agents-planner/ai.koog.agents.planner.goap/GOAPPlanner

GOAPPlanner

open class GOAPPlanner<State> : AIAgentPlanner<State, GOAPPlan<State>> (source)

Goal-Oriented Action Planning (GOAP) implementation for AI agents.

GOAP is a planning system that uses goals, actions with preconditions and effects, and a search algorithm to find the optimal sequence of actions to achieve a goal.

Parameters

State

The type of the state.

actions

The list of defined actions.

goals

The list of defined goals.

Properties

stateType

val stateType: KType

KType of the State.

Functions

execute

suspend fun execute(context: AIAgentFunctionalContext, input: State): State

Executes the main loop for the planner, which involves building and executing plans iteratively until the plan is considered successfully completed or a max number of iterations is reached.