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

FunctionalAgentBuilder

class FunctionalAgentBuilder<Input, Output>(strategy: AIAgentFunctionalStrategy<Input, Output>, promptExecutor: PromptExecutor? = null, toolRegistry: ToolRegistry = ToolRegistry.EMPTY, id: String? = null, config: AIAgentConfig, clock: Clock = Clock.System, featureInstallers: MutableList<FunctionalAIAgent.FeatureContext.() -> Unit> = mutableListOf())(source)

A builder class for constructing instances of FunctionalAIAgent with customizable configuration.

This builder simplifies the configuration process by providing a fluent API to set various parameters for the FunctionalAIAgent, including its behavior strategy, prompt details, model settings, tool registry, and additional features. The builder enforces the presence of required configurations and allows the addition of optional parameters to tailor the agent's functionality.

Parameters

Input

The type of input that the resulting AI agent will process.

Output

The type of output that the resulting AI agent will produce.

Constructors

FunctionalAgentBuilder

constructor(strategy: AIAgentFunctionalStrategy<Input, Output>, promptExecutor: PromptExecutor? = null, toolRegistry: ToolRegistry = ToolRegistry.EMPTY, id: String? = null, config: AIAgentConfig, clock: Clock = Clock.System, featureInstallers: MutableList<FunctionalAIAgent.FeatureContext.() -> Unit> = mutableListOf())

Functions

agentConfig

fun agentConfig(config: AIAgentConfig): FunctionalAgentBuilder<Input, Output>

Configures the FunctionalAgentBuilder with the provided AI agent configuration settings.

build

fun build(): AIAgent<Input, Output>

Builds and returns an instance of AIAgent<Input, Output> based on the current configuration of the FunctionalAgentBuilder. This method ensures that all required fields are set, and applies any configured feature installers to the agent.

fun id(id: String?): FunctionalAgentBuilder<Input, Output>

Sets the unique identifier for the FunctionalAgent.

install

fun <TConfig : FeatureConfig> install(feature: AIAgentFunctionalFeature<TConfig, *>, configure: ConfigureAction<TConfig>): FunctionalAgentBuilder<Input, Output>

Installs and configures a given feature into the functional agent builder.

llmModel

fun llmModel(model: LLModel): FunctionalAgentBuilder<Input, Output>

Configures the functional agent builder to use a specific Large Language Model (LLM) for processing.

maxIterations

fun maxIterations(maxIterations: Int): FunctionalAgentBuilder<Input, Output>

Sets the maximum number of iterations for the functional agent builder.

numberOfChoices

fun numberOfChoices(numberOfChoices: Int): FunctionalAgentBuilder<Input, Output>

Configures the number of choices to be considered during the functional agent's operations.

prompt

fun prompt(prompt: Prompt): FunctionalAgentBuilder<Input, Output>

Sets the specified prompt for the functional agent being built.

promptExecutor

fun promptExecutor(promptExecutor: PromptExecutor): FunctionalAgentBuilder<Input, Output>

Sets the provided PromptExecutor instance for the current FunctionalAgentBuilder.

systemPrompt

fun systemPrompt(systemPrompt: String): FunctionalAgentBuilder<Input, Output>

Sets the system-level prompt for the agent being built. The system-level prompt provides contextual instructions to the agent.

temperature

fun temperature(temperature: Double): FunctionalAgentBuilder<Input, Output>

Sets the temperature parameter for the functional agent. Temperature controls the randomness of the agent's outputs, where higher values result in more diverse outputs and lower values result in more focused outputs.

toolRegistry

fun toolRegistry(toolRegistry: ToolRegistry): FunctionalAgentBuilder<Input, Output>

Sets the tool registry to be used by the FunctionalAgentBuilder.