Slack MCP for Your Team's AI Agent

Connect MCP servers to a shared AI agent in Slack. Use Runbear's managed catalog with OAuth or add your own server, and the agent calls those tools from the thread it is already in.
Or, book a demo

Teams already running Runbear

  • Matillion
  • Fireflies.ai
  • BCG
  • Minted
  • Mercor
  • Tabs
  • TaxDome
  • Peak Ace
  • Storykit
  • SYPartners
Slack Daily Digest

Native Slack Experience

Interact naturally with your AI right inside Slack, avoiding any disruption from switching apps.

Jira
Notion
Google Calendar

Managed and Custom Servers

Connect servers from Runbear's managed catalog with OAuth, or add any remote MCP server reachable over HTTPS with Streamable HTTP or SSE.

Flexible AI Model Selection

Choose Claude, GPT, or Gemini for each agent on the default Claude Agent SDK runtime.

Per-User or Shared Access

Connect a tool Per-User, so the agent acts with each member's own permissions, or Shared, with one account for everyone.

How MCP Works with Slack

Get started with MCP-enabled AI in Slack in three simple steps

  1. Set up your AI Agent

    Pick the model, then point the agent at the knowledge it should answer from: Google Drive, Confluence, Notion.

  2. Enable MCP Integrations

    Connect servers from the managed catalog, or add your own MCP server with its HTTPS URL, transport, and authentication.

  3. Interact in Slack

    Automate workflows, answer queries, and streamline operations directly from your Slack workspace.

Connect with Your Favorite Apps

Connect MCP to the apps the work already lives in like Notion, Jira, and GitHub.

Browse MCP Servers

Use Cases

AI Workflow Enhancements for Your Team

Customer Onboarding Nudge Bot: Pipeline Madness for CS

Customer onboarding automation for CS teams: an always-on Salesforce bot that nudges every new account through onboarding milestones with daily Slack standups. AI for customer success that surfaces stuck accounts on day one, not week three.

Automate Email Rental Orders into Your System with AI

Use Runbear to watch your Gmail inbox, parse rental orders with AI, push them into your rental management system via MCP, and notify your team in Slack—cutting order processing time from 15–30 minutes to under 2 minutes.

AI Executive Dashboard: KPI Reports Delivered to Slack

Get KPI dashboards and executive summaries delivered to Slack on demand. AI agents pull data from your connected business tools and generate visual reports with charts, trends, and anomaly alerts without logging into a dashboard.

Automate KPI Reporting: AI Agent with MCP for Slack

Set up AI agents that deliver live KPI summaries and data visualizations directly into Slack using MCP. Connect your data sources and get automated reports on the metrics that matter, on a schedule or on demand.

AI Article Summarizer for Slack

Share a URL in Slack and get an instant AI summary of the article or web page. Your team saves time by reading key takeaways instead of clicking through every link shared in channels.

AI Calendar Scheduling from Slack Using MCP

Schedule meetings, check availability, and manage your calendar directly from Slack using MCP. The AI agent reads your calendar context and handles coordination so you skip the back-and-forth.

Slack AI Agent That Joins Conversations Automatically

Build an AI agent for Slack that proactively joins conversations and provides answers before anyone asks. The agent monitors channels, detects questions, and responds with relevant information from your connected knowledge sources.

Build a Routing Chatbot for Workplace Questions in Slack

Create a chatbot that routes workplace questions to the right person or team in Slack. Instead of guessing who to ask, employees type their question and the bot directs it to the correct department or answers it automatically.

Writing Blog Posts with a Chatbot

Use a chatbot in Slack to draft blog posts from outlines, talking points, or raw notes. The AI writes initial drafts, suggests headlines, and helps iterate on content without switching to a separate writing tool.

AI-Powered Q&A for Healthcare Teams in Slack

Healthcare professionals can ask clinical reference questions and get answers from uploaded medical documents directly in their workspace. Speeds up research without leaving patient-facing workflows.

AI-Powered Q&A for Legal Teams in Slack

Legal professionals can ask case-related questions and get answers from uploaded documents directly in Slack. The AI agent searches your legal knowledge base and returns relevant precedents and policy details.

AI Compliance Review Assistant for Slack Teams

Connect a compliance review assistant to your team in Slack. The AI agent checks documents and processes against your compliance requirements and flags issues before they reach audit, saving hours of manual review.

A Bot That Summarizes Daily News on Selected Topics

Build a Slack bot that collects news articles on topics you choose and delivers a daily summary. The AI scans sources, filters by relevance, and posts concise briefings so your team stays informed without browsing multiple sites.

AI Assistants for Customer Support

Create custom AI assistants that help your customer support team answer questions faster in Slack, Teams, or HubSpot. The AI pulls from your docs and past tickets to draft accurate responses.

How to Add a Proofreading Assistant to Slack

Add a proofreading bot to your Slack workspace that checks grammar, spelling, and tone in real time. Paste any text and get corrections instantly, no coding required to set up.

Frequently Asked Questions

MCP (Model Context Protocol) is an open standard for connecting AI agents to external tools. "Slack MCP" usually means one of two things. Slack's own MCP server lets an AI client outside Slack, such as Claude or Cursor, search channels and send messages. An MCP client inside Slack works the other way: an agent that lives in your channels calls other MCP servers to look up data or take action. Runbear is the second kind.

Yes. As of October 2026, Slack's developer docs describe the Slack MCP server at https://mcp.slack.com/mcp. It uses Streamable HTTP, and each MCP client must be backed by a registered Slack app that is published in the Slack directory or internal to the workspace. Slack lists Claude.ai, Claude Code, Perplexity, and Cursor as partner clients.

As of October 2026, Slack's docs describe a Slackbot MCP Client that lets developers connect remote MCP servers to Slackbot through a Slack app. Servers must use Streamable HTTP, a user can have up to 5 active MCP servers at a time, and the user approves each third-party tool call, once or always, before Slackbot runs it. Runbear is an alternative when you want a team agent you configure in a dashboard, with its own instructions, knowledge sources, and tools, that runs in Slack and Microsoft Teams.

Open the agent's Tools tab. Pick a server from the managed catalog and complete its OAuth sign-in, or choose Add Custom MCP and enter a server name, an HTTPS server URL, the transport, and the authentication. Runbear performs a handshake and lists the tools it discovered, and the agent calls them when a message in Slack needs them.

Streamable HTTP, the recommended default, and SSE for older servers that use long-lived event streams. Authentication can be OAuth, a static header such as an API key or bearer token, or none, for servers protected another way such as an IP allowlist or VPN.

Not directly. Runbear runs in the cloud, so the server must be reachable over HTTPS. Run a local or STDIO server behind a tunnel that gives it a stable HTTPS URL, such as ngrok, Cloudflare Tunnel, or Tailscale Funnel, and pair the tunnel with static header authentication so only Runbear's requests are accepted.

For catalog servers and custom servers that use OAuth, you choose Per-User or Shared. Per-User means each member links their own account and the agent acts with that member's permissions. Shared means one account serves everyone on the agent, so every user inherits that account's access. A custom server with a static header uses that one credential for everyone.

Yes, with tool approvals, which Runbear is rolling out gradually. When an admin turns on Require Tool Approval, new MCP tools added to agents wait for admin approval before they can be used. Approvals can be narrowed to individual tools, so you can allow a server's read-only tools and withhold the ones that write. Contact support@runbear.io to enable it.

Start Enhancing Slack with MCP Today

Experience streamlined, powerful AI integration within your Slack workspace.

Want a walkthrough first? Book 20 minutes with our CEO