Workflow Patterns

Sequential, parallel, conditional, and looping patterns for orchestrating agents, teams, and functions.

Workflows orchestrate agents, teams, and functions through defined control-flow patterns. Each pattern covers a specific control flow, from sequential steps to branching, loops, and parallel execution, and patterns can be combined in the same workflow.

Building Blocks

The core building blocks of Agno Workflows are:

ComponentPurpose
StepBasic execution unit
AgentAI assistant with specific role
TeamCoordinated group of agents
FunctionCustom Python logic
StepsNamed sequential group
ParallelConcurrent execution
ConditionConditional execution
LoopIterative execution
RouterDynamic routing
WorkflowNested workflow

Sequential Workflows

Linear execution with step-by-step processing

Parallel Workflows

Concurrent execution for independent tasks

Conditional Workflows

Branching logic based on conditions

Iterative Workflows

Repeat steps until an exit condition or iteration cap

Branching Workflows

Dynamic routing and path selection

Grouped Steps

Named sequences for use in larger control flows

Advanced Patterns

Step-Based Workflows

Name steps for event tracking and output lookup

Custom Function Steps

Custom Python functions as workflow steps

Fully Python Workflow

Replace the step list with one orchestration function

Nested Workflows

Run one workflow as a step inside another

Multi-Pattern Combinations

Complex workflows combining multiple patterns