If you work with an agent, you know the ritual afterward: you have to explain to your team what the thing actually did. That’s exactly the gap Anthropic closes with Artifacts in Claude Code. The idea: instead of reporting status at the end, the work itself becomes a shareable page.
The session becomes a web page
Claude Code can now capture work progress as an artifact — a living, interactive web page that updates while the session runs. That could be a PR walkthrough, a system explainer, a dashboard you can filter and sort, or a release checklist that checks itself off as work gets done.
The key part: the page is built from the full context of your session — your code, your connectors, and the conversation itself. An incident page can bring together the failing test and the function behind it from your code, the error spike from a connected monitoring tool, and the root-cause reasoning from the session. You don’t wire up data sources or stand up infrastructure — you ask for a page, and Claude Code builds it from what already exists.
Live for everyone, private to the org
When Claude Code updates an artifact, the open page refreshes in place — teammates see changes the moment they’re published. Every publish is a new version at the same link, with version history to restore and a gallery for every artifact you’ve made.
Anthropic’s favorite example from internal testing is debugging: an engineer kicks off an incident investigation before standup, Claude Code works through the logs and publishes a timeline with suspect commits and an error-rate chart. By the time standup begins, Claude has republished the page twice — and everyone is looking at the same view instead of asking to be walked through what the agent found.
Every artifact is private by default and viewable only by authenticated members of the organization — making it public is deliberately not possible. Admins control access with an org toggle and role-based scoping, set retention policies, and keep oversight through the compliance API. Artifacts is available in beta for Team and Enterprise orgs, from the CLI and desktop app.
My take
Artifacts hits a real pain point. The most expensive part of agentic work is often not the doing but the communicating of the result — that loop of “quickly explain what happened.” A page that updates itself and draws on the real session context turns agent output into something the team can use directly. The fact that it all stays org-internal and is visible through the compliance API shows Anthropic is thinking about this feature for the enterprise from day one — not as a nice gimmick.