Slack MCP: The Official Server at mcp.slack.com
Yes — Slack has an official MCP server. It’s been generally available since February 2026, hosted at mcp.slack.com, and it lets MCP-compatible AI tools search your workspace, read messages and files, and post on your behalf. If you’re searching “Slack MCP,” the connection you want already exists and it’s Slack’s own.
The thing worth knowing before you set it up: an MCP server hands your workspace to an AI inside a conversation you start. It’s a doorway, not a worker. Nothing watches Slack for you, nothing fires when a message lands in a channel, and nothing runs while the chat is closed. Here’s exactly what the Slack MCP does, how to turn it on, where it stops — and what to use when you want Slack work that runs on its own.
Slack is one of many — the complete list of MCP servers covers every app that ships one, grouped by category.
What the Slack MCP server does
Model Context Protocol (MCP) is the open standard that lets an AI client — Claude, ChatGPT, Cursor, VS Code, and others — talk to an outside app through a shared interface. Slack maintains its own official remote server at mcp.slack.com/mcp, speaking JSON-RPC 2.0 over Streamable HTTP. With it connected, an MCP-compatible AI client can:
- Search messages and files — filtered by channel, user, date, or content, so a question like “what did the team decide about the Q3 launch” gets answered from real threads.
- Look up users and channels — pull user profiles, statuses, and channel metadata instead of guessing who’s who.
- Send messages — post to any conversation, draft formatted messages, and reply in a thread.
- Create and share canvases — turn a summary or a plan into a formatted Slack canvas without leaving the chat.
- Add reactions and create conversations — small actions that used to mean tabbing over to Slack.
For most people, this is what “Slack MCP” means: an AI that can read your workspace and act in it, one request at a time, grounded in what’s actually in your channels rather than a stale guess.
How to set up the Slack MCP server
Slack’s server is a hosted, no-code path for the handful of clients it partners with directly:
- In a supported AI client — Slack names Claude.ai, Claude Code, Cursor, and Perplexity as partners — add the Slack connector or point it at
https://mcp.slack.com/mcp. - Authorize it through Slack’s OAuth 2.0 flow (
slack.com/oauth/v2_user/authorize), granting the scopes the tools need — for examplechat:writeto let it send messages. - A workspace admin has to approve the integration. Only apps published in the Slack Marketplace, or internal apps, are allowed to use MCP at all — an unlisted app can’t connect this way.
- Once approved, start a chat and ask the AI to search, summarize, or post — the tools show up automatically in a supported client.
If your AI client isn’t one of Slack’s official partners, community MCP servers exist on GitHub that wrap Slack’s Web API directly, but those are third-party and carry their own token and permission model — not the same thing as the official mcp.slack.com server.
Where the Slack MCP stops
None of this is a knock on MCP — it’s just the shape of the protocol. Four limits show up the moment you want more than a conversation:
- It only works inside a chat you start. Close the window and nothing happens. The AI doesn’t watch Slack; it waits for you to ask.
- No triggers. A new message in
#support, someone getting @mentioned, a thread going quiet — none of these can start anything through MCP. There’s no “when this happens in Slack, do that.” - It’s one app at a time. The Slack MCP knows Slack. Getting a flagged message into a ticket in your helpdesk, a row in a sheet, and a summary email means wiring up a separate MCP server for each and hoping your client can juggle them in one turn.
- You own the approvals and the scopes. Workspace-admin sign-off, OAuth tokens, and the blast radius of an AI that can post as you are all on you.
So the Slack MCP is a great way to ask Slack things and send one-off messages. It is not a way to make Slack run — to have work happen on a schedule or in reaction to an event, across the other tools a conversation touches.
Running Slack work that doesn’t need a chat open
That “run on its own, across apps” gap is exactly where Carly fits. Carly connects to Slack natively — no MCP server to authorize, no admin approval flow to chase — and to the ~260 other apps it supports, plus anything with a public API through your own key. The difference from MCP is the important part: Carly’s workflows are triggered and scheduled, so Slack work happens whether or not anyone has a chat window open.
A few things that MCP can’t do but a Carly workflow can:
- When a customer message lands in
#support→ create a ticket in your helpdesk, tag it by urgency, and post the ticket link back to the thread — automatically, the moment it happens. - Every Friday afternoon → pull the week’s decisions out of a project channel and email a recap to the team.
- When a teammate is @mentioned in an urgent channel → check their calendar, and if they’re in a meeting, DM the sender an ETA instead of leaving them on read.
The non-AI steps — the moving, matching, and routing between apps — are free and unlimited, the Zapier-style backbone of the workflow. The AI steps (drafting, summarizing, deciding) start at $35/month. You describe the outcome in plain language and Carly wires up the Slack connection and everything downstream.
If you just want to search your workspace or send a message from a chat, Slack’s official MCP server is the right tool and it’s free to connect. If you want Slack to actually do things — on a trigger, on a schedule, across every app a conversation touches — that’s the job MCP wasn’t built for, and it’s the one Carly was.
FAQ
Does Slack have an official MCP server?
Yes. Slack’s own remote MCP server at mcp.slack.com/mcp has been generally available since February 2026, giving MCP-compatible AI tools search, messaging, and canvas access to a workspace.
Is the Slack MCP server free? Connecting to it is free — you’re authorizing an AI client against your existing Slack account and permissions. A workspace admin does need to approve the integration, since only Marketplace-listed or internal apps can use MCP.
Can the Slack MCP trigger automations? No. MCP is request/response inside an AI chat — it has no triggers and nothing runs when the conversation is closed. For event- or schedule-driven Slack work across apps, you need a workflow tool like Carly rather than an MCP server.
What AI tools can connect to Slack over MCP? Slack names Claude.ai, Claude Code, Cursor, and Perplexity as its official partner clients, with more expected. Other MCP-compatible clients can reach Slack through community servers, but those aren’t Slack’s own.
Can I connect Slack to AI without coding or hosting a server? Yes. You don’t have to touch MCP at all. Carly connects to Slack for you and lets you build the automation in plain language — describe what you want to happen and it wires up Slack and the other apps involved, with no server to host and no code to write.
Ready to automate your busywork?
Carly schedules, researches, and briefs you—so you can focus on what matters.
See what people say
"Before Carly, I relied on a Calendly link, but the whole process felt impersonal and not very professional. Carly changed that by handling all the back-and-forth, so I'm no longer stuck in endless email threads trying to line up schedules.
Now Carly reaches out to candidates, shares my real-time availability, lets them pick a slot, then sends a Zoom link and drops it straight into my calendar. She sends reminders to both of us before each call, which has significantly reduced no-shows and last-minute confusion.
On top of scheduling, Carly acts like a full executive assistant, sending me my schedule the night before so I can prepare for each call. It reminds me of the old x.ai assistant, but Carly is noticeably smarter, faster, and better suited to my healthcare recruitment business."
