Anthropic has extended its Artifacts feature to Claude Code, the company's AI coding tool.
According to The Decoder, Claude Code can now turn the results of a coding session into interactive web pages called "artifacts" that teams can share with one another.
The Decoder reports that these pages draw on the full context of a working session, rather than just an isolated snippet. That means an artifact reflects what the AI assistant actually did and produced during the work, not a manually copied excerpt.
Two features stand out. First, the pages update automatically when something changes, so a shared link stays current instead of going stale the moment the underlying work moves on. Second, each artifact keeps a version history, giving teams a record of how the output evolved over time.
Artifacts began as a way to view and interact with content Claude generates in a single space. Bringing the feature into Claude Code aims it squarely at software teams, where the output of a session — code, summaries, or analysis — often needs to be passed around and reviewed by people who weren't in the session themselves.
Why it matters: sharing the output of AI coding work has been an awkward, copy-and-paste affair, and a live, self-updating, version-tracked page lowers the friction of turning one person's AI session into something a whole team can actually use.