Overview
Claude Tag is a new way for teams to work with Claude. Tag @Claude in a Slack channel to delegate a request, and it completes the work autonomously while keeping context across the channel’s conversations. Instead of each person working with Claude in a separate tool, your team hands off work where it already talks — and sees the results together.
Key features
-
@Claude is multiplayer
A one-on-one assistant stays locked in each person’s own conversation, so teams can’t easily share progress. With Claude Tag, everyone in a channel interacts with the same Claude, so teammates can see the work it’s doing and build on it.
-
@Claude learns over time
To cut down on re-explaining background every time, Claude Tag builds up context from channel history and connected data sources. It picks up where the team left off without repeated explanations.
-
@Claude takes initiative
When ambient behavior is enabled, Claude Tag surfaces relevant information and follows up on unresolved threads. It steps in when it’s useful, without you having to call on it each time.
-
@Claude works asynchronously
Hand off longer-running tasks and Claude Tag completes them autonomously. You can also schedule work to run at a set time, so you don’t have to keep a conversation open waiting on results.
-
Per-channel access control and separate identities
Administrators define tool and data access per channel. This lets you create separate Claude identities for different uses, granting access that matches each channel’s purpose.
Notes
- Beta for Claude Enterprise and Team — available today in beta on Slack, using Opus 4.8, with broader expansion planned.
- Replaces the existing Claude in Slack app — Claude Tag replaces the existing app; administrators can opt in within 30 days to migrate, and launch credits are issued to eligible Enterprise and Team organizations.
- Doesn’t report from private channels — Claude Tag doesn’t report from private channels, and memories stay scoped to their channels.
- Spend limits — administrators can set limits for token spend, both for the organization and for individual channels.