Agentic Workflow for Splitting Long Responses

​

Introduction

1. Generate a long response in a structured JSON format using the OpenAI: Generate text action.
2. Split the response into smaller, more digestible parts using the Get Data from JSON action.
3. Present the split content as individual, sequential messages for a better user experience.

​

Workflow Structure

1. OpenAI: Generate text node: Produces a structured JSON response where each key represents a smaller part of the overall message.
2. Get Data from JSON action: Extracts each section of the JSON response and maps it to individual variables.
3. Send Messages nodes: Sequentially displays each part of the message to avoid overwhelming the user with one long block of text.

​

Step-by-Step Implementation

1

Configure the OpenAI: Generate Text Node

1. Add an OpenAI: Generate text node to your workflow.
2. In the Custom Instructions/Prompt section, instruct the AI to format its response as JSON with clearly defined parts. For example:
   Copy

   Please split your response into multiple parts as a JSON object, with keys like "part1", "part2", "part3", etc. Each part should contain concise information. Example format:
   {
     "part1": "This is the introduction...",
     "part2": "This is the detailed explanation...",
     "part3": "This is the conclusion..."
   }
   
3. Save the AI-generated JSON response to a custom variable, such as {{ai_json}}.

2

Use the Get Data from JSON Action to Extract Sections

1. Add the Get Data from JSON action immediately after the OpenAI: Generate text node.
2. In the JSON Data field, reference the variable where the JSON response is stored (e.g., {{ai_json}}).
3. Map each JSON key to custom variables:
   • .part1 → MessagePart1
   • .part2 → MessagePart2
   • .part3 → MessagePart3
4. Add as many mappings as needed to cover all the parts of the JSON response.

Example JSON:

Copy

{
  "part1": "Step 1: Understand the basics of nutrition.",
  "part2": "Step 2: Include more vegetables in your meals.",
  "part3": "Step 3: Stay consistent with healthy eating habits."
}

Mapped variables:

• {{MessagePart1}} → “Step 1: Understand the basics of nutrition.”
• {{MessagePart2}} → “Step 2: Include more vegetables in your meals.”
• {{MessagePart3}} → “Step 3: Stay consistent with healthy eating habits.”

3

Display the Extracted Sections as Split Messages

1. Add Send Messages nodes to display the extracted sections sequentially.
2. Insert each mapped variable into separate messages:
   • First message: {{MessagePart1}}
   • Typing delay
   • Second message: {{MessagePart2}}
   • Typing delay
   • Third message: {{MessagePart3}}

This approach ensures that each part of the response is displayed as an individual message, breaking up the long content for clarity.

​

Best Practices

​

Example Workflow

{
  "part1": "Step 1: Set clear and achievable goals.",
  "part2": "Step 2: Minimize distractions by creating a focused workspace.",
  "part3": "Step 3: Use techniques like the Pomodoro method to stay productive.",
  "part4": "Step 4: Take breaks to avoid burnout and maintain focus."
}

1. Message 1: Step 1: Set clear and achievable goals.
2. Message 2: Step 2: Minimize distractions by creating a focused workspace.
3. Message 3: Step 3: Use techniques like the Pomodoro method to stay productive.
4. Message 4: Step 4: Take breaks to avoid burnout and maintain focus.

​

Conclusion