Skip to main content

Agentic AI Workflow Templates

Get started quickly with AgentMap using our curated collection of ready-to-use agentic AI workflow templates. Each template demonstrates autonomous agent behavior, multi-agent collaboration, and intelligent decision-making patterns that you can customize for your specific use cases.

AgentMap Template Library

Loading templates...

📥 Quick Download Templates​

Download these essential workflow templates to get started immediately:

🌟 Starter Templates​

🚀 Basic Agentic Workflow Template
Perfect for beginners - autonomous agents with intelligent error handling
Autonomous Input → Reasoning → Output with Agent Collaboration

🌤️ Multi-Agent Weather System
Demonstrates agent specialization and LLM orchestration
Features: Custom API agents, LLM reasoning agents, intelligent routing

📊 Agentic Data Pipeline
Autonomous ETL with intelligent validation and decision-making
Capabilities: Self-validating agents, AI transformation, adaptive processing

🎯 Multi-Agent Customer Support
Sophisticated intent classification with specialized handler agents
Architecture: Intent agents, routing logic, specialized response agents, interaction logging

Quick Start with Agentic AI Templates
  1. Right-click any template link above → Save link as...
  2. Save the CSV file to your AgentMap project directory
  3. Run with: agentmap run --csv your_agentic_workflow.csv
  4. Customize agent behaviors and routing for your specific use case

How to Use Templates​

1. Choose Your Template​

Browse the template library above to find a workflow that matches your needs. Use the category and difficulty filters to narrow down your options.

2. Copy the CSV Content​

Click the "📋 Copy CSV" button on any template to copy the workflow definition to your clipboard.

3. Load into AgentMap​

Paste the CSV content into AgentMap using one of these methods:

  • Command Line: Save as a .csv file and run:

    agentmap execute your_workflow.csv

  • Python API: Load directly in your code:

    from agentmap import AgentMap

    # Paste CSV content here
    csv_content = """graph_name,node_name,next_node,context,agent_type,next_on_success,next_on_failure,input_fields,output_field,prompt
    ..."""

    agent_map = AgentMap()
    result = agent_map.execute_from_csv_string(csv_content)

  • Playground: Use the "🚀 Open in Playground" button to launch the template directly in the AgentMap web interface.

4. Customize Configuration​

Most templates include configuration notes with specific customization instructions. Common customizations include:

  • File Paths: Update input/output paths for your directory structure
  • API Keys: Configure LLM providers and external services
  • Prompts: Modify prompts to match your specific domain or tone
  • Agent Parameters: Adjust temperature, model selection, and other agent settings

Template Categories​

🤖 Automation​

Workflows that automate repetitive tasks and processes:

  • Weather Notification Bot: Daily weather alerts with intelligent notifications
  • Email Classifier: Automatic email categorization and priority routing

📊 Data Processing​

Templates for data transformation, analysis, and reporting:

  • Daily Report Generator: Automated data collection and report generation
  • Data ETL Pipeline: Extract, transform, and load data between systems

🧠 AI/LLM​

AI-powered workflows leveraging language models:

  • Customer Feedback Analyzer: Sentiment analysis and issue categorization
  • Document Summarizer: Multi-level document processing and summarization
  • Translation Workflow: Multi-language translation with quality assurance
  • Content Moderator: AI-powered content moderation and compliance

👁️ Monitoring​

Real-time monitoring and alerting systems:

  • Social Media Monitor: Track mentions with sentiment analysis and alerts
  • API Health Checker: Monitor endpoint health with automated reporting

🔗 Integration​

Templates for connecting different systems and services:

  • Data ETL Pipeline: Seamless data movement between formats

🛠️ Utility​

General-purpose workflows for common tasks:

  • Various utility templates for file processing, data validation, and more

Difficulty Levels​

🟢 Beginner​

Perfect for new users learning AgentMap:

  • Simple, linear workflows
  • Basic agent types (echo, input, llm)
  • Minimal configuration required
  • Clear documentation and examples

