Skip to main content

Overview

The Workflow Copilot Agent Node provides an AI agent that can understand and manipulate workflow structures. This node enables you to:
  • Analyze existing workflows and explain their logic
  • Suggest improvements to workflow designs
  • Programmatically modify workflow structures
  • Answer questions about workflow configurations
  • Construct new workflow segments

Configuration Parameters

Node Configuration

  • Query: Ask questions or give instructions about the workflow
    Tell me about this workflow
    What does this workflow do?
    Add a node that sends an email after the LLM node
  • Workflow Data: The workflow structure (nodes and edges) in JSON format. Typically passed from workflow inputs.
  • System Prompt: Customize the copilot’s behavior and capabilities
  • Model: Select the language model (must support function calling)
  • Max Steps: Maximum reasoning and tool-use iterations (1-50)

Expected Inputs and Outputs

  • Inputs:
    • The node accepts workflow data and queries about workflow structure
  • Outputs:
    • output: The copilot’s response or modified workflow structure
    • conversation: Detailed log of the copilot’s reasoning and actions

Use Case Examples

  1. Workflow Analysis: Analyze complex workflows to understand their logic, identify inefficiencies, or generate documentation.
  2. Automated Workflow Generation: Programmatically create workflow segments based on natural language descriptions.
  3. Workflow Optimization: Get AI-powered suggestions for improving workflow structure, performance, or maintainability.

Error Handling and Troubleshooting

  • Invalid Workflow Data: Ensure the workflow_data parameter contains valid JSON with nodes and edges arrays.
  • Model Compatibility: This node requires models with function calling support. Verify your selected model has this capability.
  • Complex Modifications: The copilot works best with clear, specific instructions. Break down complex changes into smaller steps.
If you encounter any issues not covered in this documentation, please reach out to our support team for assistance.

Relevant Nodes