News

Figma Gives AI Agents a Design Rulebook With New Specs Feature

Figma has announced new features designed to help AI agents work more consistently with design systems and code. Its Specs feature lets teams document design rules in Markdown and provide agents with additional project context. The company is also expanding Code Layers, which supports code-based interactive prototypes, and adding more integrations through its Model Context Protocol (MCP) connections.

Written By : Soham Halder
Reviewed By : Pranchal Srivastava

AI tools can create designs and write code in seconds, but getting them to follow a company’s exact design rules is another challenge. Figma is trying to address that problem with a new set of features that give AI agents more context about how a product should look and work.

The design platform announced the updates at its first Config event in India on Thursday (October 14, 2026). The biggest addition is Figma Specs, a Markdown-based editor that lets teams document their design language and give both human designers and AI agents a shared reference to work from. The idea is to reduce guesswork when AI generates designs or makes changes to existing projects. 

Figma Specs Gives AI Agents a Design Rulebook

Figma Specs allows teams to describe their design systems in plain text, including the rules and guidelines that define a brand’s visual identity. The document can sit inside a Figma canvas or exist as a separate file.

Teams can also connect Specs to existing designs, FigJam diagrams and code layers. Design libraries, code libraries and other guidelines can be brought together so agents have more information before they start creating or editing a project. By giving agents a clearer reference, Figma hopes to make their output more predictable and easier for teams to review.

Code Layers Opens to More Users

Figma is also moving Code Layers into public beta after introducing the feature in a private beta in June. It allows users to add code directly to designs on the canvas, making it possible to build more interactive prototypes.

Users can ask an agent to turn a design into code or modify existing code while refining a prototype. This brings design and development closer together, potentially reducing the back-and-forth involved in testing an idea before building the final product.

The feature is now open for wider testing, although its availability and capabilities may continue to evolve during the beta period. 

Also Read: Webflow vs Figma vs Adobe XD: Which One Should You Learn?

Figma Expands AI Connections and Adds Payload

The company also announced support for more than 20 additional tools through its Model Context Protocol (MCP) integrations. These include OpenCode, Dia, Grok Bot and Indian AI startup Sarvam.

Figma is also integrating Payload, the open-source headless content management system it acquired last year, into Figma Make. The integration is intended to help users build dynamic websites and manage their content within the same workflow.

Join our WhatsApp Channel to get the latest news, exclusives and videos on WhatsApp

HODL vs Active Crypto Trading: How the Two Strategies Differ

How to Measure Solana Network Adoption in the Coming Years

AI Agents Drive 138 Million x402 Payments as US Leads Crypto Growth

Bitcoin and Ethereum Liquidations Hit USD 483M as CPI Surges

BingX Secures EU Crypto License in Austria, Targets 29 European Markets