Figma’s New Markdown Tools Control AI Agents on Canvas

According to TechCrunch, Figma is addressing the unpredictable nature of artificial intelligence by releasing tools designed to keep autonomous agents aligned with strict brand guidelines. During its Config event in India, the platform introduced a Markdown-based editor that acts as a centralized rulebook for both human teams and automated systems.
How Figma Specs and Code Layers Work
The new Figma Specs feature functions as a simple Markdown editor where users can define their design language directly within the workspace. Teams can embed FigJam diagrams, reference canvas elements, and link design libraries to create a single source of truth.
Alongside Specs, Figma expanded its code layers feature into public beta. This allows developers and designers to ask the canvas agent to turn static designs into code or modify prototypes instantly, bridging the gap between initial ideation and functional output.
How to Test the New AI Control Features
To use these new governance tools in your daily workflow, follow these steps:
- Open your Figma workspace and look for the new Markdown editor option to create a Specs document.
- Document your core design systems, style guidelines, and code libraries inside the Specs file so agents can read them as context.
- Access the canvas agent and instruct it to build or modify elements while referencing your newly created design rules.
- Opt into the public beta for code layers if you want to test generating and tweaking interactive prototype code directly on the canvas.
Verdict: Who Should Use This Right Now?
If your team struggles with AI agents producing off-brand variations or messy prototypes, these tools are worth adopting immediately. By forcing models to read centralized Markdown specs, you cut down on the random trial-and-error typical of generative design tools.
Sources
Frequently asked questions
- What is Figma Specs?
- Figma Specs is a new Markdown editor that lets users define design rules and guidelines in a simple document to serve as a reference for teams and AI agents.
- How do code layers help with AI in Figma?
- Code layers let users test out interactive prototypes and ask AI agents to turn visual designs into actual code or modify existing code blocks directly on the canvas.
Comments
0 commentsDeixe seu comentário
Be the first to comment.


