For the complete documentation index, see llms.txt. This page is also available as Markdown.

Intents, Solvers and Markets

1. What is an Intent?

An intent is an outcome‑predicated expression over future settlement transactions: a declaration that encodes invariants about state transitions while remaining agnostic to how they are accomplished.

Think of it as a goal plus guardrails: what must be true, not the step‑by‑step recipe.

Analogy: “Get me from NYC to Paris tomorrow in < 8 hours for under $800.” The traveller constrains price, timing, and destination; the planner optimizes the itinerary across airlines, layovers, insurance, and ground transport.

2. What is a Solver?

A solver is an off‑chain agent or service that:

  1. Discovers open intents expressed by users.

  2. Generates a concrete execution plan that satisfies the intent’s outcome invariants.

  3. Executes a on-chain transaction that convinces the on-chain settlement program that the user's intents are satisfied.

  4. Receives payment only if the settlement transaction is accepted.

Put simply, solvers turn “what” into “how”, but are only rewarded when their how is executed and verified as correct.

3. What is a Market?

Within an intent‑driven system, a market is an on‑chain protocol defined by three pillars:

  • Intent standard – the schema every intent must follow.

  • Market structure – how intent publishers and solvers discover each other, how solutions are matched with intents, and ranked to be cleared.

  • Settlement conditions – the on-chain verification requirements that release funds once an intent is fulfilled.

A market may expose multiple entry points: on‑chain, off‑chain, or a combination, depending on its structure. Simple, first‑come‑first‑served or fixed‑price venues can operate entirely on‑chain, whereas more sophisticated formats such as auctions or RFQs blend on‑chain settlement with off‑chain negotiation to balance order‑flow visibility, fairness, censorship resistance and cost.

Terminology is vantage‑point dependent. Intent publishers experience a solver market**:** a network of specialized agents competing to satisfy their request. Solvers observe an intent market: a live order book of opportunities priced by attached incentives. Economically and technically, they are two faces of the same protocol primitive.

4. Outcome Markets

Intent markets are best understood as outcome markets. Instead of treating workflows as ephemeral sequences of synchronous executions (a fragile "call graph"), outcome markets make the intent or task a first-class object. Each task is an instance of a fixed state machine that outsources execution, with typed fields (what must be true at settlement) and a clear lifecycle (published → solved → settled / expired / cancelled).

Coordination over Execution

This shift is critical for use cases where the hard part is not execution, but coordination: discovering who can do the job, selecting the best executor, and settling reliably when work spans time, chains, or off-chain systems.

With typed orders and fixed state transitions, integrators get a deterministic interface for "what success means," rather than encoding domain knowledge in bespoke calldata. Solvers and off-chain agents benefit because the work surface is normalized: tasks are legible, comparable, and automatable. The result is a market that can continuously optimize (price, latency, reliability) because many executors can compete over the same well-defined object.

Async by Default

Outcome markets unlock a cleaner intent-driven, escrow-based, asynchronous flow. Instead of an upstream integrator orchestrating a brittle sequence of contract calls, where every step is a failure mode, the user escrows resources and publishes intents against specific outcomes. Executing agents can take the work, acquire inventory, run off-chain computation, and coordinate across systems before settling back into the task’s state machine. This "async by default" posture makes blockchain-adjacent automation viable, as real coordination often depends on external timing, liquidity movement and retries.

Scaling by Composition

The real unlock for service or agent-heavy ecosystems is that outcome markets provide scalable semantics through composition. Agents can compose tasks across markets (e.g., acquire inventory → execute route → deliver outcome) without requiring bespoke bilateral integrations between every upstream app and every downstream executor.

As new solvers, strategies, and infrastructure appear, they plug into the same task types and market interfaces; integrators don’t need to be redeployed to benefit. This is semantic scalability: the system grows by adding new markets and solver capabilities, not by multiplying one-off integration code paths. This creates a future-proof integration surface: integrators publish "what I want," and any compliant agent can discover, compete, and execute.

Last updated