk in different layers. Flashnet is in development and is not live on mainnet as of mid-2026.

First appears: Chapter 3

See also: FOCIL, Private order flow, MEV-Boost.

Frankendancer

Short: Jump Crypto's hybrid Solana validator client — replaces Agave's networking layer with Firedancer's lower-level C/C++ networking stack while keeping Agave's consensus and execution.

Long: Frankendancer is the production-focused intermediate step between Agave (stable, Rust-based) and pure Firedancer (full C/C++ rewrite, slower mainnet adoption). The hybrid design captures the networking-layer performance gains that matter most for Solana validator economics — the bandwidth bottleneck is increasingly the constraint — while preserving the stability of Agave's consensus and execution layers. Frankendancer ran approximately 21% of Solana stake in October 2025 and approximately 12% by March 2026; the relative decline reflects validators migrating onward to JitoBAM and Harmonic rather than away from the Jump stack. Like full Firedancer, Frankendancer supports user-selectable scheduler modes via its pack tile.

First appears: Chapter 5

See also: Firedancer, Agave, Pack tile (Firedancer), Validator client.

Frontrun

Short: A transaction the searcher places so it executes immediately before a target transaction whose contents the searcher has already seen.

Long: Frontrunning requires that the target be visible somewhere before it executes — most often in a public mempool, sometimes via a private flow leak. The searcher reads the target, simulates its market impact, and submits a transaction designed to execute first and benefit from that impact. Frontrunning is the prerequisite for sandwich attacks and is the single technique most often cited when researchers talk about MEV's harm to retail traders.

TradFi analogue: An HFT desk acting on order-book information ahead of a large client order — a practice regulated in equities, structurally legal in most on-chain settings.

First appears: Chapter 4

See also: Backrun, Sandwich attack, Mempool.

Gas

Short: The unit of compute work a transaction consumes, priced by the chain in its native token; the user pays for it whether the transaction succeeds or fails.

Long: Gas exists because every transaction consumes some amount of validator computation, and the chain has to price that work to prevent denial-of-service attacks. Ethereum measures gas in units of work multiplied by a per-unit price (gwei, which is 10⁻⁹ of an ETH); Solana uses a fixed base fee per signature (5,000 lamports) plus optional priority fees. The dollar cost varies wildly across chains: in May 2026, a typical Uniswap V3 swap on Ethereum L1 cost roughly $0.25–$0.39, against an average of about $0.017 per transaction on Solana and $0.01–$0.05 on Base.

TradFi analogue: No clean analogue. The closest is an exchange's per-trade clearing fee plus the variable cost of running its matching engine — except on-chain, the cost is paid by the user directly rather than absorbed into the venue's spread.

First appears: Chapter 1

See also: Block, Priority fee, Transaction.

Harmonic

Short: A Solana validator-selection-layer protocol; an open block-building marketplace where validators continuously receive candidate blocks from multiple independent builders and select the most valuable one per slot, with three off-the-shelf scheduling strategies plus a custom-build engagement.

Long: Harmonic is not a fork of any single Solana client and not a single scheduler. It is an open block-building system that aggregates candidate blocks from multiple builders — Jito's Block Engine, Temporal, JitoBAM, Paladin — and lets the validator select per slot. Harmonic ships three off-the-shelf scheduling strategies, each operating under three architectural commitments (no sandwiching, no censorship, no malicious MEV): 50ms FBA with Priority Fee Ordering (SFDP-compliant; ~120–130% of median block rewards); FIFO (SFDP-compliant; ~90–100% of median block rewards); and MREV (Maximum Revenue Strategy; not SFDP-compliant; ~130–150% of median block rewards). Harmonic's own documentation describes MREV as "not the network-aligned choice." A Custom Scheduler option is available as a partnership-style engagement. As of March 2026, Harmonic ran approximately 17% of Solana network stake; Nasdaq-listed DeFi Development Corp adopted it in December 2025.

First appears: Chapter 3 (cameo); full strategy treatment in Chapter 5.

See also: BAM, Block builder, Validator, Validator client.

Health factor

Short: A real-time numeric ratio used by Aave-style lending protocols to determine whether a borrower's position is solvent; values below 1.0 are liquidatable.

Long: The health factor is computed continuously from the borrower's collateral value, the borrower's debt, and protocol-set liquidation thresholds for each asset. As long as it stays above 1.0, the borrower is safe; once it crosses below, any third party can liquidate a portion of the position and claim the protocol's liquidation bonus. Aave V4 introduced a refinement in which liquidation severity scales with how far below 1.0 the health factor falls, replacing the older binary regime.

TradFi analogue: A margin-call trigger at a prime broker, plus the bid-to-cover ratio at the resulting forced unwind.

First appears: Chapter 4

See also: Liquidation, Liquidation bonus.

HIP-3 (Hyperliquid builder-deployed perpetuals)

Short: The Hyperliquid Improvement Proposal that enables permissioned-by-stake deployment of new perpetual markets on top of HyperCore. Launched on mainnet 13 October 2025 with a 500,000 HYPE (~$25M) staking requirement.

Long: A HIP-3 deployer who posts the 500K HYPE stake can permissionlessly launch any number of perpetual markets with full control over oracle selection, contract specifications, maximum leverage, margin ratios, and open-interest caps. From the user's perspective, fees on HIP-3 markets are 2× the standard validator-operated perp fees; the deployer receives 50% of the fee, the protocol the other 50%. The deployer's stake is slashable: up to 100% for invalid state transitions, 50% for brief downtime, 20% for network degradation; slashed tokens are burned rather than redistributed. As of May 2026, aggregate 30-day HIP-3 volume reached approximately $2.82 billion with 20,348 unique traders, and HIP-3 markets represent more than 35% of all Hyperliquid trading volume. TradeXYZ holds >90% of HIP-3 open interest with $12.7B cumulative volume; other named operators include Ventuals, Felix Protocol, Markets by Kinetiq, HyENA, Dreamcash, Paragon.

TradFi analogue: A regulated futures-exchange's permissioning of new product listings — anyone can propose a new contract, but the exchange's listing-committee approval (plus a capital posting requirement) gates the operator set to institutional teams.

First appears: Chapter 5 (cameo); full treatment in Chapter 9

See also: HyperCore, HyperEVM, HLP, Assistance Fund (Hyperliquid).

HIP-4 (Hyperliquid outcome markets)

Short: The Hyperliquid Improvement Proposal that extends the HIP-3 builder-deployed-perpetuals pattern to outcome markets (prediction-market-style binary or multi-outcome contracts). Launched on mainnet 2 May 2026 with a 1 million HYPE Phase 2 deployer stake — 2× the HIP-3 threshold.

Long: HIP-4 was launched as Phase 1 with a single debut market (Outcomexyz BTC daily price prediction); Phase 2 is the permissionless-by-stake deployment mode requiring 1M HYPE (approximately $50M at $50/HYPE) for any operator to deploy outcome markets independently. The 2× threshold relative to HIP-3 makes the deployer-stake moat steeper; the structural argument the book develops in Chapter 12 is that HIP-4 represents the appchain pattern continuing to relocate value capture into deployer-stake-moated operator economics rather than addressing the operator-concentration question. The Hyperliquid protocol receives 50% of fees (mirroring HIP-3) and the deployer receives 50%.

First appears: Chapter 12

See also: HIP-3 (Hyperliquid builder-deployed perpetuals), HLP, HyperCore.

HLP (Hyperliquid liquidity provider vault)

Short: Hyperliquid's protocol-owned liquidity vault that quotes against incoming flow and absorbs liquidation surplus, with deposits open to any participant.

Long: HLP functions as an AMM-of-MMs inside Hyperliquid's protocol; depositors fund it, and it quotes against trades that other market makers haven't already filled. HLP also inherits the surplus or shortfall from forced liquidations on the venue. Its disclosed lifetime CAGR through early 2025 was approximately 42% (trailing-twelve-month ~22%); TVL was approximately $391 million as of May 2026. The March 2025 JELLY exploit, in which an attacker forced HLP to inherit a $4.1 million short and then pumped the underlying spot price by 400%, demonstrated the vault's exposure to targeted adversarial flow and triggered tighter leverage rules.

TradFi analogue: A primary dealer's house book, except that anyone can deposit into the book and earn its returns.

First appears: Chapter 2 (cameo); full treatment in Chapter 9.

See also: CLOB, Liquidity provider, Market maker.

HyperBFT

Short: Hyperliquid's consensus protocol — a pipelined HotStuff-derivative BFT design with up to 1/3-malicious tolerance, ~70ms block intervals, and end-to-end median finality of ~200ms (p99 ~900ms).

