What Claude Tag is
Claude Tag is Anthropic's new way for teams to work with Claude, and it starts inside Slack. Once it is added to a workspace, anyone in a channel can tag @Claude and hand it a request, the same way you would tag a coworker. Instead of opening a separate app and copying context back and forth, you delegate the task right where the conversation already lives.
The shift is small to describe and large in practice. For two years the default way to use Claude was to start a session, give it context, and prompt it. Claude Tag flips that. Claude becomes a participant in the room your team already works in, watching the thread, holding the context, and waiting to be handed work.
What makes it different from a chatbot
Claude Tag is not a private chat window bolted onto Slack. Four properties make it behave like a teammate rather than a tool, and each one changes how you would actually use it.
- Multiplayer: one Claude instance lives in the channel and is visible to everyone, so the whole team shares the same context instead of each person briefing their own private bot.
- Learning: it builds context over time from the channel history and the data sources you connect, so it gets sharper about your project the longer it sits there.
- Proactive: when ambient behavior is enabled, it can flag relevant information on its own rather than only answering when asked.
- Asynchronous: it works independently and can schedule tasks for itself, pursuing a project autonomously over hours or days.
That last property is the one to sit with. An agent that can schedule its own follow-up work and run a project over days is the same idea behind running Claude Code on a loop. If that pattern is new to you, read [what are agent loops in Claude Code](/blog/claude-code-agent-loops) and [schedule Claude Code agents while you sleep](/blog/schedule-claude-code-agents-while-you-sleep), because Claude Tag is that capability brought into your team chat.
What it can do
Anthropic positions Claude Tag for real work, not novelty. The tasks it lists are the ones a team actually hands off: generating code, analyzing data and tracking metrics, managing support tickets, and investigating bugs down to root cause. You tag it, describe the job, and it runs.
Picture the loop in a single channel. A bug report lands, someone tags @Claude to investigate, it reads the thread and the connected repo, proposes a root cause, and drafts the fix. Later that day it checks back on a metric it was asked to watch. None of that required a person to leave Slack or re-explain the project. That is the appeal.
Who can use it today, and the honest caveats
Here is the part most hype posts skip. Claude Tag is in beta starting June 23, 2026, and it is limited to Claude Enterprise and Team customers. It runs on Opus 4.8. If you are a solo builder on an individual plan, you cannot tag @Claude in your Slack today. That is a real limit, not a detail to gloss over.
There are also sensible guardrails for the teams that do get it. Claude Tag replaces the existing Claude in Slack app with a 30-day opt-in window, so it is a migration, not a brand new install for everyone. Admins can restrict Claude's access by channel, set token spend limits, and view activity logs. If you have read our warnings about never letting an agent spend without a ceiling, you will recognize why those controls matter.
The standing-agent shift, and what it means for builders
We call the bigger pattern here the standing-agent shift, and it is the part worth internalizing even if you never get Claude Tag itself. The shift is from prompting an agent on demand to keeping a standing agent that lives where your work happens, holds context, and is ready to be handed a task without a fresh briefing every time.
You do not need an Enterprise seat to practice it. In the Claude Code desktop app you already have a standing agent in your project. The move is to stop treating each session as a cold start, write a CLAUDE.md so it keeps your context, and let it run loops on real jobs. Claude Tag is simply the team version of the habit our members build solo every day.
If you want to feel the difference now, read [Claude Code auto mode runs on its own](/blog/claude-code-auto-mode-runs-on-its-own) and then point a standing agent at one repetitive job in your week. The teams getting value from Claude Tag are not doing anything magic. They are delegating to an agent that already knows the project, and you can build that exact muscle today.
Start building the standing-agent habit
Inside Claude Code Club we teach the standing-agent shift from day one, with a community that reviews your setup and a course that turns it into a repeatable habit instead of a one-off experiment. Start with the [curriculum](/curriculum) or join us at https://www.skool.com/claudecodeclub/about and bring the one workflow you most want off your plate.
Frequently asked questions
What is Claude Tag?
Claude Tag is Anthropic's new way for teams to work with Claude, starting inside Slack. Anyone in a channel can tag @Claude and delegate a task to it. One Claude instance lives in the channel for the whole team, learns from the channel history and connected data sources, and can work asynchronously, including scheduling tasks for itself over hours or days.
Who can use Claude Tag?
As of the beta that began June 23, 2026, Claude Tag is available to Claude Enterprise and Team customers only, and it runs on Opus 4.8. Individual and solo plans do not have access yet. It also replaces the existing Claude in Slack app, with a 30-day opt-in window for teams migrating over.
How is Claude Tag different from using Claude normally?
Normal use means you start a session and prompt Claude on demand. Claude Tag is a standing agent that lives in your Slack channel, is visible to the whole team, holds context over time, can act proactively when ambient behavior is on, and can run tasks asynchronously. You delegate to it in the channel instead of opening a separate app.
Can I get the same benefit without Claude Tag?
Yes, in spirit. You cannot tag @Claude in Slack without an Enterprise or Team plan, but you can practice the same standing-agent habit in the Claude Code desktop app: keep a CLAUDE.md so the agent retains your context, stop treating each session as a cold start, and let it run loops on real jobs. That is the solo version of what Claude Tag does for teams.
Last reviewed by David Iya on June 29, 2026
