$ today_in_10_seconds
Anthropic launched Claude Tag, an agentic AI teammate that lives inside your Slack channels — you @Claude it like a colleague and it breaks down tasks, uses connected tools, and remembers context across conversations. It replaces the old Claude in Slack app and is in beta for Enterprise and Team tiers. The pitch: AI as a standing team member, not a private chat window.
$ lead_story
Anthropic launches Claude Tag, a persistent AI coworker for Slack
Anthropic dropped Claude Tag today — an agentic feature that integrates directly into Slack, letting teams tag @Claude as a collaborative team member to delegate tasks, automate workflows, and build shared organizational context. The agentic part is what matters: when invoked, it breaks assigned tasks into stages, executes them using connected tools, and delivers results in a Slack thread — multi-step autonomy in the surface where work actually gets assigned. Anthropic describes it as an evolution of Claude Code, and the dogfooding numbers are loud: the company says 65% of its product team's code is created by its internal version, with uses spanning data lookup, support tickets, and bug finding. For builders, the form factor is the insight — being able to tag the agent the same way you would a coworker.
// why_it_matters: Agents just moved from a tab you open to a teammate you summon — and the Slack channel is now the battleground.
$ also_worth_attention
Each channel gets its own scoped Claude identity
Admins control exactly which tools, data, and channels each instance can touch — a sales-configured Claude won't share memory or data access with an engineering channel instance. Admins can set monthly token spend limits at org and channel level and audit a full log of every task Claude ran, including who requested it. For anyone running AI governance, this is the multi-tenant isolation story that makes autonomous agents deployable without a blast radius.
link →It's proactive, not just reactive
Claude Tag can read conversations, break down tasks, and proactively flag relevant updates across an organization without being asked. Cat Wu's own example: she gave Claude Tag access to Gmail so it reads her messages, flags when an important person writes, and posts to her on Slack. Ambient monitoring plus initiative is the line between an assistant and an agent — and it's now on by default in beta.
link →Slackbot, meet your replacement — migration clock is ticking
Claude Tag runs on Opus 4.8, replaces the existing Claude in Slack app, and gives admins a 30-day window to opt in before the legacy app is retired on August 3. Anthropic plans to expand the capability to other platforms in the coming weeks. If you're on Enterprise or Team, this isn't optional drift — it's a dated migration with launch credits attached.
link →
$ bigger_picture
The trend behind the headlines
The interesting shift isn't the model — it's the placement. Slack itself has been positioning as an 'agentic operating system,' and Salesforce announced more than 30 new Slackbot capabilities in March. Anthropic's counter is to make the agent multiplayer: shared identity, shared memory, ambient presence. Every channel that grants persistent access — connected tools, accumulated context, monitoring on — is a far deeper integration than a chatbot session, and the switching costs rise accordingly. The old question of 'where does AI fit in our stack' is being answered by the vendors planting flags in the collaboration layer first.
Claude Tag
The obvious pick. Grant it access to selected channels, connect your tools, data, and codebases, and anyone in the channel can tag @Claude to delegate. If you're evaluating, start where Anthropic recommends: pair with Slack, connect tools, set a spend limit, and test in a private channel before going live.
$ subscribe --free
Get the next edition in your inbox
// DAIly Agent covers agentic AI three times a week
// free, no spam, exit anytime