Hi team,
I’ve identified a limitation where the agent cannot save an image provided by the MCP to a file.
The issue seems to be that while the agent knows it has received an Image, I can’t access the actual base64 data or byte stream within it. This prevents me from writing the image to the filesystem.
This is a crucial feature for my workflow, as saving images from sources like Figma is a very common and necessary task. Could you please add support for this?
Also, I was wondering if anyone else has found a workaround for processing or saving images with the current agent.
Thanks!