Long: HyperBFT is the consensus layer that makes Hyperliquid's matching-engine-inside-consensus architecture viable for active market making. The combination of (a) no mempool visibility — order placement and matching collapse into a single consensus operation — and (b) ~70ms block intervals — a market maker's cancel reaches the matching engine before opportunistic takers can lift the stale quote — closes the visibility-and-latency surface that doomed every Solana spot CLOB attempt (Phoenix, OpenBook). HyperBFT runs both HyperCore (perpetuals, spot, order books, liquidations) and HyperEVM (general-purpose EVM-compatible programmability layer) in the same consensus and the same blocks while maintaining independent state.

TradFi analogue: Closest to a centralised matching engine inside an exchange's primary data centre — except that the matching code is the protocol itself, the participants are pseudonymous, and the consensus rules are publicly verifiable.

First appears: Chapter 2 (cameo); full treatment in Chapter 9

See also: HyperCore, HyperEVM, HLP, CLOB.

HyperCore

Short: The perpetuals + spot + order books + liquidations component of the Hyperliquid protocol. Runs inside HyperBFT consensus with matching as protocol code rather than smart-contract code.

Long: HyperCore is where the chain's roughly $619 billion of Q1 2026 perpetual-trading volume is matched. Published throughput: approximately 200,000 orders per second sustained. The chain's HLP vault, the validator-operated perpetual markets, and the spot order books all live inside HyperCore. HIP-3 builder-deployed perpetuals are HIP-3-stake-gated extensions of HyperCore's product surface. HyperCore and HyperEVM share consensus and blocks via HyperBFT but maintain independent state; canonical wrapped HYPE (WHYPE) and natively-minted USDC connect the two layers.

First appears: Chapter 2 (cameo); full treatment in Chapter 9

See also: HyperBFT, HyperEVM, HLP, HIP-3.

HyperEVM

Short: Hyperliquid's general-purpose EVM-compatible programmability layer; live on mainnet since 18 February 2025. Approximately $1.9 billion TVL as of mid-2026.

Long: HyperEVM is the chain's answer to "how do we add general-purpose programmability without compromising the HyperCore matching engine?" EVM-compatible smart contracts run inside HyperBFT consensus but with independent state from HyperCore; canonical wrapped HYPE (WHYPE) connects the two layers and natively-minted USDC was linked across both via Circle's CCTP in 2026. As of mid-2026, HyperEVM hosts approximately $1.9 billion TVL with 60,000+ weekly active addresses and 100+ dApps. Named ecosystem firms: Kinetiq (kHYPE liquid staking ~$639M TVL, peak ~$2.28B); Felix Protocol (HyperEVM lending + feUSD stablecoin, ~$300M TVL); LayerZero / Stargate (bridge layer); Bitget Wallet (UX integration). The structural framing: HyperEVM is genuinely smaller than Solana DeFi (~1/50th) and Ethereum (~1/30th) by TVL — the chain has not optimised to be a general-purpose platform. It has optimised to be a derivatives venue with sufficient programmability to allow secondary protocols without compromising matching-engine performance.

First appears: Chapter 9

See also: HyperBFT, HyperCore, HLP.

Impermanent loss (IL)

Short: The difference between what an LP would have held by simply holding the underlying assets and what they actually hold in the pool, as the assets' relative prices move.

Long: IL is a measurement, not a fee. It captures the gap between an LP's pool position and the hypothetical "just held" position. The term is "impermanent" because the loss is realised only when the LP withdraws — if the relative prices return to their starting point, IL goes back to zero. The Loss-Versus-Rebalancing framework is a sharper successor that compares the LP not to HODL but to an arbitrageur continuously rebalancing the same position against a centralised exchange, and gives a worse (more honest) answer.

First appears: Chapter 2

See also: Loss-Versus-Rebalancing, AMM, Liquidity provider.

Infrastructure firm (B2B trading infrastructure)

Short: A company whose business is to operate one or more layers of the on-chain trading infrastructure stack — RPC, OFA, builder, relay, block engine — and capture a share of the trading flow that passes through.

Long: The infrastructure layer emerged as a distinct firm category between approximately 2020 and 2024 in response to operational gaps on permissionless chains: running RPC at scale is expensive; constructing optimal blocks under MEV-Boost is a hard engineering problem; refunding MEV requires sophisticated routing logic and trust relationships. Named firms in 2026 include Helius, Jito Labs, Titan / Gattaca, BuilderNet, bloXroute, Flashbots, Triton One, Temporal, Harmonic, Anza, Jump Crypto, Astralane, Kolibrio, Paladin, Aestus, the Special Mechanisms Group at Consensys, and CoW DAO. The structural pattern: each function has a small number of operators with measurable share; operators often span multiple layers; concentration runs across layers as much as within them. The aggregate take across both chains in 2026 is in the mid-hundreds-of-millions to low-billions of US dollars annual run rate, though no single published source aggregates the number.

TradFi analogue: The B2B infrastructure behind every modern equity exchange — DTCC for clearing, NYSE's matching engine, the wholesalers (Citadel Securities, Virtu) that internalise retail flow, the data vendors (Refinitiv, Bloomberg). The on-chain ecosystem emerged in roughly four years (2022–2026) versus four decades in equities.

First appears: Chapter 6

See also: RPC provider, Order Flow Auction (OFA), Block builder, Relay (Ethereum), Searcher, Validator.

Inclusion list / FOCIL (EIP-7805)

Short: Ethereum's proposed fork-choice-enforced inclusion-list mechanism: a 16-validator committee per slot publishes a list of transactions that the block proposer must include or have its block rejected.

Long: FOCIL (Fork-Choice enforced Inclusion Lists) is specified in EIP-7805 and addresses the censorship-resistance side of the visibility-versus-extraction trade-off. Under FOCIL, attesters vote for a proposer's block only if it includes all transactions on the stored inclusion lists; conditional omission is permitted only when gas or block space is unavailable. The design's argument: even if 100% of block-building capacity were sold to a single hostile builder or proposer, transactions on any inclusion list would still have to be included. FOCIL has been selected as the consensus-layer headliner for Hegotá, the post-Glamsterdam Ethereum upgrade, framed as a late-2026 cleanup-and-hardening fork. No firm mainnet date has been published.

First appears: Chapter 3

See also: Flashnet, Block builder, MEV-Boost.

Intent

Short: The trader's stated desire to make a trade, before the trade is matched or executed.

Long: Intent is the first of the three phases every trade passes through. On a centralised exchange, intent is the moment the user clicks Buy or Sell in the app; the exchange knows the user wants to trade, but no one outside the exchange does. On most chains, intent is the moment the user signs and broadcasts a transaction — at which point, depending on the chain, the intent may become visible to anyone running a node or paying for an RPC service. Visibility of intent is the surface that makes most extractive on-chain trading techniques possible; Chapter 3 develops it in detail.

First appears: Chapter 1

See also: Settlement, Finality, Mempool, Wallet.

JIP-24

Short: The 2025 Jito DAO proposal that routes 100% of Jito Block Engine and BAM tip fees to the Jito DAO treasury in perpetuity, replacing the prior 3%/3% split between Jito Labs (the company) and the DAO.

Long: Pre-JIP-24, the 6% Jito Block Engine and BAM tip take was split equally between Jito Labs and the Jito DAO. Post-JIP-24, the entire 6% routes to the Jito DAO treasury — Jito Labs no longer takes a direct fee cut from these flows. Jito Labs the company continues to operate the products (Block Engine, BAM, JitoSOL, the Jito-Solana client) and to earn JitoSOL's 4% management fee on staking rewards, but the Block Engine + BAM revenue stream has been routed entirely through the DAO. The structural argument the chapter makes from this: a major infrastructure firm voluntarily restructured its revenue to route fees through a community-governed DAO — the cleanest case in the book of this pattern.

First appears: Chapter 6

See also: BAM, Jito tip, Infrastructure firm.

Jito tip

Short: On Solana, a payment a transaction sender can attach to be routed through Jito's block engine, increasing the chance that the transaction is included in the next block.

Long: Jito operates a block engine and an auction marketplace that competes for the leader-validator's attention; transactions that include a tip flow through this path and are ordered in part by the tip's size. Mechanically distinct from the network's base priority fee, the tip is paid not to the network but to the validator that includes the bundle. For sophisticated searchers, tips are the dominant inclusion signal on Solana: in the Vpe sample studied by Helius (Dec 2024 – Jan 2025), tips paid to validators accounted for 34% of the program's gross sandwich revenue.

TradFi analogue: Payment for order flow paid by a market maker to a broker, except that on-chain the payment is per-transaction and visible.

First appears: Chapter 4

See also: Priority fee, Block builder, MEV.

Jupiter Beam

Short: Jupiter's transaction-landing engine — the default protected-routing surface inside Jupiter Ultra, routing signed Solana transactions to Jito as private bundles rather than over the chain's public submission paths. Opened as a public Transaction Submission API in April 2026.

