Figma opens its canvas to AI agents via MCPs as Google’s free Stitch tool puts pressure on the company
Figma’s stock dropped 8.8% last week after Google launched a free AI design tool. Five days later, Figma opened its canvas to AI agents.
Figma launched an open beta today letting AI agents design directly on its canvas using a new use_figma MCP tool. Claude Code, Codex, Copilot, Cursor, and other MCP clients can now generate and edit Figma files tied to a team’s existing design system.
The timing is notable. Last week, Google updated its free AI design tool Stitch with voice input, an infinite canvas, and interactive prototyping, triggering an 8.8% single-day drop in Figma stock. The market read it as a direct threat to Figma’s core value proposition.
- What
use_figmadoes: Agents create and modify components using real design system variables, auto layout, and existing components — not generic AI output detached from team conventions. - What makes it different from Stitch: Figma agents work within your design system. Generic AI tools generate from scratch and produce output that has no connection to a team’s existing standards.
The key mechanism is “skills” — markdown files that encode team-specific workflows, naming conventions, and design rules. Skills make agent output predictable and brand-aligned instead of generically competent.
Nine community-built skills are available at launch, including /figma-generate-library for building components from a codebase, /apply-design-system for connecting existing designs to system components, and /create-voice for generating screen reader specs from UI designs.
- Free during beta: Will eventually be a usage-based paid feature.
- Works with: Augment, Claude Code, Codex, Copilot CLI, Copilot in VS Code, Cursor, Factory, Firebender, and Warp.
The Bottom Line: Figma is doing what every major software company is doing right now — opening up to AI agents before someone else does it for them.
Source: Figma
If you need on-demand GPUs for training, fine-tuning, inference, or running open-source models, give RunPod a try.
- Available hardware: H100, H200, A100, L40S, RTX 4090, RTX 5090, and 30+ more
- Cost: significantly cheaper than AWS or GCP, billed per second, no contracts
- Setup: spins up in under a minute, 30+ regions worldwide
Get the core business tech news delivered straight to your inbox. We track AI, automation, SaaS, and cybersecurity so you don't have to.
Just read what you want, and be done with it.
