Anthropic brings live artifacts to Claude Code for Team and Enterprise users

Anthropic added artifacts to Claude Code on June 18, giving Team and Enterprise users a way to turn a coding-agent session into a live, private web page rather than another terminal transcript. The beta lets Claude Code publish PR walkthroughs, system explainers, dashboards, release checklists, and incident pages that update as the session continues. 1

The new feature lets Claude Code capture work progress as an artifact: a live, shareable visual page built from the session's context. Anthropic says that context can include the codebase, connected tools, and the conversation, so an incident page could combine a failing test, the related function, monitoring data, and Claude Code's root-cause reasoning in one browser view. 1

The page is not static output. When Claude Code republishes an artifact, the same link refreshes in place for viewers, with a version history and a gallery for managing created artifacts. 1 Anthropic's examples are aimed at teams that already use Claude Code inside engineering workflows: PR explanations for reviewers, license audits for legal teams, security findings linked to exact lines of code, cloud-cost maps from Terraform, and incident timelines for SREs. 1

Availability is limited. Artifacts are in beta for Claude Team and Enterprise organizations, from the Claude Code CLI and desktop app, with pages viewable in a browser. 1 The Claude Code documentation says artifacts require a Team or Enterprise plan and a session signed in with /login; if the feature is unavailable, Claude Code will write a local HTML file or say it cannot publish. 2

This is a product capability update, not a model release. It matters because it moves Claude Code's output into a reviewable team surface. A coding agent can already inspect code, run tools, and produce long explanations; artifacts give that work a persistent page with the current version, prior versions, and organization-scoped sharing. 2

The security posture is also part of the product design. Anthropic says every artifact is private to its author by default, can only be shared with authenticated members of the same organization, and cannot be made public. Admins can manage access with an organization-level toggle, role-based scoping, retention policies, and compliance API visibility. 1

There are practical limits. The docs describe artifacts as one self-contained page with no backend, no external runtime requests, and a 16 MiB rendered-size ceiling. 2 That makes the feature better suited to work-in-progress explanations, dashboards, and review packets than to replacing internal apps. For enterprise buyers, the signal is that Anthropic is turning Claude Code from a solo terminal agent into a collaborative workflow product.