Long: Beam is the mechanism that closed the formal-sandwich surface for the majority of Solana retail flow. Published Ultra V3 slippage on Beam-protected paths runs approximately +0.006% versus −0.14% on unprotected paths — a sign-reversal that puts Beam-routed retail on the correct side of slippage on average. Ultra handles approximately 95% of Jupiter's aggregator volume and approximately 50% of all Solana DEX volume; Beam is the routing layer underneath. The April 2026 opening to non-Jupiter signed transactions broadens Beam from "Jupiter's internal transaction landing" to "Solana's default protected transaction-submission surface." The structural fact: Beam routes retail flow into Jito's private bundle market, which means most retail transactions on Solana are no longer publicly observable until the block is built.

TradFi analogue: A retail broker's smart order router that internalises by default to a small set of trusted wholesalers — same structural shape as US equity PFOF flow, with the rebates flowing back through the wallet layer (Chapter 7).

First appears: Chapter 3 (cameo); full treatment in Chapter 8

See also: Jupiter, Jito tip, BAM, Private order flow, RPC provider.

Kolibrio

Short: A VC-backed Order Flow Auction startup founded in early 2022 that operates an RPC-layer mempool surface — wallets, bots, and solvers route user transactions through Kolibrio, which auctions pre-leader transaction visibility to searchers and rebates a share back to the integrator.

Long: Kolibrio raised a $2 million seed in February 2023 led by Jump Crypto, with Delta Blockchain Fund and Everstake Capital. Its product, sometimes branded as "Meow RPC," is mechanically a paid-tier version of what Jito's public mempool used to do — the integrator sees a private-routing endpoint; searchers willing to pay see a private mempool feed; the integrator and Kolibrio split the resulting MEV rebates. Kolibrio operates at the RPC layer, distinguishing it from the validator-side private mempools the Solana Foundation enforced against in June 2024 — but the visibility surface it sells is functionally similar. Live on both Solana and Ethereum.

TradFi analogue: A payment-for-order-flow wholesaler — except that the wholesale "sale" happens at the RPC layer rather than at the broker layer, and the searchers it sells to are not registered market makers but anonymous on-chain bots.

First appears: Chapter 3

See also: Private order flow, MEV-Blocker, Block builder, Mempool.

Liquidation

Short: The forced closure of an under-collateralized loan on a DeFi lending protocol, executed by a third party in exchange for a protocol-paid bonus.

Long: When a borrower's health factor falls below 1.0, any third party (typically a searcher) can repay a portion of the borrower's debt to the protocol, seize a corresponding portion of the borrower's collateral, and keep a protocol-defined bonus on the seized value. Liquidations keep the lending protocol solvent — they are the mechanism by which bad debt is removed before it can become a hole in the protocol's balance sheet. The borrower would have lost the position under any healthy protocol; the searcher's role is to perform the transaction quickly enough that the collateral is still worth more than the debt.

TradFi analogue: A margin call followed by a broker-executed forced sale, with the difference that DeFi liquidations are competitive — any third party can fill the role.

First appears: Chapter 4

See also: Liquidation bonus, Health factor, Searcher.

L2 sequencer

Short: The role of ordering transactions and producing blocks on an Ethereum Layer 2 (rollup). In May 2026 every major L2 in production operates with a single centralised sequencer.

Long: The L2 sequencer is structurally analogous to a single block builder with no proposer-side competition — it controls transaction ordering, MEV extraction, and fee revenue at the operator without a separate proposer to bid against. Operators in May 2026: Arbitrum One (Offchain Labs); Base (Coinbase); OP Mainnet (OP Labs); Unichain (Uniswap Labs); Linea (Consensys; multi-node QBFT consensus but Consensys-affiliated operators). Combined L2 sequencer revenue runs approximately $150–250 million annually in 2025–2026; Base captures approximately 62% of all L2 fee revenue (~$67M/year, ~$870M cumulative, ~$55M net profit in 2025 as the only profitable L2). Shared / decentralised sequencer designs (Espresso, Astria, Optimism Superchain) are in production but not yet at material scale; realistic horizon for major L2 sequencer decentralisation is late 2026 to 2027.

TradFi analogue: An exchange's matching engine that is also the exchange operator and the regulated broker-dealer — three roles a market-structure regulator would not normally permit a single firm to combine.

First appears: Chapter 10

See also: PBS, Block builder, Cross-chain MEV, Timeboost.

Liquidation bonus

Short: The premium (typically 5–15% on Aave V3, ~5% on MarginFi) the searcher receives on the value of seized collateral when performing a liquidation.

Long: The liquidation bonus is paid out of the borrower's seized collateral, not by the protocol itself; mechanically, the searcher repays $X of debt and receives $X × (1 + bonus) of collateral. Bonuses are set by governance and vary by asset — more volatile collateral attracts a higher bonus to compensate the liquidator for price risk during the few seconds between trigger and settlement. In Aave V4, the bonus scales dynamically with how far below 1.0 the health factor has fallen, rather than firing at a single static rate.

First appears: Chapter 4

See also: Liquidation, Health factor.

Liquidity

Short: The supply of resting orders or pooled inventory ready to fill a trade at a known price; without it, a market has only intent.

Long: Liquidity is the precondition for trading. A venue is liquid to the extent that a trader can transact at the size they want with a small spread and minimal price impact. On-chain, liquidity is supplied by passive LPs depositing into AMM pools, by professional market makers posting resting orders on CLOBs, and — increasingly — by prop-AMMs and solvers that quote against incoming flow without exposing a public position. The architecture of the venue determines who is willing to supply it.

TradFi analogue: The same word, with the same meaning. A liquid stock or futures contract is one whose positions can be opened or closed in size without moving the price.

First appears: Chapter 2

See also: AMM, CLOB, Liquidity provider, Spread.

Liquidity provider (LP)

Short: A participant who supplies inventory to a market — most often by depositing into an AMM pool — in exchange for a share of trading fees.

Long: On-chain LPs come in two varieties. A passive LP deposits into a pool and accepts the algorithm's pricing; their P&L is fees earned minus impermanent loss (or, under the LVR framework, fees earned minus the arbitrageur's profit). An active LP manages the position — adjusts ranges on concentrated-liquidity pools, rebalances, hedges externally. The 2021–2022 empirical work on Uniswap V3 found that roughly half of passive LPs underperformed simply holding the underlying assets net of impermanent loss; no comprehensive 2025–2026 re-measurement has been published.

TradFi analogue: A specialist market maker at the NYSE in the open-outcry era — but without the regulatory protections, the bid-ask spread, or the privileged information access.

First appears: Chapter 2

See also: AMM, Impermanent loss, Loss-Versus-Rebalancing, Passive LP, Market maker.

Loss-Versus-Rebalancing (LVR)

Short: A measure of the loss a liquidity provider incurs against an arbitrageur who continuously rebalances the same position via a centralised exchange — a sharper formulation of LP loss than impermanent loss.

Long: Developed by Milionis, Moallemi, Roughgarden, and Zhang, LVR reframes the LP's economic problem. An LP is not, in the relevant comparison, competing with a HODLer; they are competing with someone who would have closed the same position by rebalancing continuously against a centralised exchange. Under LVR, an AMM pool experiencing 5% daily volatility loses approximately 3.125 basis points per day, or about 11% annualized, to the arbitrageurs who keep the pool's price aligned with the broader market, before fees. LVR scales with the square of volatility — doubling volatility quadruples LVR.

First appears: Chapter 2

See also: Impermanent loss, AMM, Liquidity provider.

Market maker (MM)

Short: A participant who supplies two-sided quotes to a venue, earning the spread between bids and asks.

Long: Market makers are the participants who make prices in an order book. On Hyperliquid, the named MMs are now part of the regulatory record — Wintermute is quoting across 76 markets with approximately $199 million in resting notional as of January 2026, alongside Flowdesk and others named in Bitwise's April 2026 ETF S-1 amendment. On Ethereum solver markets, the same firms appear in different shape: SCP and Wintermute together fill more than 90% of UniswapX volume. The MM business spends on inventory risk and infrastructure; the moat is increasingly access to the venue's flow, not raw algorithmic edge.

TradFi analogue: A market-making desk at Citadel Securities, Jane Street, Optiver, or DRW.

First appears: Chapter 2

See also: Searcher, Liquidity provider, Spread, Solver.

Mempool

Short: The public queue of pending transactions that have been broadcast to a network but not yet included in a block.

Long: When a user submits a transaction on a chain like Ethereum, it does not go directly into the next block; it sits in a public-facing pending state visible to anyone running a node. The mempool is the surface that makes most sandwich attacks possible: searchers read the queue, identify exploitable patterns, and submit their own transactions to capture them. Solana does not run a single global mempool in the Ethereum sense — its closest equivalent is the leader-validator's TPU and the public RPC paths that feed it. Hyperliquid has no mempool at all: orders match inside HyperBFT consensus with no pending state to observe.

TradFi analogue: A pre-trade order book visible to all market participants, except that mempools have historically been visible to everyone, not just registered market makers.

