Workflow YAML Configuration Reference

Overview

This guide provides comprehensive documentation for configuring CodeMie Workflows using YAML. CodeMie Workflows is a powerful orchestration system that enables you to create complex, multi-step AI-powered processes by coordinating multiple assistants, tools, and custom processing nodes. Built on top of LangGraph (a framework for building stateful, multi-agent applications), workflows transform simple linear AI interactions into sophisticated, automated processes.

Documentation Structure

This documentation is organized into the following sections:

Core Documentation

1. Introduction & Getting Started

   • What are CodeMie Workflows?
   • Core Architecture Components
   • Your First Workflow
   • YAML Configuration Basics

2. Configuration Reference

   • Assistants Configuration
   • Workflow-Level Settings
   • Tools Configuration
   • States Configuration
   • Custom Nodes Configuration
   • MCP Server Configuration

3. Workflow States

   • State Types
   • Agent State Configuration
   • Tool State Configuration
   • Custom Node State Configuration

4. State Transitions

   • Simple Transitions
   • Parallel Transitions
   • Conditional Transitions
   • Switch/Case Transitions
   • Iterative Transitions (Map-Reduce)

5. Context Management

   • Context Store
   • Context Configuration Options
   • Dynamic Value Resolution

Advanced Topics

6. Advanced Features

   • Map-Reduce Patterns
   • Memory Management
   • Retry Policies
   • Workflow Interruption
   • Structured Output
   • Performance Tuning

7. Specialized Node Types

   • State Processor Node
   • Bedrock Flow Node
   • Document Tree Generator

8. Integration Capabilities

   • Data Source Integration
   • Tool Integration
   • MCP (Model Context Protocol) Integration

Best Practices & Troubleshooting

9. Best Practices

   • Workflow Design Principles
   • Context Management Best Practices
   • Performance Optimization
   • Error Handling
   • Security Considerations

10. Complete Examples

    • Code Review Workflow
    • Document Processing Pipeline
    • Multi-Branch Processing

11. Troubleshooting

    • Common Issues
    • Debugging Techniques
    • Validation Process

Quick Links

• New to workflows? Start with Introduction & Getting Started
• Building a workflow? Check Configuration Reference
• Need examples? See Complete Examples
• Having issues? Visit Troubleshooting

Key Concepts

What Makes Workflows Powerful?

• Task Decomposition: Break complex problems into manageable steps
• Parallel Processing: Execute multiple operations concurrently
• Conditional Logic: Branch execution based on results
• Context Sharing: Maintain state across all workflow steps
• Tool Integration: Connect to cloud platforms, databases, and APIs
• Memory Management: Automatic summarization for long-running processes

When to Use Workflows

Use workflows when:

• Tasks require multiple distinct steps
• You need to process multiple items in parallel
• Different steps require different AI configurations
• Conditional logic based on intermediate results is needed
• Context must be preserved across multiple steps

Use single assistants when:

• The task can be completed in one interaction
• No parallelization or complex branching is needed
• Maximum flexibility for AI to determine its own approach

Getting Help

• Review the documentation sections above
• Check Complete Examples for common patterns
• Consult Best Practices for optimization tips
• See Troubleshooting for common issues

---

Version: 1.0 Last Updated: 2025-01-20