I ran into a limitation while testing my AnnaVisual Brand MVP.
The app can analyze brand materials, create a Brand Profile, and generate a prompt / JSON package. But when I reach the final visual generation step, the local app cannot directly call an image generation model and return the image.
Steps to reproduce
-
Open AnnaVisual Brand MVP.
-
Upload brand materials.
-
Generate and confirm the Brand Profile.
-
Create the final prompt package.
-
Try to generate the final image from the app.
Expected behavior
The app should send the prompt package to Anna’s image model and display the generated image inside AnnaVisual Brand MVP.
Actual behavior
The app can only prepare the prompt / JSON. The final image must still be generated manually in Anna chat.
This makes it hard to test the full MVP flow end to end: brand upload → brand profile → prompt package → final image.
If direct image generation is not supported for local apps yet, it would be helpful to show this clearly in the UI or docs.
