Intent-Based

Intent-based design flips transactions upside down: instead of specifying how ("exact contract calls, exact route, exact slippage"), you declare what you want ("swap 1 ETH for at least $3,000 USDC") and competing solvers race to fill it under the best available conditions. UniswapX, Across, and CowSwap pioneered the pattern in production. How it works An intent is a signed statement of outcome, not a recipe: You sign the goal — e.g., "receive 3,000 USDC for exactly 1 ETH, any route, before this deadline, minimum acceptable rate X." Solvers fill it — specialized actors with inventory, liquidity, and routing models compete to execute the outcome best; the winner keeps the spread as profit, so competition pushes the price up to the user's favor. Dutch auctions tune the price — UniswapX-style fillers bid downward until someone is willing to execute; the auction discovers the optimal fill instead of you paying a fixed wedge. Settlement — fillers execute on-chain (transfers, swaps, pre-signed fills) and you receive the outcome; MEV that would have sandwiched your direct trade is partly recaptured as price improvement because solvers bid to win your flow. Because the user declares outcomes rather than mechanism, the marketplace — not the chain's mempool — becomes where quality of execution is decided. Why it matters for decentralization Intents shift power from block infrastructure to execution marketplaces: value moves from "who orders blocks" (validators, searchers) to "who fills orders" (solvers). That can return value to users — searchable competition beats routine MEV extraction — but it also introduces a new centralization variable: solver oligopoly. A market with three dominant fillers is the new version of mining-pool concentration, and exclusive order-flow deals can quietly recreate gatekeepers. Under the capital and software pillars of the W3D model, filler diversity and outcome transparency are the metrics; the same way we audit validator sets, we should audit who is allowed to fill an intent and who can see the order flow. Example: from route anxiety to one signed outcome Today on Arbitrum or Ethereum, a user wants to convert ETH to a stablecoin across several venues. In an intent-based swap, they sign "give me EURC-equivalent, any venue, within 30 seconds, at ≥ rate X" — multiple solvers bid, a Dutch auction converges, and the user receives the fill at a price better than a single-route trade would have found. The flow is already live in UniswapX, CowSwap, and Across-adjacent bridges: same wallet, far fewer intermediaries, more of the MEV sandwich recaptured as price improvement. On Base, the same pattern powers consumer-friendly cross-chain swaps that settle like same-chain ones. Risks & limitations Solver oligopoly: if a few fillers see the fullest order flow, competition thins and fills silently worsen. Opaque execution: users verify outcomes, not paths — great for UX, dangerous if filler incentives diverge (partial fills hidden in fine print). Order-flow deals: exclusive routing agreements can hand fillers advantages and users worse prices than the open marketplace would produce. Complexity escapes auditing: multi-venue, multi-chain fills are harder to verify than a single contract call; outcome guarantees matter more than mechanism promises. The intent landscape is real and growing, and understanding the standards helps separate hype from production-ready design: ERC-7683 (cross-chain intents): a standard for intent-based cross-chain settlement, co-designed by Uniswap and Across, that defines how a user signs a cross-chain goal and how fillers prove they executed it. This is the standard behind "instant bridge" UX on Arbitrum, Base, and Optimism. MEV-share and MEV recapture: intent marketplaces like Flashbots MEV-Share let users capture some of the MEV their trades generate by directing it back as price improvement. The solver network competes not just on fill price but on how much MEV it returns — a measurable, auditable metric. Filler economics: solvers earn the spread between what the user signed and what they executed. In a competitive market, this spread shrinks to near-zero; in practice, exclusive order-flow deals can inflate it. Measuring the spread across a week of fills is the most direct audit of marketplace health. Intent gas and verification: intents require on-chain verification (proofs, signatures, settlement logic), which costs gas. Intent-aware chains (Essential, Anoma) optimize for this by native verification; on Ethereum, intents settle through contract logic on top of the same EVM. The decentralization question for intents is the same as for block builders: who sees the order flow, and who can fill it? If the answer is "a few professional solvers with exclusive deals," the market is centralizing. If the answer is "anyone with capital and a competitive model," the market is healthy. The capital pillar of the W3D model treats solver-set diversity as a measurable metric, just as it treats validator-set diversity. Anatomy of one intent fill Trace a cross-chain intent through ERC-7683 to see where the trust rests: Sign: you sign an intent order defining the input (1 ETH) and the outcome (USDC on Arbitrum before a deadline, ≥ a minimum rate). Publish: the order posts to an intent marketplace (UniswapX's fill contract or Across's relayers), visible to all fillers. Fill: a solver locks liquidity on the destination side — they bridge or self-provide — so your outcome is real before you lose the input. Verify: the on-chain contract checks the covenant (deadline, minimum rate, correct token pair); a filler that fails pays penalties, not the user. Settle: input releases to the filler only after outcome verification, and any MEV recapture from the auction returns to you as price improvement. Because the user's asset is released only against a verified outcome, filler failure risk stays with the filler — the intent model converts "trust the router" into "trust the verification contract," which anyone can audit. Filler-set diversity then becomes the free-market health check for the whole marketplace. Frequently asked questions Intents vs limit orders — what is the difference? A limit order pins price on one venue; an intent specifies an outcome across all venues and asks the marketplace to fill it best. The intent is strictly more flexible. Do intents fix MEV? They redirect it: solvers compete to give you better prices instead of sandwiching you. The equilibrium improves, though the players are often the same sophisticated searchers. What can go wrong with intents? Solver failure mid-fill, exclusive-flow centralization, and users approving intents they do not understand. The safeguard is clear outcome guarantees — read the guarantee, not the mechanism. Sources & methodology UniswapX whitepaper — Dutch-auction intent fills. Across intent docs — proposal-based cross-chain intents. W3D methodology + academy dataset. Related terms MEV · DEX aggregator · DeFi · Liquidity · Slippage Chain audits: Ethereum · Arbitrum · tool: Nakamoto coefficient calculator