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

Package-level declarations

Types

class AIAgentEdge<IncomingOutput, OutgoingInput>

Represents a directed edge connecting two nodes in the graph of an AI agent strategy. This edge facilitates the transmission of data between a source node and a destination node, allowing transformation or filtering of output values from the source node before they reach the destination node.

AIAgentNode

open class AIAgentNode<Input, Output> : AIAgentNodeBase<Input, Output>

Represents a simple implementation of an AI agent node, encapsulating a specific execution logic that processes the input data and produces an output.

AIAgentNodeBase

abstract class AIAgentNodeBase<Input, Output>

Represents an abstract node in an AI agent strategy graph, responsible for executing a specific operation and managing directed edges to other nodes.

AIAgentStateManager

class AIAgentStateManager

Manages the state of an AI agent by providing thread-safe access and mechanisms to update the internal state using a locking mechanism.

AIAgentStorage

class AIAgentStorage

Concurrent-safe key-value storage for an agent. You can create typed keys for your data using the createStorageKey function and set and retrieve data using it by calling set and get.

AIAgentStorageKey

data class AIAgentStorageKey<T : Any>(val name: String)

Represents a storage key used for identifying and accessing data associated with an AI agent.

AIAgentStrategy

class AIAgentStrategy<Input, Output>(val name: String, val nodeStart: StartNode<Input>, val nodeFinish: FinishNode<Output>, toolSelectionStrategy: ToolSelectionStrategy, serializer: Json = Json { prettyPrint = true }) : AIAgentSubgraph<Input, Output>

Represents a strategy for managing and executing AI agent workflows built as subgraphs of interconnected nodes.

AIAgentSubgraph

open class AIAgentSubgraph<Input, Output>(val name: String, val start: StartNode<Input>, val finish: FinishNode<Output>, toolSelectionStrategy: ToolSelectionStrategy, llmModel: LLModel? = null, llmParams: LLMParams? = null) : AIAgentNodeBase<Input, Output> , ExecutionPointNode

AIAgentSubgraph represents a structured subgraph within an AI agent workflow. It serves as a logical segment containing a defined starting and ending point. The subgraph is responsible for executing tasks in a step-by-step manner, managing iterations, and handling tool selection strategies.

ExecutionPoint

data class ExecutionPoint(val node: AIAgentNodeBase<*, *>, val input: Any? = null)

Represents a point of execution within the AI agent's strategy graph. An execution point consists of a specific node and an optional input value.

ExecutionPointNode

interface ExecutionPointNode

Represents a node in the execution graph of an AI agent that can explicitly enforce execution at a specified node with optional input data. This interface provides the ability to define a forced node and input, overriding default execution behavior.

FinishNode

class FinishNode<Output> : AIAgentNode<Output, Output>

Represents a specialized node within an AI agent strategy graph that marks the endpoint of a subgraph. This node serves as a "finish" node and directly passes its input to its output without modification.

StartNode

class StartNode<Input> : AIAgentNode<Input, Input>

Represents the base node for starting a subgraph in an AI agent strategy graph. This node acts as an entry point for executing subgraphs.

SubgraphMetadata

data class SubgraphMetadata(val nodesMap: Map<String, AIAgentNodeBase<*, *>>, val uniqueNames: Boolean = false)

Represents metadata associated with a subgraph in an AI agent strategy graph.

ToolSelectionStrategy

sealed interface ToolSelectionStrategy

Represents a strategy to select a subset of tools to be used in a subgraph during its execution.

Functions

createStorageKey

inline fun <T : Any> createStorageKey(name: String): AIAgentStorageKey<T>

Creates a storage key for a specific type, allowing identification and retrieval of values associated with it.