Anthropic Launches New /design Skill for Claude Code UI Workflows

Anthropic has introduced a new /design skill for Claude Code, expanding its developer-focused AI platform into user interface design workflows. The feature is currently available as a research preview.

It brings Claude Design’s artboard-based experience to both the Claude Code command-line interface and desktop application. The new capability allows developers to create editable UI artboards directly from a Claude Code session.

Instead of moving between a design tool, written specifications, and a coding environment, users can ask Claude to generate interface concepts, select an option, refine the layout, and instruct the model to implement the approved design.

Anthropic said the workflow is built on Artifacts, its format for generating and editing interactive outputs. Developers can invoke the feature by entering /design, after which Claude can produce visual interface drafts for an application or feature request. The resulting artboards are editable, allowing users to adjust a proposed UI before converting it into working code.

Anthropic Launches /design Skill for Claude Code

This workflow could reduce friction in early-stage product development, where design decisions often move through several tools and teams before implementation begins. Claude Code can now serve as a bridge between natural-language product requirements, visual prototyping, and code generation.

For example, a developer building an internal security operations dashboard could specify requirements such as an alert queue, severity filters, an incident timeline, an IOC search panel, and analyst notes.

The /design skill could generate several dashboard layouts as editable artboards. The developer can choose a preferred structure, revise sections such as alert prioritization or dark-mode contrast, and then ask Claude to implement the selected interface.

The announcement is notable because it moves Claude Code beyond its existing role as an AI coding assistant. Agentic coding tools are increasingly able to inspect repositories, modify files, execute commands, and help debug software.

Adding a visual design layer enables a more complete development loop. However, developers will still need to review generated code, validate accessibility, and test application security before deployment.

From a cybersecurity perspective, teams adopting AI-assisted UI development should treat generated front-end code as untrusted until it has been reviewed.

Developers should check authentication flows, input validation, client-side data exposure, API endpoint handling, dependency changes, and secret-management practices.

Visual prototypes can also unintentionally normalize insecure patterns, such as displaying sensitive identifiers, verbose error messages, or operational data without appropriate access controls.

According to a post from ClaudeDevs on X, the /design feature is available to Claude Pro, Max, Team, and Enterprise subscribers, who may need to update Claude Code to access it.

The feature signals a broader shift toward AI development environments that combine planning, design, coding, and iteration into a single workflow.

For engineering teams, the value will depend on whether the generated artboards remain easy to edit, accurately reflect requirements, and translate into maintainable, secure production code.

 Strengthen Your SOC by Accelerating Threat Detection & Rapid Investigations. -> Integrate ANY.RUN With Your SOC Now.