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

FunctionalAgentServiceBuilder

class FunctionalAgentServiceBuilder<Input, Output>(source)

A builder class for constructing a FunctionalAIAgentService, enabling a fluent configuration style.

This class allows customization of various parameters and components required for the AI agent service, including strategies, prompts, models, tool registries, and other operational settings.

Parameters

Input

The type of data input to the AI agent during execution.

Output

The type of data output by the AI agent after processing.

Functions

agentConfig

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

Configures the FunctionalAgentServiceBuilder with the given AI agent configuration.

build

fun build(): FunctionalAIAgentService<Input, Output>

Builds and returns a configured instance of FunctionalAIAgentService.

install

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

Installs the specified feature into the functional service builder and applies the provided configuration.

llmModel

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

Sets the specified Large Language Model (LLM) to be used by the Functional Service.

maxIterations

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

Sets the maximum number of iterations to use for the service.

numberOfChoices

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

Sets the number of choices for the service builder configuration.

prompt

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

Assigns the given prompt to the builder configuration.

promptExecutor

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

Sets the PromptExecutor for this FunctionalServiceBuilder. The PromptExecutor is responsible for processing prompts during the execution of the functional AI service being built.

systemPrompt

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

Configures a system-level prompt message for the functional service builder. This method sets the prompt for the agent, providing necessary system context or instructions that guide the behavior of the language model.

temperature

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

Sets the temperature parameter for the functional service builder. The temperature controls the randomness of the model's output, where a lower value makes the output more deterministic, and a higher value increases randomness.

toolRegistry

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

Sets the tool registry to be used by the FunctionalServiceBuilder.