Claude Tag: Persistent, Multiplayer AI Teammate for Slack

The article announces Claude Tag, a new product from Anthropic that lets teams integrate Claude directly into Slack channels as a collaborative team member. The core problem it addresses is the friction of working with AI in isolated chat windows or single-user sessions, where context resets and coordination with a human team is clunky. Claude Tag aims to make AI interaction feel more like working with a proactive, asynchronous teammate who builds shared context over time, rather than a stateless tool that requires re-explanation for every task.

Technically, Claude Tag operates as a persistent, multiplayer agent within Slack. It has ambient awareness of channel conversations, can be tagged with requests, and breaks down complex tasks autonomously, reporting results in threads. It maintains a single shared context per channel, remembers information across interactions, and can schedule future work. System administrators have granular control over which tools, data sources, and channels Claude can access, with strict scoping to prevent cross-contamination between, say, sales and engineering use cases. The product is initially available in beta for Claude Enterprise and Team customers, replacing the older Claude in Slack app, and runs on Opus 4.8.

For serious builders, the key takeaway is the operational shift toward delegation as a team workflow. Anthropic reports that 65% of its own product team’s code is now created via their internal version of Claude Tag, and the usage pattern is spreading beyond engineering to product metrics, support tickets, and debugging. This suggests the real value isn’t just in coding assistance, but in embedding a persistent, asynchronous AI worker into the organization’s communication fabric. The tight permission controls and channel-scoped memory are critical details for any team considering deployment in sensitive or compliance-heavy environments.