Interface LlmAgentConfig<T>

Configuration for LlmAgent

interface LlmAgentConfig<T> {
    name: string;
    description: string;
    subAgents?: BaseAgent[];
    beforeAgentCallback?: BeforeAgentCallback;
    afterAgentCallback?: AfterAgentCallback;
    model?:
        | string
        | string & {}
        | LanguageModelV2
        | T;
    instruction?: string | InstructionProvider;
    globalInstruction?: string | InstructionProvider;
    tools?: ToolUnion[];
    codeExecutor?: BaseCodeExecutor;
    disallowTransferToParent?: boolean;
    disallowTransferToPeers?: boolean;
    includeContents?: "default" | "none";
    outputKey?: string;
    planner?: BasePlanner;
    plugins?: BasePlugin[];
    memoryService?: BaseMemoryService;
    sessionService?: BaseSessionService;
    artifactService?: BaseArtifactService;
    userId?: string;
    appName?: string;
    generateContentConfig?: GenerateContentConfig;
    inputSchema?: ZodType<unknown, unknown, $ZodTypeInternals<unknown, unknown>>;
    outputSchema?: ZodType<unknown, unknown, $ZodTypeInternals<unknown, unknown>>;
    beforeModelCallback?: BeforeModelCallback;
    afterModelCallback?: AfterModelCallback;
    beforeToolCallback?: BeforeToolCallback;
    afterToolCallback?: AfterToolCallback;
}

Type Parameters

Properties

name description subAgents? beforeAgentCallback? afterAgentCallback? model? instruction? globalInstruction? tools? codeExecutor? disallowTransferToParent? disallowTransferToPeers? includeContents? outputKey? planner? plugins? memoryService? sessionService? artifactService? userId? appName? generateContentConfig? inputSchema? outputSchema? beforeModelCallback? afterModelCallback? beforeToolCallback? afterToolCallback?

Properties

name

name: string

Name of the agent

description

description: string

Description of the agent

OptionalsubAgents

subAgents?: BaseAgent[]

Sub-agents that this agent can delegate to

OptionalbeforeAgentCallback

beforeAgentCallback?: BeforeAgentCallback

Callback or list of callbacks to be invoked before the agent run

OptionalafterAgentCallback

afterAgentCallback?: AfterAgentCallback

Callback or list of callbacks to be invoked after the agent run

Optionalmodel

model?:
    | string
    | string & {}
    | LanguageModelV2
    | T

The LLM model to use When not set, the agent will inherit the model from its ancestor

Optionalinstruction

instruction?: string | InstructionProvider

Instructions for the LLM model, guiding the agent's behavior

OptionalglobalInstruction

globalInstruction?: string | InstructionProvider

Instructions for all the agents in the entire agent tree ONLY the global_instruction in root agent will take effect

Optionaltools

tools?: ToolUnion[]

Tools available to this agent

OptionalcodeExecutor

codeExecutor?: BaseCodeExecutor

Code executor for this agent

OptionaldisallowTransferToParent

disallowTransferToParent?: boolean

Disallows LLM-controlled transferring to the parent agent

OptionaldisallowTransferToPeers

disallowTransferToPeers?: boolean

Disallows LLM-controlled transferring to the peer agents

OptionalincludeContents

includeContents?: "default" | "none"

Whether to include contents in the model request

OptionaloutputKey

outputKey?: string

The output key in session state to store the output of the agent

Optionalplanner

planner?: BasePlanner

Instructs the agent to make a plan and execute it step by step

Optionalplugins

plugins?: BasePlugin[]

Extend or intercept the agent’s behavior. Each plugin may implement callbacks for model calls, tool calls, events, and agent lifecycle.

OptionalmemoryService

memoryService?: BaseMemoryService

Memory service for long-term storage and retrieval

OptionalsessionService

sessionService?: BaseSessionService

Session service for managing conversations

OptionalartifactService

artifactService?: BaseArtifactService

Artifact service for file storage and management

OptionaluserId

userId?: string

User ID for the session

OptionalappName

appName?: string

Application name

OptionalgenerateContentConfig

generateContentConfig?: GenerateContentConfig

Additional content generation configurations NOTE: not all fields are usable, e.g. tools must be configured via tools, thinking_config must be configured via planner in LlmAgent.

OptionalinputSchema

inputSchema?: ZodType<unknown, unknown, $ZodTypeInternals<unknown, unknown>>

The input schema when agent is used as a tool

OptionaloutputSchema

outputSchema?: ZodType<unknown, unknown, $ZodTypeInternals<unknown, unknown>>

The output schema when agent replies NOTE: when this is set, agent can ONLY reply and CANNOT use any tools

OptionalbeforeModelCallback

beforeModelCallback?: BeforeModelCallback

Callback or list of callbacks to be called before calling the LLM

OptionalafterModelCallback

afterModelCallback?: AfterModelCallback

Callback or list of callbacks to be called after calling the LLM

OptionalbeforeToolCallback

beforeToolCallback?: BeforeToolCallback

Callback or list of callbacks to be called before calling a tool

OptionalafterToolCallback

afterToolCallback?: AfterToolCallback

Callback or list of callbacks to be called after calling a tool

Settings

Member Visibility

On This Page

Properties