Workflow Builder & Execution

Workflows are constructed using the WorkflowBuilder class, which provides a fluent API for defining the workflow structure:

using Microsoft.Agents.AI.Workflows;

var processor = new DataProcessor();
var validator = new Validator();
var formatter = new Formatter();

// Build workflow
WorkflowBuilder builder = new(processor); // Set starting executor
builder.AddEdge(processor, validator);
builder.AddEdge(validator, formatter);
var workflow = builder.Build();

Workflows are constructed using the WorkflowBuilder class:

from agent_framework import WorkflowBuilder

processor = DataProcessor()
validator = Validator()
formatter = Formatter()

# Build workflow
builder = WorkflowBuilder(start_executor=processor)
builder.add_edge(processor, validator)
builder.add_edge(validator, formatter)
workflow = builder.build()

The workflow package provides a graph-based execution model where executors are connected by edges.

• Executor - A processing unit that receives input and produces output
• Edge - Connects the output of one executor to the input of another
• Builder - Constructs workflows by defining executors and edges
• Run - Executes a workflow with given input

import (
    "github.com/microsoft/agent-framework-go/workflow"
    "github.com/microsoft/agent-framework-go/workflow/inproc"
)

uppercase := workflow.NewExecutor("UppercaseExecutor", func(input string) string {
    return strings.ToUpper(input)
}).Bind()

reverse := workflow.NewExecutor("ReverseExecutor", func(input string) string {
    runes := []rune(input)
    slices.Reverse(runes)
    return string(runes)
}).Bind()

wf, err := workflow.NewBuilder(uppercase).
    AddEdge(uppercase, reverse).
    WithOutputFrom(reverse).
    Build()
if err != nil {
    return err
}

using Microsoft.Agents.AI.Workflows;

// Streaming execution — get events as they happen
StreamingRun run = await InProcessExecution.RunStreamingAsync(workflow, inputMessage);
await foreach (WorkflowEvent evt in run.WatchStreamAsync())
{
    if (evt is ExecutorCompletedEvent executorComplete)
    {
        Console.WriteLine($"{executorComplete.ExecutorId}: {executorComplete.Data}");
    }

    if (evt is WorkflowOutputEvent outputEvt)
    {
        Console.WriteLine($"Workflow completed: {outputEvt.Data}");
    }
}

// Non-streaming execution — wait for completion
Run result = await InProcessExecution.RunAsync(workflow, inputMessage);
foreach (WorkflowEvent evt in result.NewEvents)
{
    if (evt is WorkflowOutputEvent outputEvt)
    {
        Console.WriteLine($"Final result: {outputEvt.Data}");
    }
}

# Streaming execution — get events as they happen
async for event in workflow.run(input_message, stream=True):
    if event.type == "output":
        print(f"Workflow completed: {event.data}")

# Non-streaming execution — wait for completion
events = await workflow.run(input_message)
print(f"Final result: {events.get_outputs()}")

Use RunStreaming when you want events as they happen:

stream, err := inproc.Default.RunStreaming(context.Background(), wf, "Hello, World!")
if err != nil {
    return err
}
defer stream.Close(context.Background())

for evt, err := range stream.WatchStream(context.Background()) {
    if err != nil {
        return err
    }
    if output, ok := evt.(workflow.OutputEvent); ok {
        fmt.Printf("Workflow completed: %v\n", output.Output)
    }
}

Use Run when you want to wait for workflow completion and then inspect the collected events:

run, err := inproc.Default.Run(context.Background(), wf, "Hello, World!")
if err != nil {
    return err
}

for evt := range run.NewEvents() {
    if output, ok := evt.(workflow.OutputEvent); ok {
        fmt.Printf("Final result: %v\n", output.Output)
    }
}

You can also inspect executor events collected by a non-streaming run:

for evt := range run.NewEvents() {
    if evt, ok := evt.(workflow.ExecutorCompletedEvent); ok {
        fmt.Printf("%s: %v\n", evt.ExecutorID, evt.Result)
    }
}