Slack MCP Server: Let AI Manage Your Workspace

In this guide

1. What a Slack MCP server does
2. Use cases by team
3. How to set it up
4. Available tools
5. Security and permissions
6. FAQ

What a Slack MCP server does

Slack is where work conversations happen, but finding information in Slack is painful. Important decisions get buried in threads. Updates span dozens of channels. A Slack MCP server gives your AI assistant direct access to all of it.

• "What did the engineering team decide about the migration timeline?" — searches relevant channels and threads, summarizes the decision
• "Post in #sales-alerts: 3 deals over $100K closing this week" — sends a formatted message to the channel
• "Summarize what I missed in #product today" — pulls the day's messages and gives you the highlights
• "Find every message mentioning the Acme deal in the last 2 weeks" — searches across channels and returns context
• "Send a DM to Sarah: can you review the proposal by Friday?" — sends a direct message
• "What questions are customers asking in #support this week?" — analyzes support channel activity

Use cases by team

Sales

Post deal alerts to team channels automatically. Search for customer mentions across all channels to get context before calls. Generate end-of-day pipeline summaries and post them to #sales-standup.

Engineering

Search for deployment updates and incident discussions. Post automated alerts when builds fail or monitoring thresholds are crossed. Summarize long technical threads into actionable decisions.

Management

Get daily summaries from channels you can't keep up with. Search for specific topics across the workspace without scrolling through history. Post weekly updates to multiple channels simultaneously.

Customer support

Search for similar issues that were previously discussed and resolved. Escalate tickets to engineering channels with full context. Track recurring themes in support conversations.

How to set it up

Step 1: Create a Slack App

Go to api.slack.com/apps and create a new app. Under OAuth & Permissions, add the scopes you need:

• Read-only start: channels:history, channels:read, search:read, users:read
• If you want to send messages: add chat:write
• For DMs: add im:write, im:history

Install the app to your workspace and copy the Bot User OAuth Token (xoxb-...).

Step 2: Add to your AI client

{
  "mcpServers": {
    "slack": {
      "command": "npx",
      "args": ["-y", "@anthropic/mcp-server-slack"],
      "env": {
        "SLACK_BOT_TOKEN": "xoxb-your-token",
        "SLACK_TEAM_ID": "T0123456789"
      }
    }
  }
}

Find your Team ID in Slack: click your workspace name → Settings → scroll to the bottom.

Step 3: Test it

Restart your AI client and ask "What channels am I in on Slack?" or "What's the latest message in #general?"

Available tools

• search — search messages across the entire workspace
• list_channels — get all channels the bot can access
• get_channel_history — read recent messages from a channel
• get_thread_replies — read a full thread
• post_message — send a message to a channel
• reply_to_thread — reply in a thread
• get_users — list workspace members
• get_user_profile — get details about a specific user

Security and permissions

Slack MCP access is controlled by your Slack App's OAuth scopes. Best practices:

• Start read-only. Add chat:write only when you're comfortable with AI sending messages on your behalf.
• Limit channel access. By default, the bot can only see channels it's been invited to. Don't add it to sensitive channels unless needed.
• Private channels require explicit invite. The bot can't read private channels unless someone adds it.
• Audit with Slack's API logs. Slack tracks all API calls — review periodically.
• Be careful with chat:write. AI-sent messages appear as the bot, not as you. Make sure your team knows.

Related guides

• What Is an MCP Server? — start here if MCP is new to you
• HubSpot MCP Server — connect AI to your CRM
• Notion MCP Server — AI-powered workspace and docs