Claude Tag: Anthropic’s New Way to Work with AI in Slack

Anthropic just dropped something interesting: Claude Tag. It’s a new way for teams to work with Claude, starting on Slack. Think of it as Claude becoming an actual team member you can @mention in channels.

The idea is straightforward. You add Claude to selected Slack channels, connect it to your tools, data, even codebases. Then anyone in the channel can tag @Claude and delegate tasks while they focus on other stuff. Claude remembers context from the channels it’s in and can plan out future tasks. It’s basically a proactive, persistent AI colleague.

I’ve been testing this internally at Anthropic, and honestly, it’s become one of the main ways we get things done. Today, 65% of our product team’s code is created by our internal version of Claude Tag. That’s not a vanity metric—it’s real. And it’s spreading beyond engineering. We’re tagging Claude to chase down product metrics, work through support tickets, or help find the root cause of tricky bugs.

How it works

If you’ve used Claude Code or Cowork, this will feel familiar. Tag @Claude with a request in plain English, and it breaks the task into stages, works through them using its connected tools, and responds in a Slack thread when done. But tagging brings some new tricks.

Multiplayer. Within a channel, there’s one Claude that interacts with everyone. Anyone can see what it’s working on and pick up where someone left off. It’s collaborative, not siloed.

Learns over time. Claude builds context as it follows along in channels. You don’t have to explain things from scratch every time. It can even learn from other channels and data sources if you let it—though it won’t report from private channels. That tacit knowledge is a big deal for consistency.

Takes initiative. If “ambient” mode is on, Claude proactively flags relevant info from across channels and tools. It’ll follow up on threads or tasks that have gone quiet without resolution. It’s like having a diligent junior dev who actually remembers things.

Works asynchronously. Set a task, walk away, and Claude handles it. It can even schedule tasks for itself over hours or days. At Anthropic, we now spend way more time delegating to multiple Claudes in parallel. It’s a shift in how we think about work.

You can also DM Claude for private responses using your personal tools and connectors.

Getting started

This is clearly designed for teams with real security concerns. Sysadmins control which tools and info Claude accesses in which channels. Think of it as creating separate Claude identities for different uses—sales won’t leak into engineering, and vice versa. Memories stay scoped to the channels defined by admins. You can set token spend limits per org or per channel, and view a log of everything @Claude did, including who requested it.

Claude Tag is available today in beta for Claude Enterprise and Team customers. It replaces the existing Claude in Slack app, and admins have 30 days to migrate. Anthropic is also offering an introductory launch credit to eligible organizations so the whole company can try it out. It runs on Opus 4.8.

To get started, you pair it with your Slack workspace, give Claude access to your tools, set a monthly spend limit, and test it in a private channel. That’s it.

I’m curious to see where this goes. Slack is a natural starting point, but if Anthropic expands this to other platforms, it could get really interesting. For now, if you’re already on Claude Enterprise or Team, it’s worth giving a shot. Just be ready for your team to start treating Claude like an actual colleague—because that’s kind of the point.