Workflows Development Guide

from agno.workflow.v2 import Step, Workflow, StepOutput

def data_preprocessor(step_input):
    # Custom preprocessing logic

    # Or you can also run any agent/team over here itself
    # response = some_agent.run(...)
    return StepOutput(content=f"Processed: {step_input.message}") # <-- Now pass the agent/team response in content here

workflow = Workflow(
    name="Mixed Execution Pipeline",
    steps=[
        research_team,      # Team
        data_preprocessor,  # Function
        content_agent,      # Agent
    ]
)

workflow.print_response("Analyze the competitive landscape for fintech startups", markdown=True)

def custom_workflow_function(workflow: Workflow, execution_input: WorkflowExecutionInput):
    # Custom orchestration logic
    research_result = research_team.run(execution_input.message)
    analysis_result = analysis_agent.run(research_result.content)
    return f"Final: {analysis_result.content}"

workflow = Workflow(
    name="Function-Based Workflow",
    steps=custom_workflow_function  # Single function replaces all steps
)

workflow.print_response("Evaluate the market potential for quantum computing applications", markdown=True)

from agno.workflow.v2 import Step, Workflow

# Named steps for better tracking
workflow = Workflow(
    name="Content Creation Pipeline",
    steps=[
        Step(name="Research Phase", team=researcher),
        Step(name="Analysis Phase", executor=custom_function), 
        Step(name="Writing Phase", agent=writer),
    ]
)

workflow.print_response(
    "AI trends in 2024",
    markdown=True,
)

from agno.workflow.v2 import Condition, Step, Workflow

def is_tech_topic(step_input) -> bool:
    topic = step_input.message.lower()
    return any(keyword in topic for keyword in ["ai", "tech", "software"])

workflow = Workflow(
    name="Conditional Research",
    steps=[
        Condition(
            name="Tech Topic Check",
            evaluator=is_tech_topic,
            steps=[Step(name="Tech Research", agent=tech_researcher)]
        ),
        Step(name="General Analysis", agent=general_analyst),
    ]
)

workflow.print_response("Comprehensive analysis of AI and machine learning trends", markdown=True)

from agno.workflow.v2 import Parallel, Step, Workflow

workflow = Workflow(
    name="Parallel Research Pipeline",
    steps=[
        Parallel(
            Step(name="HackerNews Research", agent=hn_researcher),
            Step(name="Web Research", agent=web_researcher),
            Step(name="Academic Research", agent=academic_researcher),
            name="Research Step"
        ),
        Step(name="Synthesis", agent=synthesizer),  # Combines the results and produces a report
    ]
)

workflow.print_response("Write about the latest AI developments", markdown=True)

from agno.workflow.v2 import Loop, Step, Workflow

def quality_check(outputs) -> bool:
    # Return True to break loop, False to continue
    return any(len(output.content) > 500 for output in outputs)

workflow = Workflow(
    name="Quality-Driven Research",
    steps=[
        Loop(
            name="Research Loop",
            steps=[Step(name="Deep Research", agent=researcher)],
            end_condition=quality_check,
            max_iterations=3
        ),
        Step(name="Final Analysis", agent=analyst),
    ]
)

workflow.print_response("Research the impact of renewable energy on global markets", markdown=True)

from agno.workflow.v2 import Router, Step, Workflow

def route_by_topic(step_input) -> List[Step]:
    topic = step_input.message.lower()
    
    if "tech" in topic:
        return [Step(name="Tech Research", agent=tech_expert)]
    elif "business" in topic:
        return [Step(name="Business Research", agent=biz_expert)]
    else:
        return [Step(name="General Research", agent=generalist)]

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

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

from agno.workflow.v2 import Steps, Step, Workflow

# Create a reusable content creation sequence
article_creation_sequence = Steps(
    name="ArticleCreation",
    description="Complete article creation workflow from research to final edit",
    steps=[
        Step(name="research", agent=researcher),
        Step(name="writing", agent=writer), 
        Step(name="editing", agent=editor),
    ],
)

