Anthropic has transformed Claude Code Artifacts into a workspace tool that publishes live, updating pages instead of static outputs, letting teams follow coding sessions in real time without pasting logs into Slack. The feature, announced on June 18, 2026, pairs with enterprise-managed authentication through Okta, making it practical for IT departments to govern how teams access connected tools like Linear, Asana, and Supabase.

What Makes Claude Code Artifacts Different From Ordinary Dashboards?

Claude Code Artifacts are not meant to replace permanent business intelligence dashboards or SaaS subscriptions. Instead, they solve a narrower, more immediate problem: recording what an AI coding agent is doing while it does it. When Claude Code publishes an artifact, the page updates at the same link as the session progresses, creating a shared working page that teammates can refresh without asking for a walkthrough.

The distinction matters because Claude Code Artifacts are different from ordinary Claude artifacts available in the web and desktop apps. Claude Code Artifacts are currently available only to Team and Enterprise plan holders through the Claude Code CLI and desktop app, with pages viewable only by members of the organization that published them. Ordinary Claude artifacts, by contrast, support Model Context Protocol (MCP) integration on Pro, Max, Team, and Enterprise plans.

How Are Teams Actually Using Live Artifacts in Practice?

Anthropic's own example illustrates the practical value. An engineer starts an incident investigation before a standup meeting. Claude Code publishes a timeline and an error-rate chart, with suspect commits laid out alongside, then republishes the artifact as the investigation progresses. By the time the meeting begins, everyone is looking at the same evolving page instead of asking for a walkthrough or reviewing a static screenshot from 18 minutes earlier.

The pages can include several types of content that teams actually need during active work:

• PR Walkthroughs: Reasoning, diffs, test results, and changes that occurred while Claude kept working on the pull request.
• Incident Timelines: Error-rate charts, suspect commits, and investigation progress that updates as the session continues.
• System Explainers: Documentation and context that evolves as the coding session uncovers new information.
• Release Checklists: Task lists and verification steps that update as Claude completes each item.

Why Enterprise Authentication Is the Real Story

The second piece of this announcement sits outside Claude Code Artifacts but explains why Anthropic is pushing this feature now. Model Context Protocol connectors only become normal workplace infrastructure if companies can govern them. Individual employees clicking through separate OAuth prompts for each tool is not a sustainable enterprise deployment.

Anthropic's enterprise-managed authentication lets admins authorize connectors once through an identity provider, with Okta supported at launch. Employees inherit access on first login, eliminating the friction of repeated authentication steps. The supported connectors at launch include Asana, Atlassian, Canva, Figma, Granola, Linear, and Supabase, with Slack coming soon.

This standards work represents the difference between a neat demo and something an IT department can actually let into the building. Okta's June 18 announcement names the same provider group, and joint customers Ramp, Webflow, and HubSpot are part of the beta program. The Model Context Protocol specification now includes a June 18, 2026 authorization section built around OAuth 2.1, protected resource metadata, dynamic client registration, and resource indicators.

How to Set Up Claude Code Artifacts for Your Team

• Check Your Plan: Claude Code Artifacts are available only to Team and Enterprise plan holders; Pro and Max users do not have access to this feature yet.
• Access via CLI or Desktop: Use the Claude Code CLI or desktop app to publish artifacts; pages are viewable in a browser by organization members only.
• Configure Enterprise Auth: If your organization uses Okta, work with your IT admin to authorize connectors once; employees inherit access on first login without repeated OAuth prompts.
• Connect Supported Tools: Link to Asana, Atlassian, Canva, Figma, Granola, Linear, Supabase, or Slack (when available) to pull context into your coding sessions.

What Problem Does This Actually Solve?

Cursor and Replit already know how to show a preview of what an AI coding tool just built. Anthropic is aiming at a slightly different use case: the small internal page that records what an agent is doing while it does it. For a founder, an engineering manager, or whoever is on call, that can be enough. You do not always need another dashboard subscription. Sometimes you just need the working page to keep up with the work.

The real value emerges when you have watched a long agent run. A session starts with an incident, a refactor, a cost review, or a stubborn failing test. Claude Code pulls from the repo and the conversation, plus whatever connected tools the session has, then publishes a page with the context a teammate actually needs. The page can update at the same link as the session progresses, eliminating the need to paste logs into Slack or narrate every step of a pull request.

This is not magic, and it is not a replacement for every internal tool. It is a shareable working page that evolves with the work, available now to Team and Enterprise users through the Claude Code CLI and desktop app.