Slack MCP Server: Let Your AI Agent Read and Send Messages (2026)

The Slack MCP server connects your AI coding agent to your team's Slack workspace. Your agent can read channel messages, search conversations, post updates, and react to threads through the Model Context Protocol.

Quick Answer: The Slack MCP server connects an AI agent to Slack to read and send messages, enabling the agent to gather context from team discussions for tasks like coding, summarizing channels, posting updates, and automating standup or deployment reports.

Why Connect an AI Agent to Slack?

The most common use case isn't chatbots. It's context. Developers discuss decisions, share requirements, and provide feedback in Slack. When your AI agent can read those conversations, it has the context it needs to write better code.

Practical examples:

• "Read the #frontend channel and summarize what the team decided about the new design system"
• "Post a deployment summary to #releases after I finish this migration"
• "Search Slack for any discussion about the authentication refactor"

Recommended skills

Offer Gravity Lab

by Bani Berisha

Transform messy product notes into rigorous positioning, message maps, and campa…

$20

The TweetClaw

by Martin Gunderman

Professional X/Twitter automation for AI agents: Post, monitor, extract data, an…

$7

nex-whatsapp-business-notify

by Nex AI

Enterprise-grade WhatsApp Business Cloud API integration for Python with session…

$49

Browse all Slack skills

Installation

npm install -g @anthropic/mcp-slack

Configure in your Claude Code MCP settings:

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

You'll need to create a Slack app in your workspace and generate a bot token with appropriate permissions (channels:read, chat:write, search:read at minimum).

Key Workflows

Gathering Requirements

"Read the last 50 messages in #product-feedback and list the top feature requests."

The agent reads the channel, identifies themes, and produces a structured summary. This is useful before sprint planning or when prioritizing work.

Deployment Notifications

"Post to #deployments: 'Auth service v2.3 deployed to production. Changes: OAuth2.1 flow, session management refactor, rate limiting on /token endpoint.'"

Automate deployment announcements. The agent can generate release notes from your git history and post them to the right channel.

Standup Reports

"Read my recent commits and draft a standup update for #daily-standup."

The agent compiles what you've done, formats it as a standup update, and posts it. Less time spent writing updates, more time coding.

Context for Code Changes

"Search Slack for discussions about the payment flow redesign, then implement the approach the team agreed on."

This is where Slack MCP becomes powerful. The agent gathers context from team discussions, then uses that context to inform its coding decisions. Combined with the right skills, this creates an agent that understands not just the code but the reasoning behind it.

Permissions

Set up your Slack bot token with minimal required permissions:

• channels:read for reading public channels
• channels:history for reading message history
• chat:write for posting messages
• search:read for searching conversations
• users:read for resolving usernames

Avoid giving the bot admin permissions. The agent should be able to read and post, not manage channels or users.

Pairing with Skills

A documentation skill paired with the Slack MCP server lets your agent gather requirements from Slack and generate technical documentation. A code review skill lets it summarize PR discussions happening in Slack threads.

Browse skills on Agensi that complement the Slack MCP server.