igny8/EXPECTED_RESPONSE_STRUCTURE.md

Expected Response Structure for generate_images_from_prompts

Function Flow

1. Frontend calls: POST /v1/writer/images/generate_images/ with { ids: [1, 2, 3] }

2. Backend ViewSet (ImagesViewSet.generate_images):

   • Calls run_ai_task with function_name='generate_images_from_prompts'
   • Returns response with queued_prompts if in TEST MODE

3. AIEngine.execute:

   • Validates, prepares, builds prompt (placeholder), parses (placeholder)
   • Calls save_output which queues prompts (TEST MODE)

4. GenerateImagesFromPromptsFunction.save_output:

   • Returns dict with queued prompts

Response Structure (TEST MODE)

{
  "success": true,
  "count": 3,
  "images_generated": 0,
  "images_failed": 0,
  "total_images": 3,
  "queued_prompts": [
    {
      "image_id": 1,
      "index": 1,
      "image_type": "featured_image",
      "content_title": "Your Content Title",
      "provider": "openai",
      "model": "dall-e-2",
      "formatted_prompt": "Create a high-quality realistic image to use as a featured photo for a blog post titled \"Your Content Title\". The image should visually represent the theme, mood, and subject implied by the image prompt: [original prompt]. Focus on a realistic, well-composed scene that naturally communicates the topic without text or logos...",
      "negative_prompt": null,
      "prompt_length": 250
    },
    {
      "image_id": 2,
      "index": 2,
      "image_type": "in_article_1",
      "content_title": "Your Content Title",
      "provider": "openai",
      "model": "dall-e-2",
      "formatted_prompt": "...",
      "negative_prompt": null,
      "prompt_length": 245
    }
  ],
  "test_mode": true,
  "provider": "openai",
  "model": "dall-e-2",
  "errors": null
}

Response Structure (Production Mode - when AI calls are enabled)

{
  "success": true,
  "count": 3,
  "images_generated": 3,
  "images_failed": 0,
  "total_images": 3,
  "errors": null
}

Key Points

1. TEST MODE (current):

   • images_generated: 0 (no actual images generated)
   • queued_prompts array contains all formatted prompts
   • test_mode: true
   • Each prompt includes the complete formatted prompt that would be sent to AI

2. Production MODE (when AI calls are uncommented):

   • images_generated will be > 0
   • queued_prompts will not be in response
   • test_mode: false or not present
   • Images will have image_url and status='generated'

Console Logging

The function logs to console (if DEBUG_MODE=True):

• [generate_images_from_prompts] [TEST MODE] Queued prompt X/Y
• [generate_images_from_prompts] [TEST MODE] Provider: openai, Model: dall-e-2
• [generate_images_from_prompts] [TEST MODE] Prompt length: 250 chars
• [generate_images_from_prompts] [TEST MODE] Prompt preview: ...

Frontend Console Output

When clicking "Generate Images", browser console will show:

[Generate Images] Request: { imageIds: [1, 2, 3], count: 3 }
[Generate Images] Endpoint: /v1/writer/images/generate_images/
[Generate Images] Full Response: { success: true, queued_prompts: [...], ... }
[Generate Images] Queued Prompts (TEST MODE - NOT sent to AI): [...]
[Generate Images] Provider: openai, Model: dall-e-2
[Generate Images] Prompt 1/3:
  - Image Type: featured_image
  - Content: "Your Content Title"
  - Prompt Length: 250 chars
  - Full Prompt: "Create a high-quality realistic image..."