# Use the sequence in a workflow
workflow = Workflow(
    name="Article Creation Workflow",
    steps=[article_creation_sequence]  # Single sequence
)

workflow.print_response("Write an article about renewable energy", markdown=True)

from agno.workflow.v2 import Steps, Router, Step, Workflow

# Define two completely different workflows as Steps
image_sequence = Steps(
    name="image_generation",
    description="Complete image generation and analysis workflow",
    steps=[
        Step(name="generate_image", agent=image_generator),
        Step(name="describe_image", agent=image_describer),
    ],
)

video_sequence = Steps(
    name="video_generation", 
    description="Complete video production and analysis workflow",
    steps=[
        Step(name="generate_video", agent=video_generator),
        Step(name="describe_video", agent=video_describer),
    ],
)

def media_sequence_selector(step_input) -> List[Step]:
    """Route to appropriate media generation pipeline"""
    if not step_input.message:
        return [image_sequence]
        
    message_lower = step_input.message.lower()
    
    if "video" in message_lower:
        return [video_sequence]
    elif "image" in message_lower:
        return [image_sequence]
    else:
        return [image_sequence]  # Default

# Clean workflow with clear branching
media_workflow = Workflow(
    name="AI Media Generation Workflow",
    description="Generate and analyze images or videos using AI agents",
    steps=[
        Router(
            name="Media Type Router",
            description="Routes to appropriate media generation pipeline",
            selector=media_sequence_selector,
            choices=[image_sequence, video_sequence],  # Clear choices
        )
    ],
)

# Usage examples
media_workflow.print_response("Create an image of a magical forest", markdown=True)
media_workflow.print_response("Create a cinematic video of city timelapse", markdown=True)

from agno.workflow.v2 import Condition, Loop, Parallel, Router, Step, Workflow

def research_post_processor(step_input) -> StepOutput:
    """Post-process and consolidate research data from parallel conditions"""
    research_data = step_input.previous_step_content or ""
    
    try:
        # Analyze research quality and completeness
        word_count = len(research_data.split())
        has_tech_content = any(keyword in research_data.lower() 
                              for keyword in ["technology", "ai", "software", "tech"])
        has_business_content = any(keyword in research_data.lower() 
                                  for keyword in ["market", "business", "revenue", "strategy"])
        
        # Create enhanced research summary
        enhanced_summary = f"""
            ## Research Analysis Report
            
            **Data Quality:** {"✓ High-quality" if word_count > 200 else "⚠ Limited data"}
            
            **Content Coverage:**
            - Technical Analysis: {"✓ Completed" if has_tech_content else "✗ Not available"}
            - Business Analysis: {"✓ Completed" if has_business_content else "✗ Not available"}
            
            **Research Findings:**
            {research_data}
        """.strip()
        
        return StepOutput(
            content=enhanced_summary,
            success=True,
        )
        
    except Exception as e:
        return StepOutput(
            content=f"Research post-processing failed: {str(e)}",
            success=False,
            error=str(e)
        )

# Complex workflow combining multiple patterns
workflow = Workflow(
    name="Advanced Multi-Pattern Workflow",
    steps=[
        Parallel(
            Condition(
                name="Tech Check",
                evaluator=is_tech_topic,
                steps=[Step(name="Tech Research", agent=tech_researcher)]
            ),
            Condition(
                name="Business Check", 
                evaluator=is_business_topic,
                steps=[
                    Loop(
                        name="Deep Business Research",
                        steps=[Step(name="Market Research", agent=market_researcher)],
                        end_condition=research_quality_check,
                        max_iterations=3
                    )
                ]
            ),
            name="Conditional Research Phase"
        ),
        Step(
            name="Research Post-Processing",
            executor=research_post_processor,
            description="Consolidate and analyze research findings with quality metrics"
        ),
        Router(
            name="Content Type Router",
            selector=content_type_selector,
            choices=[blog_post_step, social_media_step, report_step]
        ),
        Step(name="Final Review", agent=reviewer),
    ]
)

workflow.print_response("Create a comprehensive analysis of sustainable technology trends and their business impact for 2024", markdown=True)