Skip to Content
ReferenceAgentsgetWorkflows()

Agent.getWorkflows()

The getWorkflows() method retrieves the workflows configured for an agent, resolving them if they’re a function. These workflows enable the agent to execute complex, multi-step processes with defined execution paths.

Syntax

getWorkflows({ runtimeContext = new RuntimeContext() }: { runtimeContext?: RuntimeContext } = {}): Record<string, NewWorkflow> | Promise<Record<string, NewWorkflow>>

Parameters


runtimeContext?:

RuntimeContext
Runtime context for dependency injection and contextual information.

Return Value

Returns a Record<string, NewWorkflow> object or a Promise that resolves to a Record<string, NewWorkflow> object containing the agent’s workflows.

Description

The getWorkflows() method is used to access the workflows that an agent can execute. It resolves the workflows, which can be either directly provided as an object or returned from a function that receives runtime context.

Examples

import { Agent } from "@mastra/core/agent";
import { createWorkflow, createStep } from "@mastra/core/workflows/vNext";
import { openai } from "@ai-sdk/openai";
import { z } from "zod";
 
const generateSuggestionsStep = createStep({
  id: "generate-suggestions",
  inputSchema: z.object({
    topic: z.string().describe("The topic to research"),
  }),
  outputSchema: z.object({
    summary: z.string(),
  }),
  execute: async ({ inputData, mastra }) => {
    const researchAgent = mastra?.getAgent("researchAgent");
 
    if (!researchAgent) {
      throw new Error("Research agent is not initialized");
    }
 
    const { topic } = inputData;
 
    const result = await researchAgent.generate([
      { role: "assistant", content: topic },
    ]);
 
    return { summary: result.text };
  },
});
 
const researchWorkflow = createWorkflow({
  id: "research-workflow",
  inputSchema: z.object({
    topic: z.string().describe("The topic to research"),
  }),
  outputSchema: z.object({
    summary: z.string(),
  }),
});
 
researchWorkflow.then(generateSuggestionsStep).commit();
 
// Create an agent with the workflow
const agent = new Agent({
  name: "research-organizer",
  instructions:
    "You are a research organizer that can delegate tasks to gather information and create summaries.",
  model: openai("gpt-4o"),
  workflows: {
    research: researchWorkflow,
  },
});
 
// Get the workflows
const workflows = await agent.getWorkflows();
 
console.log(Object.keys(workflows)); // ["research"]
stream()getVoice()