Hi there,
I’m Joo, a designer contributing to the Cursor-Talk-to-Figma (CTTF) project.
First of all, thank you for releasing the official DevMode MCP — especially the image extraction feature. We’ve been successfully combining DevMode MCP with our plugin API-based third-party MCP to automate many workflows, and it’s working really well.
However, I’d like to report a recurring issue when using the get_image tool.
In our setup, we typically use Figma’s get_image to extract visuals, pass them to an AI agent for analysis, and then follow up with tools from CTTF — such as UX writing or critique.
But as shown in the screenshot below, we consistently run into an error when calling get_image. While hitting “Resume” does allow the process to continue, the repeated interruption creates some confusion.
Any idea what might be causing this?
Should I raise this with Anthropic as well?
Thanks in advance!
