Question
Has Figma changed the underlying LLMs that Agents use? Quality has dropped like a rock.
Did Figma do a rug-pull with Agents recently? (< 2 weeks) Previously, with the exact same level of detail and instructions in prompts, I was singing Agents praises. This week it won’t follow simple instructions and each time admits it’s not paying close enough attention. Did Figma make a conscientious decision to downgrade underlying LLM?
Based on recent utility, I’m better off doing it all myself. Not even out of Beta yet and already garbage.