First appears: Chapter 4 (cameo); full treatment in Chapter 3.

See also: Frontrun, Sandwich attack, MEV.

MEV

Short: The profit a privileged participant can earn by reordering, inserting around, or censoring transactions inside a block — money that comes out of other people's trades.

Long: MEV is the umbrella term for the value extractable from the right to choose transaction ordering. The acronym was originally coined as "miner extractable value" and is now most often expanded as "maximal extractable value"; the book avoids the expansion because it offers no help to a business reader. MEV manifests in three main forms: arbitrage (closing price gaps), liquidations (forcing repayment of bad debt), and sandwiches (extracting from individual traders by trading around them). The first two are useful or neutral; the third is a pure transfer from the trader to the searcher and the validator.

TradFi analogue: The aggregate revenue captured by HFT firms, exchange specialists, and payment-for-order-flow market makers — every actor with a privileged position in the order-handling pipeline.

First appears: Chapter 4

See also: Searcher, Sandwich attack, Liquidation, Arbitrage.

MEV-Blocker

Short: An Ethereum private-routing service, originally operated by CoW DAO and transferred to Consensys' Special Mechanisms Group in January 2026, that returns a share of captured backrun value back to the user.

Long: MEV-Blocker has served over 4.5 million unique wallets and paid 6,177 ETH in cumulative rebates to users since launch. Mechanically similar to Flashbots Protect — a user changes their wallet's RPC URL to MEV-Blocker's endpoint, transactions are routed as private bundles — but with a different revenue-share model and operator. The January 2026 transfer to Consensys' SMG was framed as enabling broader development of backrun auction infrastructure.

First appears: Chapter 3

See also: Flashbots Protect, Private order flow, MEV-Share.

MEV-Boost

Short: The Ethereum proposer/builder/relay separation infrastructure; the standard mechanism by which Ethereum validators outsource block construction to specialised builders.

Long: MEV-Boost, introduced post-Merge in 2022, is the architectural pattern that lets validators (the proposers) outsource block construction to specialised builders via relays. The builder constructs the most profitable block it can find, including bundles from searchers and from private-routing services; the relay forwards the block to the proposer; the proposer signs it. As of May 2026, the structural concentration is sharp: live snapshots from relayscan.io show one builder, Titan, producing approximately 50% of all Ethereum blocks, with BuilderNet and Quasar bringing the top three to approximately 91% of all mainnet blocks. Flashbots itself stopped operating a centralised builder in December 2024 and migrated its building infrastructure into BuilderNet.

TradFi analogue: The relationship between a public-stock-exchange listing, the broker-dealers who provide liquidity, and the wholesalers who internalise retail orders — except that on-chain, the proposer–builder–relay separation is explicit at the protocol level.

First appears: Chapter 3

See also: Block builder, Relay, Flashbots Protect, MEV.

Short: Flashbots' selective-disclosure mechanism: users opt into letting searchers see specific parts of a transaction in exchange for a share of any extracted value.

Long: MEV-Share occupies a middle ground between full public-mempool exposure and full private routing. A user signing a transaction can choose which fields are visible to searchers — for example, allowing the swap path and pool but hiding the amount, or vice versa — and any searcher who profits from the partial information must share the proceeds back with the original submitter. Adoption has been steady but never matched the simpler URL-swap path that Flashbots Protect offers, primarily because most users prefer the binary choice (fully private or fully public) over the partial-disclosure complexity MEV-Share introduces.

First appears: Chapter 3

See also: Flashbots Protect, Private order flow, MEV-Blocker.

Pack tile (Firedancer)

Short: The Firedancer / Frankendancer component that decides transaction ordering within a slot; supports user-selectable scheduler modes.

Long: The pack tile is functionally analogous to Agave's central scheduler but ships with explicit user-configurable modes documented in the Firedancer codebase: perf ("fill the block as fast as possible using the highest-paying transactions available… consistently 100% full blocks") and balanced (the default, "optimizes for revenue from priority fees, but can result in blocks that are not always 100% full, and can be poor at capturing MEV if using an external block builder"). The documentation explicitly recommends reverting to balanced under network congestion. A third mode — revenue — was deployed alongside perf and balanced through early 2026; Syndica's March 2026 data shows approximately 47% of Frankendancer-Jito stake on revenue at that point, roughly equal to the 47% on balanced. The revenue mode filled blocks "extremely lazily," maximising the window for high-fee transaction reordering at the cost of block fullness. It was removed in Frankendancer testnet release v0.902.40002 on 21 March 2026, one of eight bullet points in the release notes with no accompanying rationale from Jump Crypto.

First appears: Chapter 5

See also: Frankendancer, Firedancer, Central scheduler (Solana).

MEV refund

Short: A payment from a successful searcher or solver back to the original transaction submitter, typically routed through a private-routing service or Order Flow Auction operator.

Long: MEV refunds are the structural mechanism by which value extracted by a third party (a searcher running a backrun, a solver filling an intent, etc.) is partially returned to the user whose transaction created the opportunity. The split varies by operator: Helius's wallet-customer rebates run 50/50 between Helius and the wallet (Phantom, Backpack, Solflare et al.); MEV-Blocker / Special Mechanisms Group returns approximately 90% of backrun surplus to users and 10% to validators; Flashbots Protect has returned approximately 313 ETH cumulative in refunds across $43 billion of shielded volume. The refund mechanism converts MEV from a one-sided extraction into a two-sided market where the user captures a documented share of the value their transaction created.

TradFi analogue: Payment-for-order-flow rebates that some institutional retail brokers (Schwab, Fidelity) pass through to customers in the form of price improvement on retail trades. The on-chain refund market is structurally similar but more explicit — the refund is a discrete payment rather than an embedded price improvement.

First appears: Chapter 6

See also: Order Flow Auction (OFA), Flashbots Protect, MEV-Blocker, RPC provider.

Order Flow Auction (OFA)

Short: A market in which a routing entity sells the right to fill a transaction to whichever searcher or solver bids the most, with a share of the bid refunded to the original submitter.

Long: OFAs operate at different layers of the on-chain trading stack depending on the implementation. Kolibrio's OFA operates at the RPC layer: integrators (wallets, bots, solvers) route through Kolibrio's "Meow RPC" endpoint and Kolibrio auctions the visibility to searchers. Flashbots' MEV-Share operates at the wallet-RPC layer with selective disclosure of transaction details. CoW DAO's solver auction operates at the intent layer — the user signs an intent (not a specific transaction) and a pool of solvers bids to fill it. MEV-Blocker (now operated by the Special Mechanisms Group at Consensys) operates a backrun auction at the private-RPC layer. All variants share a common structural element: the routing entity captures the searcher's bid and returns a share to the user, with the operator taking a margin in between.

TradFi analogue: The wholesaler internalisation business in US equities — a wholesaler (Citadel Securities, Virtu) pays a broker (Robinhood, Schwab) for the right to fill the broker's retail orders against the wholesaler's inventory, with the wholesaler internalising the order at or inside the public quote.

First appears: Chapter 6

See also: MEV refund, Flashbots Protect, MEV-Blocker, MEV-Share, RPC provider.

Passive LP

Short: A liquidity provider who deposits into an AMM pool and accepts the pool's algorithm to manage the position, versus an active LP who manages ranges, rebalances, or hedges externally.

Long: The passive LP is the role the 2021 retail-DeFi wave most prominently filled and the role whose outcomes have been most studied. Empirical work on Uniswap V3 found that the passive LP's fee revenue often does not cover their impermanent loss; the LVR framework gives a sharper version of the same result. In 2026, on Solana, the role has been largely displaced by prop-AMMs that perform the same function with better adverse-selection pricing and keep the proceeds for the operator.

First appears: Chapter 2

See also: Liquidity provider, Impermanent loss, Loss-Versus-Rebalancing, Prop-AMM.

PBS / Proposer-Builder Separation

Short: The architectural pattern in which block proposing and block construction are performed by different actors, connected by a relay.

Long: PBS emerged on Ethereum post-Merge as an out-of-protocol optimisation. The proposer (the validator scheduled to produce a block) outsources block construction to a builder — a specialised firm whose software constructs the most valuable block possible from the available transactions — via a relay, which acts as a commitment device to prevent the proposer from extracting MEV by re-organising the builder's work. MEV-Boost is the dominant implementation in 2026. The proposed protocol-level version is ePBS (EIP-7732), which would bring the separation into the Ethereum protocol layer itself but which had not shipped as of mid-2026.

TradFi analogue: The relationship between an exchange's order-matching engine, a market-on-close auction operator, and a wholesaler that internalises retail flow — three distinct functions that on legacy markets had distinct operators, here separated explicitly at the protocol layer.

First appears: Chapter 5

See also: ePBS, MEV-Boost, Block builder, Relay (Ethereum), Proposer.

