agents-core/ai.koog.agents.core.agent.entity/AIAgentGraphStrategyBase

AIAgentGraphStrategyBase

open class AIAgentGraphStrategyBase<TInput, TOutput>(val name: String, val nodeStart: StartNode<TInput>, val nodeFinish: FinishNode<TOutput>, toolSelectionStrategy: ToolSelectionStrategy, serializer: Json = Json { prettyPrint = true }) : AIAgentStrategy<TInput, TOutput, AIAgentGraphContextBase> , AIAgentSubgraphBase<TInput, TOutput> (source)

Base class for AIAgentGraphStrategy.

Inheritors

AIAgentGraphStrategy

Constructors

AIAgentGraphStrategyBase

constructor(name: String, nodeStart: StartNode<TInput>, nodeFinish: FinishNode<TOutput>, toolSelectionStrategy: ToolSelectionStrategy, serializer: Json = Json { prettyPrint = true })

Properties

edges

var edges: List<AIAgentEdge<TOutput, *>>

The directed edges connecting this node to other nodes in the AI agent strategy graph. Each edge defines the flow and optional transformation of output data from this node to another.

finish

val finish: FinishNode<TOutput>

val id: String

Represents the unique identifier of the AI agent node.

inputType

open override val inputType: TypeToken

The TypeToken of the TInput

metadata

lateinit var metadata: SubgraphMetadata

Represents the metadata of the subgraph associated with the AI agent strategy.

name

open override val name: String

The unique identifier for the strategy.

nodeFinish

val nodeFinish: FinishNode<TOutput>

The finishing node of the strategy, marking the subgraph's endpoint.

nodeStart

val nodeStart: StartNode<TInput>

The starting node of the strategy, initiating the subgraph execution. By default, the start node receives the agent input and passes it through unchanged to the next node.

outputType

open override val outputType: TypeToken

The TypeToken of the TOutput

start

val start: StartNode<TInput>

Functions

addEdge

open fun addEdge(edge: AIAgentEdge<TOutput, *>)

Adds a directed edge from the current node, enabling connections between this node and other nodes in the AI agent strategy graph.

enforceExecutionPoint

open override fun enforceExecutionPoint(node: AIAgentNodeBase<*, *>, input: Any? = null)

Enforces execution to start at the specified node with the given input, overriding the default execution flow.

execute

open suspend override fun execute(context: AIAgentGraphContextBase, input: TInput): TOutput?

Executes the AI agent's strategy using the provided context and input.

executeUnsafe

suspend fun executeUnsafe(context: AIAgentGraphContextBase, input: Any?): Any?

Executes the node operation using the provided execution context and input, bypassing static type checks. This method internally performs an unchecked cast of input to TInput and delegates to execute. The caller is responsible for ensuring that input is actually assignable to TInput; otherwise a ClassCastException may be thrown at the call site or later during execution.

forwardTo

infix fun <OutgoingInput> forwardTo(otherNode: AIAgentNodeBase<OutgoingInput, *>): AIAgentEdgeBuilderIntermediate<TOutput, TOutput, OutgoingInput>

Creates a directed edge from this AIAgentNodeBase to another AIAgentNodeBase, allowing data to flow from the output of the current node to the input of the specified node.

getExecutionPoint

open override fun getExecutionPoint(): ExecutionPoint?

Retrieves the current execution point, which consists of a specific node in the execution graph and an optional input. If no forced node is defined, the method returns null.

resetExecutionPoint

open override fun resetExecutionPoint()

Resets the currently enforced execution point for the AI agent, including clearing any forced node and input data. Once the execution point is reset, the system will revert to its default execution behavior without targeting a specific node explicitly.

resolveEdge

suspend fun resolveEdge(context: AIAgentGraphContextBase, nodeOutput: TOutput): AIAgentNodeBase.ResolvedEdge?

Resolves the edge associated with the provided node output and execution context. Iterates through available edges and identifies the first edge that can successfully process the given node output within the provided context. If a resolvable edge is found, it returns a ResolvedEdge containing the edge and its output. Otherwise, returns null.

setExecutionPointAfterNode

suspend fun setExecutionPointAfterNode(nodePath: String, output: JSONElement, agentContext: AIAgentGraphContextBase)

Enforces the strategy's next execution to continue right after the node identified by nodePath, using output as that node's completed output.