How to Use Slack MCP Server with Your AI

Slack is where work conversations happen, but it is also where hours of your day disappear. Scrolling through channels, catching up on threads, drafting messages. The Slack MCP server gives your AI assistant direct access to your Slack workspace, so you can read, search, and send messages without ever opening the Slack app.

This is not about replacing Slack. It is about making your interaction with it faster and more intentional.

What It Does

The Slack MCP server connects your AI to the Slack API, enabling it to send messages to channels and users, search message history across your workspace, list channels and their members, read recent messages from specific channels, react to messages with emoji, and manage channel topics and purposes.

Your AI becomes a Slack power user on your behalf. It can catch you up on what you missed, draft and send updates, and search for that one message from three weeks ago that you vaguely remember.

Prerequisites

• A Slack workspace where you have admin or app installation permissions.
• A Slack Bot Token. You will create a Slack App and generate a Bot User OAuth Token with the right scopes.
• Node.js 18 or later installed on your machine.
• An MCP-compatible AI client such as Claude Desktop, Cursor, or Windsurf.

Step-by-Step Setup

1. Create a Slack App

Go to api.slack.com/apps and click "Create New App." Choose "From scratch" and give it a name like "AI Assistant." Select your workspace.

2. Configure Bot Permissions

Under "OAuth & Permissions," add the following Bot Token Scopes: channels:history, channels:read, chat:write, groups:history, groups:read, im:history, im:read, search:read, users:read, and reactions:write. Install the app to your workspace and copy the Bot User OAuth Token that starts with xoxb-.

3. Add the Server Configuration

In your AI client's MCP configuration, add:

json{
  "mcpServers": {
    "slack": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-slack"],
      "env": {
        "SLACK_BOT_TOKEN": "xoxb-your-token-here"
      }
    }
  }
}

4. Invite the Bot to Channels

Your Slack bot needs to be invited to any channel it should have access to. In each relevant channel, type /invite @YourBotName.

5. Restart Your AI Client

Save the configuration and restart your AI client to activate the Slack tools.

Things to Try

Once set up, here are practical ways to use it throughout your day:

• "Catch me up on the #engineering channel from the last 24 hours." Your AI reads recent messages and gives you a summarized digest of what was discussed, decisions made, and action items mentioned.

• "Send a message to #general saying 'Deploying v2.4 to production at 3pm EST. Expect 5 minutes of downtime.'" Draft and send messages without context-switching.

• "Search Slack for any messages mentioning the pricing change from last month." Full-text search across your workspace history, with your AI summarizing the relevant results.

• "Who has been most active in #design-feedback this week?" Get a quick pulse on channel activity and participation.

• "Draft a project update for #product-updates summarizing what our team shipped this sprint." Give your AI the context, and it will compose and send a polished update on your behalf.

Tips and Tricks

Start with read-only. If you are cautious about your AI sending messages on your behalf, begin with just the read and search scopes. You can add chat:write later once you are comfortable.

Use it for morning standup prep. Before your daily standup, ask your AI to summarize key conversations from overnight. It saves you from scrolling through dozens of messages.

Combine searching with action. Ask your AI to "find the thread where we discussed the new pricing tiers and summarize the final decision." It can search, read the thread, and distill the key points.

Be mindful of private channels. The bot only sees channels it has been invited to. If you need access to private channels, make sure to invite the bot there as well.

Tools That Pair Well

The Slack MCP server becomes a communication hub when combined with other a-gnt tools. Use the GitHub MCP server to get PR notifications summarized and posted to a Slack channel. Pair it with NNotion to pull meeting notes and share them directly in Slack. The Linear MCP server can feed project updates into your team channels automatically.

Find It on a-gnt

Explore the full listing for the Slack MCP server on a-gnt for additional configuration options and community guides.