Pipeline rebate

Short: A revenue-share arrangement under which an RPC provider routes a wallet's transaction flow through a private auction with pre-approved searchers and splits the resulting MEV capture with the wallet — typically 50/50 on Solana in 2026.

Long: The canonical case is Helius's 50/50 wallet-rebate split with the eight wallets named on the firm's Wallets product page: Phantom, Ledger, Bitgo, Exodus, Backpack, Squads, Trust, and Solflare. The mechanism: a transaction routed through the RPC is exposed to a KYC'd searcher pool that submits backrun-only bids paid in SOL in the same block; the resulting value is split between the RPC and the wallet that submitted the transaction. Whether the wallet passes the rebate through to the end user or retains it as platform revenue is not publicly disclosed — Phantom's reported 2025 revenue of approximately $79.1 million does not separately itemise the rebate component. The pipeline-rebate structure is the Solana-side functional analogue of US equity payment-for-order-flow, except that on Solana the rebate nominally flows toward the user (through the wallet) rather than directly to the broker.

TradFi analogue: The price-improvement payments a wholesaler returns to a retail broker in US equities — except that the on-chain version is layered through a wallet rather than a broker and is not subject to Rule 606 routing-disclosure requirements.

First appears: Chapter 6 (cameo); full treatment in Chapter 7.

See also: Exclusive order flow (on-chain), RPC provider, Order Flow Auction (OFA), MEV refund.

Priority fee

Short: A per-transaction fee a sender pays above the base network fee to influence the chance and ordering of inclusion in the next block.

Long: On Ethereum, the priority fee (also called the "tip") is one of the two components of the post-EIP-1559 gas price and is paid directly to the validator who proposes the block. On Solana, the priority fee is a parallel mechanism layered on top of the network's base 5,000-lamport-per-signature charge; transactions with higher priority fees are scheduled earlier within the leader-validator's slot. Priority fees are the dominant ordering signal on both chains for transactions that are not routed through private flows.

TradFi analogue: An expedited-processing fee at a clearinghouse, or the price difference between standard and white-glove execution at a prime broker.

First appears: Chapter 4

See also: Jito tip, Block builder, MEV.

Private order flow

Short: Transactions delivered directly to a block builder, sequencer, or specialised confidential-execution surface (TEE enclave), bypassing the public mempool.

Long: Private order flow is the dominant value-routing pattern on Ethereum in 2026: approximately 30% of transactions by count, more than 50% of gas, and 54.59% of total block value now flow through private channels. On Solana, the same pattern is implemented at the wallet-router layer (Jupiter Beam) and at the block-construction layer (Jito BAM). The trade-off is structural: private routing reduces extraction risk for the submitting user but increases the concentration of trust in a small number of routing operators, which has knock-on consequences for censorship resistance.

First appears: Chapter 3

See also: Public order flow, Flashbots Protect, MEV-Blocker, BAM, Block builder.

Prop-AMM

Short: A private, market-maker-operated AMM whose pricing logic is closed and whose inventory is not on a public dashboard, settling on-chain but quoted off-chain.

Long: Prop-AMMs are the 2025–2026 evolution of on-chain market-making on Solana. HumidiFi (reportedly run by Temporal), SolFi (run by Ellipsis Labs, the team that built Phoenix), and Tessera together accounted for more than half of Solana spot DEX volume by Q1 2026. None of them takes outside deposits or accepts governance constraints; each is a single firm's market-making operation that happens to settle on-chain. The economic argument for the form: a market maker who can distinguish informed from uninformed flow prices each correctly and keeps the spread, where a public AMM cannot.

TradFi analogue: A bank's principal trading desk that quotes prices on an exchange — same business, on-chain rails.

First appears: Chapter 2

See also: AMM, Solver, Liquidity provider, Market maker.

Proposer (Ethereum)

Short: The validator scheduled to propose the next block in a given Ethereum slot. On Ethereum since the Merge, the proposer's job is to sign a block constructed by another actor (the builder).

Long: The Ethereum protocol selects a proposer per slot from the active validator set, weighted by stake. Pre-Merge, the proposer would have constructed the block itself from the public mempool. Post-Merge, with MEV-Boost adoption above 90%, the proposer's actual job has become to select the highest-bidding builder's block via the relay infrastructure and sign it. The "validator" role on Ethereum in 2026 is, in practice, the proposer role; the block-construction function has migrated to specialised builder firms.

First appears: Chapter 5

See also: PBS, Block builder, MEV-Boost, Validator, Slot leader (Solana).

Public order flow

Short: Transactions broadcast through the public path (Ethereum's public mempool, Solana's public RPC), visible to anyone observing the network before they are included in a block.

Long: The public mempool was, for most of crypto's history, the substrate for both extraction (sandwich attacks, frontrunning, generalised arbitrage) and observability (anyone could see what was about to happen on the chain). By 2026, on Ethereum, the public mempool still exists but holds the value-leftovers — approximately 30% of transactions by count but only 45.4% of block value. On Solana, the equivalent surface (a private mempool operated by Jito's client from August 2022 to March 2024) has been eliminated entirely. On Hyperliquid, no public order-flow surface ever existed by design.

First appears: Chapter 3

See also: Private order flow, Mempool, Flashbots Protect.

RFQ (request for quote)

Short: A trading model in which a trader requests a fillable price from one or more market makers, who respond with quotes; the trader picks the best.

Long: RFQ is the dominant institutional model in fixed income and FX, where order books are thinly populated and dealers prefer to quote in size on demand. On-chain, the RFQ model underlies the major Ethereum solver markets — UniswapX, 1inch Fusion, parts of CoW Swap. A user states an intent; an auction of professional MMs competes to fill it from inventory the MM chooses. The end result is mechanically different from an AMM but structurally similar: a public-facing trader meets a small set of professional liquidity providers behind an intermediary that hides the matching logic.

TradFi analogue: An institutional dealer-to-customer RFQ on Bloomberg or Tradeweb.

First appears: Chapter 2

See also: Solver, Aggregator, Market maker.

RPC provider

Short: A firm that operates the API gateway between user-facing applications (wallets, aggregators, bots) and the chain itself.

Long: RPC providers handle the protocol-level reads and writes that wallets and other apps need to interact with the chain — block lookups, balance queries, transaction submission. In 2026 the largest RPC providers also operate validator stake, MEV-rebate flows, and (in some cases) private-mempool surfaces, making the RPC layer structurally adjacent to the OFA and validator layers rather than purely a read-only gateway. Helius is the dominant Solana RPC provider with paid subscription tiers ($49–$999/mo plus enterprise) and a 50/50 MEV-rebate split with its wallet customers (Phantom, Backpack, Solflare, Bitwise). Triton One operates across Solana, Sui, Monad, and Pythnet and is the Solana Foundation's RPC 2.0 partner. Astralane is the HFT-tier alternative aimed at market makers. On Ethereum the legacy multi-chain providers (Infura, Alchemy, QuickNode) coexist with specialised crypto-native private-RPC operators (Flashbots Protect, MEV-Blocker, bloXroute).

TradFi analogue: The institutional data and order-routing providers (Refinitiv, Bloomberg, Reuters' direct feeds) that sit between trading firms and the exchanges they trade on — except that on-chain, the RPC provider is also frequently the actor that captures the MEV-rebate value flowing back to the wallet.

First appears: Chapter 6

See also: Order Flow Auction (OFA), MEV refund, Infrastructure firm.

Relay (Ethereum)

Short: An Ethereum entity that receives bundles from builders and submits them to validators, mediating between the proposer/builder/relay separation that MEV-Boost defines.

Long: Relays sit between block builders and validators in the MEV-Boost architecture: a builder constructs a block, hands it to a relay, and the relay forwards it to the proposer (the validator scheduled to propose the next block). The relay's role is to validate that the builder is paying what it claims and to provide the proposer a header to sign before the full block payload is revealed — a commitment device that prevents the proposer from extracting MEV by re-organising the builder's work. As of May 2026, Ethereum relay share is concentrated: Ultrasound 35.7%, Titan Relay 26.1%, BloXroute Max-Profit 13.4%, BloXroute Regulated 12.3%, Aestus 7.3%, Flashbots Relay 2.1%. Hasu (Flashbots) has publicly framed the absence of a neutral relay as "a failure of the Ethereum ecosystem."

First appears: Chapter 3

See also: Block builder, MEV-Boost, Bundle, Validator.

Rotor

Short: Alpenglow's block-propagation component, replacing Solana's existing Turbine + gossip propagation layer.

Long: Rotor is one of two new components introduced by Alpenglow (the other is Votor). It handles the propagation of blocks to validators, working alongside Votor's BLS-aggregated off-chain vote mechanism to achieve the two-path finality targets (≥80% in round 1 = ~100ms; ≥60% in round 2 = ~150ms). Mainnet activation is expected in late Q3 or early Q4 2026.

First appears: Chapter 5

See also: Alpenglow, Votor, Validator.

