Branching Workflow

Example use cases: Expert routing, content type detection, multi-path processing A Router uses a selector function to decide at runtime which of its choices to execute. Multiple selected choices run sequentially. A selected Parallel, Loop, or other container applies its own control flow.

Selector Flexibility

The Router selector function supports multiple return types:

Return value	Behavior
Step name as `str`	Resolve the matching named choice
`Step`, `Steps`, `Condition`, `Loop`, `Parallel`, or `Router`	Execute that step-like object
`list` of supported values	Resolve each item and execute them sequentially

The selector can also receive step_choices as an optional second parameter for dynamic selection, or be a CEL expression string that returns a step name.

Example: String-based Selector

The simplest approach is to return the step name as a string:

branching_workflow.py

from typing import Union, List

from agno.agent import Agent
from agno.models.openai import OpenAIChat
from agno.workflow import Router, Step, StepInput, Workflow

tech_expert = Agent(
    name="tech_expert",
    model=OpenAIChat(id="gpt-4o-mini"),
    instructions="You are a tech expert. Provide technical analysis.",
)

biz_expert = Agent(
    name="biz_expert",
    model=OpenAIChat(id="gpt-4o-mini"),
    instructions="You are a business expert. Provide business insights.",
)

generalist = Agent(
    name="generalist",
    model=OpenAIChat(id="gpt-4o-mini"),
    instructions="You are a generalist. Provide general information.",
)

tech_step = Step(name="Tech Research", agent=tech_expert)
business_step = Step(name="Business Research", agent=biz_expert)
general_step = Step(name="General Research", agent=generalist)


def route_by_topic(step_input: StepInput) -> Union[str, Step, List[Step]]:
    """Selector can return step name as string - Router resolves it."""
    topic = step_input.input.lower()

    if "tech" in topic or "ai" in topic or "software" in topic:
        return "Tech Research"  # Return name as string
    elif "business" in topic or "market" in topic or "finance" in topic:
        return "Business Research"
    else:
        return "General Research"


workflow = Workflow(
    name="Expert Routing",
    steps=[
        Router(
            name="Topic Router",
            selector=route_by_topic,
            choices=[tech_step, business_step, general_step],
        ),
    ],
)

workflow.print_response("Latest developments in artificial intelligence", markdown=True)

Example: Using step_choices Parameter

Access available choices dynamically for more flexible routing:

dynamic_routing.py

from agno.workflow import Router, Step, StepInput, StepOutput, Workflow


def research(step_input: StepInput) -> StepOutput:
    return StepOutput(content=f"Research for {step_input.input}")


def write(step_input: StepInput) -> StepOutput:
    return StepOutput(content=f"Draft from:\n{step_input.previous_step_content}")


def review(step_input: StepInput) -> StepOutput:
    return StepOutput(content=f"Reviewed:\n{step_input.previous_step_content}")


researcher = Step(name="researcher", executor=research)
writer = Step(name="writer", executor=write)
reviewer = Step(name="reviewer", executor=review)


def dynamic_selector(step_input: StepInput, step_choices: list):
    user_input = step_input.input.lower()
    step_map = {s.name: s for s in step_choices if hasattr(s, "name") and s.name}

    if "research" in user_input:
        return "researcher"
    if "write" in user_input:
        return step_map.get("writer", step_choices[0])
    if "full" in user_input:
        return [step_map["researcher"], step_map["writer"], step_map["reviewer"]]
    return step_choices[0]


workflow = Workflow(
    name="Dynamic Routing",
    steps=[
        Router(
            name="Dynamic Router",
            selector=dynamic_selector,
            choices=[researcher, writer, reviewer],
        ),
    ],
)

workflow.print_response("Run the full pipeline")

Example: Nested Choices

Nested lists in choices become Steps containers for sequential execution:

nested_routing.py

from agno.workflow import Router, Step, StepInput, StepOutput, Workflow


def process_a(step_input: StepInput) -> StepOutput:
    return StepOutput(content=f"A: {step_input.input}")


def process_b(step_input: StepInput) -> StepOutput:
    return StepOutput(content=f"B: {step_input.input}")


def process_c(step_input: StepInput) -> StepOutput:
    return StepOutput(content=f"C after {step_input.previous_step_content}")


step_a = Step(name="single", executor=process_a)
step_b = Step(name="first_in_group", executor=process_b)
step_c = Step(name="second_in_group", executor=process_c)


def nested_selector(step_input: StepInput, step_choices: list):
    user_input = str(step_input.input).lower()
    if "single" in user_input:
        return step_choices[0]
    return step_choices[1]


workflow = Workflow(
    name="Nested Choices Routing",
    steps=[
        Router(
            name="Nested Router",
            selector=nested_selector,
            choices=[step_a, [step_b, step_c]],
        ),
    ],
)

workflow.print_response("Run the grouped branch")

Developer Resources

Reference

For complete API documentation, see Router Steps Reference.

Get Started

Basics

Advanced

Production

Providers

Other

Additional Resources

Branching Workflow

Selector Flexibility

Example: String-based Selector

Example: Using step_choices Parameter

Example: Nested Choices

Developer Resources

Reference

​Selector Flexibility

​Example: String-based Selector

​Example: Using step_choices Parameter

​Example: Nested Choices

​Developer Resources

​Reference

Selector Flexibility

Example: String-based Selector

Example: Using step_choices Parameter

Example: Nested Choices

Developer Resources

Reference