Agent
The Agent class is the foundation for creating AI agents in Mastra. It provides methods for generating responses, streaming interactions, and handling voice capabilities.
Importing
import { Agent } from "@mastra/core/agent";Constructor
Creates a new Agent instance with the specified configuration.
constructor(config: AgentConfig<TAgentId, TTools, TMetrics>)Parameters
name:
string
Unique identifier for the agent.
instructions:
string | ({ runtimeContext: RuntimeContext }) => string | Promise<string>
Instructions that guide the agent's behavior. Can be a static string or a function that returns a string.
model:
MastraLanguageModel | ({ runtimeContext: RuntimeContext }) => MastraLanguageModel | Promise<MastraLanguageModel>
The language model to use for generating responses. Can be a model instance or a function that returns a model.
tools?:
ToolsInput | ({ runtimeContext: RuntimeContext }) => ToolsInput | Promise<ToolsInput>
Tools that the agent can use. Can be a static object or a function that returns tools.
defaultGenerateOptions?:
AgentGenerateOptions
Default options to use when calling generate().
defaultStreamOptions?:
AgentStreamOptions
Default options to use when calling stream().
workflows?:
Record<string, NewWorkflow> | ({ runtimeContext: RuntimeContext }) => Record<string, NewWorkflow> | Promise<Record<string, NewWorkflow>>
Workflows that the agent can execute. Can be a static object or a function that returns workflows.
evals?:
Record<string, Metric>
Evaluation metrics for assessing agent performance.
memory?:
MastraMemory
Memory system for the agent to store and retrieve information.
voice?:
CompositeVoice
Voice capabilities for speech-to-text and text-to-speech functionality.