Rakurai

Short: A closed-source fork of the Agave Solana validator client, built by a small team with ASIC and system-on-chip engineering backgrounds; positioned as the "no timing games" alternative to Jito-Solana for institutional operators.

Long: Rakurai raised a $3 million seed in 2025 led by Anagram Ventures (with Slow Ventures, Robot Ventures, Crypto.com, P2P.org, GlobalStake, and Cyber Fund). CEO Ali Rizvi is ex-Apple; the team's engineering background is in low-latency systems rather than crypto-native infrastructure. Mainnet stake share was approximately 2% in January 2026, rising to approximately 6% by March 2026. Figment's published migration to Rakurai on 2 March 2026 produced a 32-basis-point increase in Staking Reward Rate (6.85% to 7.17%), with non-vote transactions per block up 49%, revenue per block 2.3×, priority fees up 60%, and MEV capture approximately 5× — gains achieved through scheduler optimisation rather than block-timing manipulation. Figment's migration report explicitly notes that Rakurai "do[es] not resort to block timing game manipulation with intentional delays past the target block time of 400ms," positioning Rakurai as the legitimate alternative for institutional operators concerned about SFDP compliance and block-timing-game discourse.

First appears: Chapter 5

See also: Agave, Validator client, Timing games, Frankendancer.

Sandwich attack

Short: A searcher attack in which a buy is placed immediately before a victim's swap and a sell immediately after, so the victim fills at a price the attacker has just moved.

Long: A sandwich requires that the victim's transaction be visible before it executes, that the victim's swap be large enough to move the relevant pool's price by more than the round-trip cost, and that the attacker can win priority over the victim and over other searchers watching the same opportunity. The attacker's profit is the price degradation the victim experiences — the difference between the pre-frontrun pool price and the price the victim's transaction actually fills at, minus the attacker's costs (gas, base fees, priority fees and tips paid to validators). Sandwiches are pure extraction: they perform no useful function for the protocol or the network, and their profit is the victim's loss.

TradFi analogue: Closest to front-running a known large client order, then closing the position immediately after — a practice regulated against in equities markets but structurally legal in most on-chain venues until protocol-level mitigations like Solana's April 2026 patch took effect.

First appears: Chapter 4

See also: Frontrun, Backrun, MEV, Mempool.

Searcher

Short: A firm or operator that runs latency-optimized software to find and capture MEV opportunities on-chain.

Long: Searchers are typically small teams (two to fifteen engineers) operating dedicated infrastructure colocated with the relays, block engines, and RPC nodes that route transaction flow. They earn from arbitrage, liquidations, and sandwiches; they spend on gas paid on every attempt (win or lose), on infrastructure, and on engineering. By 2025, the searcher business had consolidated sharply — the number of active Ethereum CEX–DEX searchers fell from 23 in late 2024 to 11 in Q1 2025, with the top three capturing 90% of volume — and the competitive moat is increasingly the relationship with one or more block builders, not raw algorithmic speed.

TradFi analogue: An HFT arbitrage desk at Citadel Securities or Jane Street circa 2010.

First appears: Chapter 4

See also: MEV, Block builder, Validator.

Settlement

Short: The moment a trade is matched and executed at the venue — on a centralised exchange, when the matching engine fills the order; on-chain, when the transaction is included in a block.

Long: The word has two slightly different meanings depending on context. In TradFi, "settlement" usually means the final cash-and-asset transfer at a clearinghouse (T+1 for US equities since May 2024, T+2 before that). In crypto, "settlement" is more often used colloquially to mean inclusion-in-a-block — the chain has recorded the trade, even if it has not yet finalised. The book uses both meanings carefully: TradFi settlement is the legal final transfer; on-chain settlement is the on-chain record of the trade, with finality as a separate concept covering irrevocability.

TradFi analogue: T+1 cash-and-asset transfer at DTCC for US equities; same-day at Hong Kong Exchanges and Clearing for HK equities.

First appears: Chapter 1

See also: Trade, Transaction, Finality, Block.

SFDP (Solana Foundation Delegation Program)

Short: The Solana Foundation's stake-delegation pool — the largest single-source delegated stake on the network and the chain's de facto soft-governance lever. The program delegates Foundation-controlled stake to qualifying validators that meet operational, geographic, and behavioural criteria; removal from SFDP is the Foundation's primary enforcement mechanism against extractive or concentrated validator behaviour.

Long: Starting 1 May 2026, SFDP participants must operate on an ASN and hosting provider that holds less than 25% of network stake, and the program enforces a separate <15% data-centre concentration cap across all staked validators at any single data-centre provider — an explicit counter to the Frankfurt/Amsterdam colocation gravity that drives the chain's geographic concentration. The "onboard 1, offboard 3" policy retires validators with less than 1,000 SOL of external stake after 18 months of program participation. Approximately 150 validators are projected to lose Foundation stake under the May 2026 rule. The program's anti-extraction lever is the censorship-resistance and "no sandwich-enabling private mempools" rules — the June 2024 enforcement that removed approximately 32 validators for participating in DeezNode-style arrangements was executed under these rules (Chapter 3, Chapter 7). Tim Garcia (Solana Foundation validator-relations lead) has described the enforcement as "ongoing" but the Foundation has not published an itemised list of subsequent removals.

TradFi analogue: The discretionary order-routing relationships a large institutional asset manager maintains with broker-dealers — the relationship is not regulated, but the threat of withdrawal is the asset manager's primary lever over broker behaviour.

First appears: Chapter 5 (cameo); full treatment in Chapter 8

See also: Validator, Validator client, Private order flow, Sandwich attack.

Slashing

Short: The protocol-defined penalty applied to a validator's stake for misbehaviour (double-signing, missed attestations, etc.).

Long: On Ethereum, slashing results in stake destruction — the validator loses a portion of its 32-ETH bond and is forcibly exited from the active validator set. From the Beacon Chain's December 2020 launch through September 2025, fewer than 500 validators were slashed against an active count over 1.2 million in 2025; the dominant cause is operator misconfiguration rather than malicious behaviour. The largest 2025 cluster was a 39-validator simultaneous slashing on 10 September 2025 traced to SSV Network distributed-validator-technology operators during routine maintenance. On Solana, there is no stake-slashing equivalent for vote-skipping; penalties take the form of reduced vote credits and reduced inflation-share rewards, with Alpenglow's Validator Admission Ticket (~1.6 SOL burned per epoch) as the first explicit per-epoch fixed cost.

TradFi analogue: A clearinghouse member's posted-collateral confiscation following a settlement failure — except that crypto slashing is rules-based and automatic rather than discretionary.

First appears: Chapter 5

See also: Validator, Validator Admission Ticket (VAT), Staking Reward Rate (SRR).

Slippage

Short: The difference between the price a trader expected and the price they actually got, expressed as a percentage of the trade or in basis points.

Long: Slippage on an AMM is a function of pool depth and trade size — the larger the trade relative to the pool, the more the swap moves the price as it executes. Slippage on a CLOB is a function of how much of the spread the trader has to cross and how much depth they consume. Aggregators reduce slippage by splitting a trader's order across multiple venues; solver markets reduce slippage by routing the order through a private MM that fills from inventory, though the trader pays the MM via the price they're filled at.

TradFi analogue: The same word. Equity HFT firms make their money in part by reducing the slippage they charge institutional clients while widening the slippage they charge retail flow.

First appears: Chapter 2

See also: Spread, AMM, Aggregator.

Slot leader (Solana)

Short: The validator scheduled to produce the current Solana slot — the block-construction window.

Long: Solana publishes its leader schedule per epoch (approximately every two days), so slot leaders are known in advance. Unlike Ethereum's proposer role since the Merge, the Solana slot leader is also the block constructor: the validator's choice of client software and scheduler configuration determines what transactions enter the block and in what order. The slot-leader role on Solana is structurally where the proposer + builder roles meet, which is why the choice of validator client has become the network's largest single revenue lever.

First appears: Chapter 5

See also: Validator, Validator client, Agave, Frankendancer, Proposer (Ethereum).

Solver

Short: A firm that fills a user's on-chain trading intent from inventory it chooses, in an auction against other solvers — the Ethereum analogue of a prop-AMM.

Long: Solvers are the participants in CoW Swap, UniswapX, and 1inch Fusion. A user states an intent ("get me at least X ETH for my Y USDC"); solvers bid to fill it; the best bid wins. The solver supplies the liquidity from wherever — its own inventory, an AMM pool, a CEX hedge — and captures the difference between what the user receives and what the solver paid. Two firms (SCP and Wintermute) together account for more than 90% of UniswapX volume; Barter is targeting more than 50% of CoW solver share. The solver model lets the same professional MM firms that quote on Hyperliquid and on centralised exchanges fill Ethereum retail flow without ever depositing into a public pool.

TradFi analogue: The principal-trading desk that fills a client's order in an RFQ — except that on-chain the auction is fully visible and the solver competition is real-time.

