Agent
data class Agent(val name: String, val provider: String, val version: <Error class: unknown class> = Semver(), val description: String, val conditions: Set<Condition> = emptySet(), val actions: List<Action>, val goals: Set<Goal>, val stuckHandler: StuckHandler? = null, val opaque: Boolean = false, val domainTypes: Collection<DomainType> = mergeTypes(
agentName = name,
defaultDataTypes = emptyList(),
actions = actions,
)) : AgentScope
An agent defines a set of actions and conditions that enable planning.
Parameters
name
The name of the agent.
provider
The provider of the agent.
version
The version of the agent. Defaults to 0.1.0
description
A description of the agent. Required
goals
The goals the agent can achieve
stuck Handler
The handler to call when the agent is stuck, if provided
conditions
Well-known conditions that can be referenced by actions
actions
The actions the agent can use
opaque
whether to hide the agent's actions and conditions from the outside world, defaults to false.
domain Types
Data types used in this agent
Constructors
Link copied to clipboard
constructor(name: String, provider: String, version: String, description: String, goals: Set<Goal>, actions: List<Action>, conditions: Set<Condition> = emptySet(), stuckHandler: StuckHandler? = null)
constructor(name: String, provider: String, version: <Error class: unknown class> = Semver(), description: String, conditions: Set<Condition> = emptySet(), actions: List<Action>, goals: Set<Goal>, stuckHandler: StuckHandler? = null, opaque: Boolean = false, domainTypes: Collection<DomainType> = mergeTypes(
agentName = name,
defaultDataTypes = emptyList(),
actions = actions,
))
Properties
Functions
Link copied to clipboard
Creates a transformation action from an agent
Link copied to clipboard
Create a new agent from the given scope
Link copied to clipboard
Link copied to clipboard
Return a version of the agent with actions and conditions pruned to the given pruned planning system.
Link copied to clipboard
Link copied to clipboard
Return a version of the agent with the single goal