Slack MCP SERVER

🚀 Slack AI MCP Server

The Slack AI MCP Server connects AI assistants and automation systems directly to Slack workspaces using the Model Context Protocol (MCP). This Actor enables AI models to read, send, and analyze Slack messages, manage channels, and interact with workspace data — all through standardized MCP interfaces.

It bridges Slack’s API with Apify’s automation ecosystem, allowing you to build intelligent ChatOps workflows, AI assistants, and conversational agents that can operate securely inside your organization’s Slack.

🧠 Key Features

✅ Real-time Slack interaction Read and send messages, react with emojis, and reply in threads directly through MCP tools.

✅ Comprehensive workspace access List channels, view user profiles, read message history, and search workspace conversations.

✅ Automated AI workflows Trigger actions based on keyword detection or message patterns (e.g., respond to “deploy” automatically).

✅ Secure authentication Authenticate with your Slack workspace using a Bot User OAuth Token (xoxb-...), securely stored locally or via environment variables.

✅ Standardized MCP protocol Integrates with Claude Desktop, LangGraph, and other AI orchestration platforms for seamless AI ↔ Slack connectivity.

✅ Scalable multi-channel architecture Supports multiple concurrent AI agents across many Slack channels.

🧩 Tools Exposed by MCP Server

Tool	Description
slack_auth_prompt	Guides user to create a Slack app and obtain the Bot User OAuth Token.
save_slack_bot_token(bot_token)	Stores and activates a Slack Bot Token (xoxb-...).
slack_status	Checks if the saved token is valid and displays bot identity info.
list_channels(limit, types)	Lists all public/private channels the bot can access.
create_channel(name, is_private)	Creates a new Slack channel.
invite_user(channel_id, user_id)	Invites a user into a channel by ID.
post_message(channel_id, text, thread_ts)	Sends or replies to a message in a channel.
add_reaction(channel_id, timestamp, emoji)	Adds an emoji reaction to a message.
read_channel_history(channel_id, limit, oldest)	Fetches recent messages from a channel.
search_messages(query, count)	Searches across workspace messages.
get_user_profile(user_id)	Returns profile info for a user.
detect_and_reply(channel_id, keyword, reply_text)	Scans messages and replies to any containing a specific keyword.

Demo video for Slack MCP Server

⚙️ Input Parameters

Field	Type	Editor	Description
slack_bot_token	string	password	Your Slack Bot User OAuth Token (starts with xoxb-...).
slack_app_token	string	password	(Optional) Slack App-Level Token for advanced Socket Mode usage.

Example Input

{
  "slack_bot_token": "xoxb-1234567890-abcdef123456",
  "slack_app_token": ""
}

🔑 How to Get Your Slack Bot Token

1️⃣ Go to https://api.slack.com/apps 2️⃣ Create a new app → “From Scratch” 3️⃣ Under OAuth & Permissions, add these Bot Token Scopes:

• chat:write
• channels:read, channels:history
• groups:read, im:read, users:read
• reactions:write, search:read 4️⃣ Click Install to Workspace and Authorize 5️⃣ Copy the Bot User OAuth Token (xoxb-...) 6️⃣ Paste it in your Apify Actor input or call:

save_slack_bot_token("xoxb-...")

🧰 Required Slack Scopes

Category	Scopes
Reading	channels:read, channels:history, groups:read, im:read, users:read, search:read
Writing	chat:write, reactions:write
Managing	channels:manage, groups:write (optional)

📌 Grant scopes in Slack App → OAuth & Permissions → Bot Token Scopes → Save & Reinstall App.

🧠 Example Workflows

💬 Send and Read Messages

post_message(channel_id="C12345678", text="Hello team, this is your AI assistant!")
read_channel_history(channel_id="C12345678", limit=20)

🔍 Keyword Automation

Auto-reply to messages that mention “deploy”:

detect_and_reply(channel_id="C12345678", keyword="deploy", reply_text="✅ Deployment acknowledged.")

😄 Add Reactions

add_reaction(channel_id="C12345678", timestamp="1722234455.002300", emoji="thumbsup")

📚 Get Channel and User Info

list_channels(limit=50)
get_user_profile(user_id="U045XYZ12")

🧠 Example Output

{
  "ok": true,
  "bot_user": "apify-mcp-bot",
  "team": "Acme Corp",
  "channels": [
    {"id": "C04ABCD12", "name": "general"},
    {"id": "C05EFGH34", "name": "devops"}
  ]
}

⚙️ MCP Configuration Example

To connect this Actor to Claude Desktop or any MCP client, add to your config:

{
  "mcpServers": {
    "slack-mcp-server": {
      "url": "https://bhansalisoft--slack-mcp-server.apify.actor/mcp"
    }
  }
}

Once connected, you’ll see all Slack MCP tools automatically listed in your client.

🔄 Typical Usage Flow

1️⃣ slack_auth_prompt → instructions for creating a Slack app 2️⃣ save_slack_bot_token("xoxb-...") → authenticate 3️⃣ slack_status → confirm token validity 4️⃣ list_channels → choose a target channel 5️⃣ post_message → send AI-generated messages 6️⃣ detect_and_reply → automate responses or trigger workflows

🧾 Example Environment Variables

Key	Description
SLACK_BOT_TOKEN	Bot User OAuth Token (xoxb-...)
SLACK_APP_TOKEN	Optional App-level Token (xapp-...)

Set them in your Apify Actor Settings → Environment Variables, or via MCP tools.

💡 Benefits

• Integrate AI directly into Slack workflows
• Enable ChatOps with automated responses
• Reduce manual communication tasks
• Standardized interface via MCP — works with Gemini, Claude, or OpenAI models
• Scalable across multiple channels, bots, or workspaces

🧠 Example Use Case: AI DevOps Assistant

Use your AI agent to automatically:

• Watch the #deployments channel
• Detect words like “deployed”, “rollback”, “error”
• Send Slack replies summarizing next steps via Gemini AI
• Trigger Jenkins or GitHub actions from Slack messages (via Apify actors)

🧾 Example Log Output

✅ Slack MCP Server started successfully
✅ Passing SLACK_BOT_TOKEN to subprocess
✅ Connected as @apify-mcp-bot
✅ Listening for MCP requests...

📚 Learn More

• Slack API: OAuth & Permissions
• Apify Model Context Protocol Docs
• Slack SDK for Python
• Apify Actors & Python SDK