🟡 Intermediate​

For users comfortable with AgentMap basics:

  • Multi-step workflows with branching
  • Multiple agent types and data formats
  • Some external integrations
  • Customizable parameters

🔴 Advanced​

Complex workflows for experienced users:

  • Sophisticated routing and orchestration
  • Multiple data sources and outputs
  • External API integrations
  • Advanced error handling

Customization Guide​

Modifying Prompts​

LLM agent prompts can be customized to match your specific needs:

graph_name,node_name,next_node,context,agent_type,next_on_success,next_on_failure,input_fields,output_field,prompt
MyWorkflow,Analyzer,,"{'temperature': 0.3}",llm,Next,,input,analysis,"Analyze this data for trends and insights: {input}. Focus on actionable recommendations."

Tips for prompt customization:

  • Be specific about the desired output format
  • Include examples of good responses
  • Set the appropriate tone (formal, casual, technical)
  • Use field placeholders like {input} for dynamic content

Configuring Agent Context​

Many agents accept context parameters for fine-tuning:

Context
"{'temperature': 0.7, 'model': 'gpt-4', 'max_tokens': 500}"
"{'format': 'records', 'encoding': 'utf-8'}"
"{'chunk_size': 1000, 'should_split': true}"

Common context options:

  • LLM agents: temperature, model, max_tokens
  • File agents: encoding, mode, chunk_size
  • CSV agents: format, delimiter, id_field

Adding Error Handling​

Robust workflows include proper error handling:

graph_name,node_name,next_node,context,agent_type,next_on_success,next_on_failure,input_fields,output_field,prompt
MyWorkflow,ProcessData,,Process the data,llm,SaveResult,ErrorHandler,input,result,"Process: {input}"
MyWorkflow,ErrorHandler,,Handle errors gracefully,echo,End,,error,error_msg,"Error occurred: {error}"

File Path Configuration​

Update file paths to match your directory structure:

# Input files
"data/input.csv"
"config/settings.json"

# Output files  
"reports/daily_summary.md"
"output/processed_data.csv"

Best Practices​

1. Start Simple​

Begin with beginner templates and gradually work up to more complex workflows as you become comfortable with AgentMap concepts.

2. Test Incrementally​

When customizing templates:

  • Make small changes at a time
  • Test each modification before adding more
  • Use the error messages to guide troubleshooting

3. Organize Your Files​

Create a clear directory structure for your workflows:

my_agentmap_project/
├── workflows/
│   ├── daily_reports.csv
│   ├── content_moderation.csv
│   └── data_processing.csv
├── data/
│   ├── input/
│   └── output/
├── config/
│   └── settings.json
└── reports/

4. Version Control​

Keep your customized workflows in version control to track changes and collaborate with team members.

5. Document Customizations​

When modifying templates, document your changes:

  • What was changed and why
  • Any new dependencies or requirements
  • Expected input/output formats

Troubleshooting​

Common Issues​

CSV Format Errors

  • Ensure all rows have the same number of columns
  • Check for unescaped commas in text fields
  • Verify column headers match expected format

Agent Configuration

  • Validate JSON syntax in Context fields
  • Check that required agent types are available
  • Ensure file paths exist and are accessible

Missing Dependencies

  • Install required Python packages
  • Configure API keys for external services
  • Verify file permissions for input/output directories

Getting Help​

Need assistance with templates?

Contributing Templates​

Have a useful workflow template to share? We welcome contributions!

Template Requirements​

  • Well-documented use case and configuration
  • Tested and working example
  • Clear setup instructions
  • Appropriate difficulty classification

Submission Process​

  1. Create your template following our format
  2. Test thoroughly with sample data
  3. Document configuration requirements
  4. Submit via pull request with description

See our Contributing Guide for detailed submission instructions.

Ready to get started? Choose a template above that matches your use case and start building your first AgentMap workflow!