OpenAI Asset Pack

This graph generates a coordinated image, voiceover audio clip, metadata JSON, and markdown campaign brief from one prompt/script pair.

Before running it, add a graph-scoped secret named OPENAI_API_KEY in Graph Settings > Execution Profile.

High-level shape:

Creative prompt -> Image JSON body -> OpenAI image request -> Base64 image -> Image media
Voice script -> Speech JSON body -> OpenAI speech request -> Audio media
Image media + Audio media + Prompt + Script -> Metadata JSON -> Markdown campaign brief
Image media -> Media Response
Audio media -> Media Response
Markdown campaign brief -> Text Response

Recommended panels:

• Image Preview bound to Image media: media
• Audio Preview bound to Speech request: media
• Markdown Preview bound to Campaign brief: value
• JSON Inspector bound to Asset metadata: value

Why this graph exists:

• It exercises fan-out from authored inputs.
• It runs multiple provider requests in parallel once the executor supports async scheduling.
• It exercises image, audio, markdown, JSON, and media response panels together.
• It highlights the need for artifact storage because large inline data: URLs are useful locally but not ideal long term.