First appears: Chapter 2 (cameo); full treatment in Chapter 7.

See also: Prop-AMM, RFQ, Aggregator, Market maker.

Spread

Short: The difference between the best bid and the best ask in an order book, expressed in price units or basis points.

Long: The spread is the market maker's compensation for hosting liquidity. A tight spread means an MM is willing to be picked off cheaply; a wide one means the MM is pricing in adverse selection — the risk that the trader hitting the quote knows more about the underlying than the MM does. Hyperliquid's disclosed BTC perpetual top-of-book spread of approximately $1 (against Binance's ~$5.50) in early 2026 is a sign that the venue's MMs perceive lower adverse selection there than on the largest centralised venue — a counter-intuitive finding driven by HyperBFT's lack of public mempool, which prevents takers from seeing and lifting stale quotes.

TradFi analogue: The same word. Equity bid-ask spreads on NYSE-listed S&P 500 stocks are typically under a basis point; on smaller venues they widen substantially.

First appears: Chapter 2

See also: Market maker, CLOB, AMM.

Staking Reward Rate (SRR)

Short: The realised annualised yield on staked balance for a given validator, encompassing base issuance, fees, MEV-derived revenue, and operational discipline.

Long: SRR is the headline number a staking-as-a-service provider quotes its clients. On Ethereum, the network-average SRR was approximately 2.94% in Q3 2025, with ~93% derived from consensus-layer rewards (base issuance and attestations) and ~7% from execution-layer rewards (proposed blocks). Validators running MEV-Boost typically realise approximately 5.69% APY versus approximately 4% without; priority fees account for approximately 17% of total validator income; MEV-Boost contributes approximately 3% APY. On Solana, the SRR composition is materially different — Figment's published migration to Frankendancer at epoch 871 (30 October 2025) increased its SRR by 18 basis points (peak 28), with MEV / Jito tips up 9× and priority fees up 2.5× on a single client switch, demonstrating that client choice is the dominant determinant of Solana SRR.

First appears: Chapter 5

See also: Validator, MEV-Boost, Frankendancer, Validator client.

TEE / SEV-SNP

Short: Trusted Execution Environment — a hardware enclave (AMD SEV-SNP is the variant Jito BAM uses) that keeps program contents and inputs encrypted from the operator of the host machine, with cryptographic attestations that the code running inside the enclave is the code claimed.

Long: BAM Nodes on Solana run inside AMD SEV-SNP enclaves, which keep transaction contents encrypted until execution and produce attestations of the ordering decision. The privacy guarantee depends on AMD's hardware behaving as specified rather than on cryptographic primitives — this makes BAM trust-minimised rather than fully trustless. TEE overhead in BAM is approximately 2–5% of normal validator load. TEEs are not unique to crypto — Intel SGX, AMD SEV-SNP, and ARM TrustZone are used in cloud computing, content protection, and confidential computing more broadly — but their use as a substrate for block construction is a 2025–2026 development specific to the Solana stack.

TradFi analogue: A hardware security module (HSM) inside which sensitive matching or settlement logic runs — the matching engine's operator can prove the code is correct without anyone needing to trust the operator's discretion.

First appears: Chapter 3

See also: BAM, Block builder, Bundle.

Timeboost (Arbitrum)

Short: Arbitrum's sealed-bid second-price auction selling a 200-millisecond "express lane" priority for a 60-second slot. Launched April 2025. Approximately $3 million per year run-rate as of March 2026 — the L2-side exclusive-flow analogue.

Long: Timeboost is the canonical published L2-side EOF-equivalent surface in 2026. The auction's revenue routes 97% to the Arbitrum DAO and 3% to the auctioneer; the express-lane auction now accounts for approximately 20–30% of daily DEX volume on Arbitrum. Castro et al.'s September 2025 empirical analysis (arXiv 2509.22143) documents that two entities win more than 90% of Timeboost auctions, with competition declining over time and revenue trending downward as auction concentration tightened. The pattern is the L2-sequencer-layer echo of the L1 builder-direct contracts Chapter 7 documented — same exclusivity-by-stake-and-relationship dynamic, one architectural level out. Generated approximately $3 million in first three months post-April 2025 launch; approximately 26% of Arbitrum DAO income in 2025–2026.

First appears: Chapter 10

See also: L2 sequencer, Exclusive order flow (on-chain), Builder-direct relationship.

Timely Vote Credits (TVC)

Short: The Solana vote-credit mechanism (SIMD-0033, activated November 2024) that awards up to 16 vote credits for votes confirmed within 2 slots, with diminishing credits for later votes.

Long: Pre-TVC, Solana validators earned vote credits at a constant rate regardless of timing. TVC made vote-landing-latency a competitive variable: a vote confirmed within 2 slots earns the maximum 16 credits; each additional slot of latency costs one credit; the floor is 1 credit at ≥18 slots. The rule converted operational latency from a constant cost into a measurable competitive moat — better hardware, better geographic placement, and better client configuration translate directly into more vote credits and ultimately more reward. Chorus One's reported skip rate of approximately 2.03% versus a network average of 5.19% is the moat operationalised.

First appears: Chapter 5

See also: Validator, Slot leader (Solana), Alpenglow.

Timing games (Solana block production)

Short: The strategy in which a Solana validator artificially delays Proof-of-History ticking to extend the duration of its slot beyond the protocol's nominal 400-millisecond target, giving its scheduler more time to pack high-priority-fee transactions into the current block.

Long: Timing games operate at the block-production layer rather than the vote-attestation layer, distinguishing them from "vote late" / "vote early" voting strategies. Chorus One's August 2025 empirical research (Timing Games on Solana: Validator Incentives, Network Impacts, and Agave's Hidden Inefficiencies) quantified the economic payoff: combined timing-games and scheduler optimisation produced approximately +3.0% in total rewards (about 27 basis points annualised). The network-level cost is a reduction in effective inflation — extended slots translate to slower epochs, reducing nominal inflation to roughly 80% at 500-millisecond slot times. Validator clients that pack blocks well without timing manipulation — Rakurai, Firedancer/Frankendancer — capture similar gains while staying inside the 400-millisecond target, suggesting Agave's scheduler weakness rather than a fundamental protocol vulnerability.

First appears: Chapter 5

See also: Validator, Rakurai, Frankendancer, Agave, Central scheduler (Solana).

Trade

Short: The act of exchanging one asset for another at an agreed price.

Long: A trade exists when a buyer and a seller have agreed on a price for a quantity of an asset. On a centralised exchange, the match happens inside the venue's matching engine; on-chain, the match happens when a transaction is included in a block and executes against a liquidity pool or a counterparty order. Trades are the basic unit of market activity; everything else in this book is about what surrounds the trade — the actors who help it happen, the actors who profit from watching it, and the architectures that make each kind of trade possible.

TradFi analogue: The same word, same meaning.

First appears: Chapter 1

See also: Transaction, Settlement, Finality.

Transaction

Short: The on-chain instruction that, when executed, settles a trade — or transfers funds, or interacts with a smart contract.

Long: A transaction is the unit a user submits to a chain. It contains a sender, a destination (which may be another wallet or a smart contract), a nonce, a payload, and a signature. The chain's validators decide whether to include the transaction in a block and in what order. A trade is what the user wanted; a transaction is the on-chain instruction that, when executed, makes the trade happen. The user pays gas for the transaction whether it succeeds or fails.

TradFi analogue: An order ticket plus a clearing instruction, combined into one indivisible unit.

First appears: Chapter 1

See also: Trade, Settlement, Block, Gas.

Validator

Short: An operator who stakes a chain's native token, runs validator software, and participates in block production for revenue.

Long: The validator's job, as originally framed in the founding documents of most proof-of-stake chains, was neutral plumbing: software running in datacentres, taking a fixed share of issuance, staying out of the way. By 2026, the role has bifurcated. On Ethereum, the validator is the proposer but typically not the block constructor — specialised firms called builders construct the block, and the validator signs it in exchange for an MEV-Boost payment. On Solana, the validator is the slot leader and still constructs the block, but the software the validator runs — the validator client and its scheduler configuration — has become the largest single determinant of revenue, with per-block priority-fee captures ranging from baseline to +101% versus median depending on the configuration. On Hyperliquid, the validator is part of a small permissionless-top-21 set running HyperBFT consensus, with no block-builder role to separate. The structural argument: the validator role has bifurcated into a signing function (still mostly neutral) and an ordering function (commercial).

TradFi analogue: A specialist on the floor of the NYSE circa 1995 — the actor with privileged information access who handled order matching for one or more tickers, earning a share of the spread, with operational discipline as the visible criterion but information relationships as the actual moat.

First appears: Chapter 5

See also: Proposer, Slot leader (Solana), Block builder, Validator client, Slashing.

