Branching Workflow

Complex decision trees requiring dynamic path selection based on content analysis

Example Use-Cases: Expert routing, content type detection, multi-path processing

Dynamic routing workflows provide intelligent path selection while maintaining predictable execution within each chosen branch.

Selector Flexibility

The Router selector function supports multiple return types:

• String: Return step name - Router resolves it from choices
• Step: Return Step object directly
• List[Step]: Return list of steps for chaining

The selector can also receive step_choices as an optional second parameter for dynamic selection.

Example: String-based Selector

The simplest approach - 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.router import Router
from agno.workflow.step import Step
from agno.workflow.types import StepInput
from agno.workflow.workflow import 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

def dynamic_selector(step_input: StepInput, step_choices: list) -> Union[str, Step, List[Step]]:
    """
    Selector receives step_choices - can select by name or return Step directly.
    step_choices contains the prepared Step objects from Router.choices.
    """
    user_input = step_input.input.lower()

    # Build name map from step_choices
    step_map = {s.name: s for s in step_choices if hasattr(s, "name") and s.name}

    # Can return step name as string
    if "research" in user_input:
        return "researcher"

    # Can return Step object directly
    if "write" in user_input:
        return step_map.get("writer", step_choices[0])

    # Can return list of Steps for chaining
    if "full" in user_input:
        return [step_map["researcher"], step_map["writer"], step_map["reviewer"]]

    # Default
    return step_choices[0]


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

Example: Nested Choices

Nested lists in choices become Steps containers for sequential execution:

nested_routing.py

def nested_selector(step_input: StepInput, step_choices: list) -> Union[str, Step, List[Step]]:
    """
    When choices contains nested lists like [step_a, [step_b, step_c]],
    the nested list becomes a Steps container in step_choices.
    """
    user_input = step_input.input.lower()

    # step_choices[0] = Step for step_a
    # step_choices[1] = Steps container with [step_b, step_c]

    if "single" in user_input:
        return step_choices[0]  # Just step_a
    else:
        return step_choices[1]  # Steps container: step_b -> step_c


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

Developer Resources

• Router Steps Workflow
• Router with step_choices

Reference

For complete API documentation, see Router Steps Reference.