Figma introduced Figma Specs, expanded access to its on-canvas agent and code layers to give AI agents more design rules and project context. These tools can help teams keep AI-assisted design work aligned with their standards, but teams still need to review the results.
Key takeaways
- Figma Specs lets teams document design language and conventions in Markdown.
- Teams can keep Specs standalone or connect it to canvas elements and supporting materials.
- Figma is expanding access to its agent on canvas and adding ways to bring in design and code libraries.
- Code layers are intended to make prototypes more interactive than static screens.
- The usefulness of AI-generated design work depends on relevant, maintained context and team review.
Figma has introduced new tools that let teams document their design rules and provide more project context for AI agents working on its platform. The updates include Figma Specs, a Markdown-based way to describe a team’s design language, along with expanded access to Figma’s agent on canvas and new features for building interactive prototypes.
Figma announced the features at its first Config event in India. Together, they address a practical challenge for AI-assisted design workflows: helping agents produce work that fits a team’s existing system, rather than relying on broad instructions alone.
Figma Specs Lets Teams Document Design Rules for AI Agents
Figma Specs is an editor for writing down a team’s design language in Markdown. Teams can use it to record guidance and conventions, creating a reference for people working on a project as well as AI agents that need to follow its rules.
Specs can be created as a standalone file or placed alongside the design canvas. In the latter setup, teams can connect written guidance to Figma elements and bring in related materials, including FigJam diagrams, code layers and designs. That gives teams a way to keep rules close to the work they describe, rather than separating them into an unrelated document.
The goal is to make the design context available when agents are asked to help. Clear documentation can explain how a product should look and behave, while links to project elements can make that guidance more concrete. Teams still need to decide what their standards are and keep the documentation current; Specs provides a place to organize and share that context.
Design Libraries, Code Layers and a Wider Canvas Agent Add More Context
Figma is also expanding the context teams can bring into its environment, including design libraries and code libraries. The company is integrating with Payload, a headless content management system, adding another way to connect project work and supporting materials.
The updates also include wider availability for Figma’s agent on canvas. The agent can work within the design environment, where the canvas and its surrounding materials offer context that may be harder to convey through a standalone prompt. Figma’s announcement positions the agent alongside the new tools for documenting design rules and connecting project information.
Code layers are another part of the announcement. They are intended to make prototypes more interactive by allowing code to be used within a design. That can help teams explore behavior and demonstrate more than a static screen during prototyping.
What Figma’s Agent Updates Could Mean for Design Workflows
For teams using AI agents in design, the value of these features will depend on the quality of the context they provide. A documented design language, relevant libraries and connected examples can give agents clearer boundaries for their work. Without that groundwork, teams may still need to correct outputs that miss product conventions or visual details.
The combination also brings guidance, design materials and code closer together. That could make it easier for designers and developers to use agents during exploration and prototyping while keeping the work tied to a shared system. It does not remove the need for review: teams will still need to assess whether an agent’s output meets their standards.
Review your team’s design guidelines and libraries, then explore how Figma’s new tools could help agents work within them.
By the numbers
Figma announced the updates at its first Config event in India.
The article identifies the event as the announcement setting; it does not provide an event date or attendance figures.
Figma Specs uses Markdown to document design language.
Figma's described feature format is a concrete product detail; the article reports no adoption or performance metrics.
The announcement includes a Payload integration.
Figma describes Payload as a headless content management system and the integration as another way to connect project work and supporting materials.
Step by step
- 01
Review your design standards
Identify the visual rules, product conventions and interaction guidance that an AI agent should follow.
- 02
Document the design language
Use Figma Specs to record team guidance in Markdown, and keep it current as your design system changes.
- 03
Connect relevant project context
Where appropriate, link the guidance to Figma elements and supporting materials such as diagrams, designs, libraries or code layers.
- 04
Test the agent and review its work
Try the on-canvas agent with project-specific context, then check its output against your design standards before using it.
Frequently asked questions
What is Figma Specs?
Figma Specs is a Markdown-based editor for documenting a team's design language. Teams can use it to record rules and conventions for people and AI agents. Specs can be a standalone file or connected to design work on the canvas.
How can Figma Specs help AI agents?
Figma Specs gives AI agents written design guidance and project context to draw on. Teams can connect guidance to relevant canvas elements and supporting materials. The documentation still needs to reflect the team's current standards, and generated work needs review.
What are Figma code layers used for?
Figma code layers let teams use code within a design to make prototypes more interactive. They can help demonstrate behavior beyond a static screen. Figma says the feature is part of its updates for building interactive prototypes.
Do Figma's new AI tools guarantee designs will follow team rules?
No, the tools do not guarantee that an agent's output will meet a team's standards. Design documentation and connected project materials can provide useful context, but the quality of that context matters. Teams still need to assess and refine the results.