Validator Admission Ticket (VAT)

Short: Alpenglow's per-epoch fixed cost for Solana validators — approximately 0.8 SOL per day (1.6 SOL per epoch), entirely burned. The first explicit per-epoch fixed cost on the network.

Long: The VAT is one of Alpenglow's three load-bearing economic changes (the others being the removal of vote transactions from blockspace and the introduction of two-path finality). Helius's modelling suggests the VAT economics drop the profitable-validator stake threshold from approximately 4,850 SOL (~$800,000) to approximately 450 SOL (~$75,000) — a roughly tenfold reduction in the stake required to operate profitably, which would meaningfully widen who can run a validator. Mainnet activation is expected in late Q3 or early Q4 2026.

First appears: Chapter 5

See also: Alpenglow, Validator, Slashing.

Validator client

Short: The software a validator runs to participate in block production. On Solana in 2026, the choice between Agave, Jito-Solana, JitoBAM, Frankendancer, and Firedancer measurably determines revenue.

Long: Pre-2024, the validator client was treated as a neutral implementation detail — different clients implemented the same protocol and produced equivalent revenue. Post-2024, with the rise of Jito's Block Engine integration, BAM's TEE-based confidential block construction, Frankendancer's networking-layer performance, and Harmonic's validator-side preference architecture, the choice of validator client has become the largest single revenue lever a Solana validator has. As of March 2026, the stake-weighted distribution was: Agave Jito 32%, Agave JitoBAM 28%, Agave Harmonic 17%, Agave Rakurai 6%, Frankendancer 12%, Firedancer 2%. Roughly 86% of stake on Agave-based clients. On Ethereum, client diversity (Lighthouse, Prysm, Teku, Lodestar, Nimbus) is treated as a network-security property rather than a revenue lever.

First appears: Chapter 5

See also: Agave, Jito tip, Frankendancer, Firedancer, BAM, Harmonic.

Validator-quintile gap

Short: The empirical observation that on Solana in 2026 the per-block priority-fee revenue gap between a top-quintile and bottom-quintile validator is dominated by access factors (client choice and infrastructure relationships) rather than operational factors (uptime, latency, timing-game optimisation), with the access-to-operational ratio approximately an order of magnitude.

Long: Syndica's March 2026 measurement found per-block priority-fee captures of approximately +33% to +101% versus the network median for validators running access-optimised configurations (Frankendancer with Harmonic's Performance or Balanced strategy; Agave Harmonic). Chorus One's August 2025 empirical research found that exhausting every operational optimisation publicly known — block-production timing manipulation plus scheduler optimisation — produces approximately +3% in total rewards (~27 basis points annualised). Placeholder VC's September 2025 measurement of the Solana validator profit Gini coefficient at approximately 0.93 documents the resulting inequality at the aggregate level. The structural argument the book builds: the founding-document framing of the validator as a neutral plumbing role earning rewards proportional to operational excellence has been displaced by a market structure in which the validator's commercial relationships and client choice dominate the revenue distribution.

TradFi analogue: The gap between a top-tier NYSE specialist (with privileged exchange relationships, captive flow, and information rents) and a marginal specialist (operating on the same exchange floor with the same nominal obligations but without the relationships) — a gap that was a function of access rather than craft.

First appears: Chapter 7

See also: Exclusive order flow (on-chain), Validator client, Harmonic, Timing games, Staking Reward Rate.

Votor

Short: Alpenglow's vote-aggregation component, which moves Solana vote transactions off-chain using BLS-aggregated signatures.

Long: Votor is one of two new components introduced by Alpenglow (the other is Rotor). It aggregates validator votes off-chain into compact BLS signatures, which are then verified on-chain. The structural effect is large: vote transactions currently consume approximately 70% of all Solana transactions at peak load, and removing them from blockspace liberates that capacity for fee-paying transactions. Together with Rotor, Votor enables Alpenglow's two-path finality (≥80% in round 1 = ~100ms; ≥60% in round 2 = ~150ms). Mainnet activation is expected in late Q3 or early Q4 2026.

First appears: Chapter 5

See also: Alpenglow, Rotor, Timely Vote Credits (TVC), Validator.

Wallet

Short: The user-facing software that holds the trader's private keys and signs transactions on their behalf.

Long: A wallet is software, not hardware (in the most common case). It holds the user's private keys, displays balances, and produces signed transactions when the user authorises one. Wallets vary in custody model: a self-custody wallet (Phantom, MetaMask, Backpack, Solflare, Rabby) gives the user direct control of the keys; a custodial wallet (the one inside Coinbase or Binance) holds the keys on the user's behalf. The book distinguishes the two only in passing — both produce signed transactions — but the legal and operational responsibilities differ.

TradFi analogue: The brokerage account, except that a self-custody wallet does not have a brokerage between the user and the venue.

First appears: Chapter 1

See also: Trade, Transaction, Settlement.

Glossary ​

Template ​

Term ​

Entries ​

Agave ​

Aggregator ​

Alpenglow ​

AMM (automated market maker) ​

Arbitrage (on-chain) ​

AVS (Actively Validated Service) ​

Assistance Fund (Hyperliquid) ​

Backrun ​

BAM (Block Assembly Marketplace) ​

BAM Node ​

Block ​

Block builder ​

Builder-direct relationship ​

Bundle (Jito-style) ​

Central scheduler (Solana) ​

CEX–DEX arbitrage ​

Chronic loser (on-chain) ​

Consent gap ​

CLOB (central limit order book) ​

Colocation (Solana validator context) ​

Constellation ​

Cross-chain MEV ​

EigenLayer ​

ePBS (enshrined PBS) ​

Exclusive order flow (on-chain) ​

ERC-7683 (cross-chain intent standard) ​

Finality ​

Firedancer ​

Flashbots Protect ​

Flashnet ​

Frankendancer ​

Frontrun ​

Gas ​

Harmonic ​

Health factor ​

HIP-3 (Hyperliquid builder-deployed perpetuals) ​

HIP-4 (Hyperliquid outcome markets) ​

HLP (Hyperliquid liquidity provider vault) ​

HyperBFT ​

HyperCore ​

HyperEVM ​

Impermanent loss (IL) ​

Infrastructure firm (B2B trading infrastructure) ​

Inclusion list / FOCIL (EIP-7805) ​

Intent ​

JIP-24 ​

Jito tip ​

Jupiter Beam ​

Kolibrio ​

Liquidation ​

L2 sequencer ​

Liquidation bonus ​

Liquidity ​

Liquidity provider (LP) ​

Loss-Versus-Rebalancing (LVR) ​

Market maker (MM) ​

Mempool ​

MEV ​

MEV-Blocker ​

MEV-Boost ​

MEV-Share ​

Pack tile (Firedancer) ​

MEV refund ​

Order Flow Auction (OFA) ​

Passive LP ​

PBS / Proposer-Builder Separation ​

Pipeline rebate ​

Priority fee ​

Private order flow ​

Prop-AMM ​

Proposer (Ethereum) ​

Public order flow ​

RFQ (request for quote) ​

RPC provider ​

Relay (Ethereum) ​

Rotor ​

Glossary

Template

Term

Entries

Agave

Aggregator

Alpenglow

AMM (automated market maker)

Arbitrage (on-chain)

AVS (Actively Validated Service)

Assistance Fund (Hyperliquid)

Backrun

BAM (Block Assembly Marketplace)

BAM Node

Block

Block builder

Builder-direct relationship

Bundle (Jito-style)

Central scheduler (Solana)

CEX–DEX arbitrage

Chronic loser (on-chain)

Consent gap

CLOB (central limit order book)

Colocation (Solana validator context)

Constellation

Cross-chain MEV

EigenLayer

ePBS (enshrined PBS)

Exclusive order flow (on-chain)

ERC-7683 (cross-chain intent standard)

Finality

Firedancer

Flashbots Protect

Flashnet

Frankendancer

Frontrun

Gas

Harmonic

Health factor

HIP-3 (Hyperliquid builder-deployed perpetuals)

HIP-4 (Hyperliquid outcome markets)

HLP (Hyperliquid liquidity provider vault)

HyperBFT

HyperCore

HyperEVM

Impermanent loss (IL)

Infrastructure firm (B2B trading infrastructure)

Inclusion list / FOCIL (EIP-7805)

Intent

JIP-24

Jito tip

Jupiter Beam

Kolibrio

Liquidation

L2 sequencer

Liquidation bonus

Liquidity

Liquidity provider (LP)

Loss-Versus-Rebalancing (LVR)

Market maker (MM)

Mempool

MEV

MEV-Blocker

MEV-Boost

MEV-Share

Pack tile (Firedancer)

MEV refund

Order Flow Auction (OFA)

Passive LP

PBS / Proposer-Builder Separation

Pipeline rebate

Priority fee

Private order flow

Prop-AMM

Proposer (Ethereum)

Public order flow

RFQ (request for quote)

RPC provider

Relay (Ethereum)

Rotor