Intent-based execution
Intent-based execution is the design pattern where users define desired outcomes and constraints, and the system determines the optimal path to fulfillment.
What it refers to
Intent-based execution is a design approach where the user specifies what they want to happen, not how it should happen.
Instead of constructing a sequence of transactions across networks, the user submits an intent: a structured declaration of the desired outcome (such as "swap 100 USDC on Base for ETH on Arbitrum") along with constraints (minimum amount received, expiration time, acceptable slippage).
The system, specifically the Solver, determines the execution path. It evaluates liquidity, gas costs, bridge routes, and network conditions to find the most reliable way to fulfill the intent.
This pattern is distinct from transaction-based execution, where the user or application must specify every step.
Why this concept exists
Cross-network financial actions are inherently complex. A single swap across two networks may involve bridging, local swaps, liquidity sourcing, gas management, and settlement coordination.
In transaction-based models, applications must embed all of this logic. They must know which bridges to use, where liquidity exists, how to handle failures mid-execution, and how to manage timing across asynchronous networks.
Intent-based execution (often searched as intent based trading or crypto intents) shifts this burden from the application to the infrastructure. The application defines the outcome. The infrastructure handles the complexity.
This is particularly important in multi-network environments where:
- Execution paths change based on real-time conditions
- Liquidity is fragmented across venues
- Optimal routing requires information the application may not have
- Failure handling needs to be coordinated across networks
What this changes for system design
Intent-based execution requires systems to:
- Separate outcome definition from execution planning
- Build solver infrastructure that evaluates and selects execution paths dynamically
- Handle fulfillment verification, ensuring intents are only settled when constraints are met
- Support intent expiration and recovery when fulfillment is not possible
- Design settlement processes that work asynchronously across networks
The pattern trades user-side complexity for system-side intelligence. It is a foundation for making cross-network actions feel as simple as local ones.
