pull down to refresh
This follows offerings from OpenAI and Claude and has support for multimodal prompting