I have a question and potential feature request regarding the AI Meta Filter photo upload tool.

Currently, when generating caricatures, the AI applies styling elements such as suit color randomly. I would love to know if it is possible now, or if it could be added in a future update, to allow dynamic customer inputs that directly influence the generated image.

For example:

• The store owner writes the core structured prompt that controls pose, style, framing, proportions, etc.

• The customer is given custom input fields such as:

Suit Color

Hobby or Interest

Accessory Preference

Then the AI would merge the base prompt with the customer’s inputs during generation.

Example use case:

If my base prompt defines the caricature style and pose, and the customer types “Navy suit” and “Golf,” the AI would generate the character wearing a navy suit while holding a golf club and whiskey glass instead of defaulting to random styling choices.

This would create a much more personalized and controllable output while still using the Meta Filter workflow. It would significantly increase flexibility for personalized products like caricatures, wedding gifts, retirement gifts, and hobby based gifts.

Is something like dynamic prompt variable injection possible within the current system, or could this be considered for a future update?