Scriptable Transactions

Scripting any user intent within the transaction

Smart accounts are an inefficient way to define the user’s intent logic (what the user wants from an app). Instead, Intentify let developers define their user’s intent within the transaction itself orchestrated by his EOA via one signature. This reduces gas costs and development complexity while providing a better UX:

Gas-optimized: Reduces gas overhead by ~90%.
Simple: Customizes the user’s intent logic without contract deployments.
Seamless: Let user express any intent using one signature from his EOA.

Scriptable transactions bring the 1-click UX of web2 apps onchain. Using an intuitive language, developers can script any form of intent logic for users in minutes – providing them what they want via 1-click.

Intent scripts are represented by workflows which define the onchain actions performed from the user’s perspective to fulfill his intent. There are two classes of workflows: monolithic and multi-party.

Monolithic Workflows

Monolithic workflows predefine all parameters of the onchain actions required to fulfill an intent. They can express logic of arbitrary complexity by combining the following building blocks:

Chained multicalls: Special type of a multicall where the output of one transaction (unknown upfront) can be used as input for the next one (e.g. swap ETH for USDC and supply the received USDC amount to a lending protocol). Currently, this type of multi-step operation requires a custom contract deployment.
Conditions: Adding conditional logic to workflows allows to check any onchain state before making a functional call to enforce constraints.
Automations: Adding time- or event-based conditions to workflows allows to perform any type of recurring actions for users enabling web2-like app experiences.

Based on the specific app use case, monolithic workflows can be either executed ad-hoc or signed to be executed at a later date (e.g. when certain conditions are satisfied). The main use cases for monolithic workflows are gas sponsorship, chain abstraction, chained multicalls and transaction automations.

Multi-Party Workflows (Super Sessions)

In addition to predefined workflows, Intentify supports multi-party workflows where intent logic can include free parameters to be provided by a third party. The user signs a partial workflow that defines incomplete onchain actions along with specific constraints. To complete the workflow and make it executable, a third party must sign a complementary workflow that provides the missing parameters while satisfying the defined constraints.

There are different types of constraints which can be enforced as conditions (within a workflow) providing an individual degree of freedom for the counterparty:

Amount: Maximum/minimum token amount the user is willing to spend/receive.
Time: Certain time frame or frequency for a valid workflow execution.
Target: Contracts (functions) that can be called within the workflow.
Operator: Operator allowed to provide missing parameters.
Values: Parameter value ranges acceptable for specific calls.

We define multi-party workflows as Super Sessions which can be seen as a novel and enhanced version of session keys enabling new onchain consumer app experiences.

By signing workflows with free parameters, users can delegate custom permission logic to an app enabling it to perform actions on their behalf (while never holding their assets or private keys). The permissions act as guardrails granting an individual degree of freedom to complete the missing parameters and execute the workflows (only executable if all constraints are satisfied). Like workflows themselves, the permission logic can be arbitrarily complex extending the capabilities of session keys.

Solvers

Beyond performing the onchain actions by themselves, app developers can also outsource (parts of) the parameter completion to incentivized third-parties acting on behalf of their users. This way, they can effectively outsource the necessary computation parts of their apps to sophisticated solvers reducing their development effort and creating better outcomes for their users.

Autonomous agents

Extended further, multi-party workflows enable to build AI agents which are financially autonomous while at the same time being controlled by pre-defined guardrails signed by users. By giving AI agents custom permissions, they can operate autonomously on the user’s behalf while being controlled by financial guardrails. These guardrails ensure that AI agents, equipped with their own private keys but without holding user funds, operate strictly within a user-defined action space by completing the parameters of partial workflows and satisfying the pre-defined constraints.