All times UTC
A new piece on enterprise AI agents argues that reliable agentic workflows are a product-design problem, not just an engineering one. That's a real opening for designers to own agent UX and review-step design.
appzime.com (opens in a new tab)
Your move
pick one repetitive task in your own workflow and sketch what a human-in-the-loop review step for an agent doing that task would actually look like.
UK design industry coverage is folding AI-workflow fluency into everyday reporting rather than treating it as a novelty. No hard hiring stats to share here, but it's a genuinely current thing to be ready to discuss.
designweek.co.uk (opens in a new tab)
Your move
before your next application or interview, run one small Figma MCP or Make experiment so you've got a real example to talk through, not a theoretical one.
As AI produces more first-draft designs, research-backed critique skills are becoming the scarcer differentiator, not prompting speed. This isn't a new method - it's an established one that just matters more now.
nngroup.com (opens in a new tab)
Your move
next time you get an AI-generated design (from an agent, Make, or elsewhere), spend 10 minutes running a heuristic evaluation on it before you ship it.
As AI agents increasingly read Figma files through MCP, messy naming and unbound tokens are becoming a real blocker. Design systems are quietly being pushed to get more legible - for agents as much as humans.
designsystems.news (opens in a new tab)
Your move
audit one component in your Figma or Penpot library this week - rename anything vague like 'Frame 42' and check its tokens are bound, not hardcoded.
Figma has opened its canvas to AI agents and expanded its MCP partner catalog, so agents can read and write Figma files directly. Penpot users don't have an equivalent yet, so it's worth tracking the gap.
figma.com (opens in a new tab)
Your move
find Figma's official MCP server setup guide, install it locally, and pull one real component into Cursor or Claude to see what context it exposes.
