Figma announces new features to define how agents work on its platform

Figma Introduces Specs to Steer AI Agents

Design platform Figma is introducing a new set of tools aimed at managing the output of AI agents on its platform. At its first Config event in India, the company announced Figma Specs, a Markdown-based editor designed to establish clear design rules and systems. This move directly addresses the challenge of AI's unpredictability, providing a framework for both human designers and AI agents to maintain brand consistency as generative features become more integrated into the core design workflow.

Establishing Guardrails for AI Design

The centerpiece of the announcement is Figma Specs, which functions as a centralized document for a company's design language. It can exist as a standalone file or live directly within the design canvas. This tool goes beyond a simple text editor by allowing teams to embed and reference design elements, code libraries, and guidelines, creating a single source of truth. Figma also announced its 'code layers' feature, which adds interactive code to prototypes, is now moving into public beta.

• Figma Specs: A Markdown-based editor to define a design language and its associated guidelines.
• Context Integration: Allows teams to embed Figma elements, FigJam diagrams, design libraries, and code layers for context.
• Payload CMS: Deeper integration with the open-source headless CMS Payload for generating dynamic websites.
• Code Layers Beta: The feature for adding code to designs for interactive prototypes is now available in public beta.

From Design Tool to Development Hub

By providing tools to guide generative AI, Figma is positioning its platform as the central environment where software teams decide what to build and ship. Loredana Crisan, Figma's chief design officer, highlighted the massive increase in code production that is used for discovery rather than deployment. These new features aim to make that AI-assisted 'discovery' process more efficient and aligned with production standards, bridging the gap between raw AI generation and a shippable product that adheres to a coherent design system.

Figma's strategy is to evolve from a collaborative design canvas into a structured development environment. By building guardrails like Figma Specs, the company is betting that the key to enterprise adoption of generative AI isn't just rapid creation, but controlled, predictable, and on-brand creation that reduces the friction between initial concepts and shippable code.