How to Connect to the Slack MCP with Claude Code (5 steps)

By connecting to Merge Agent Handler's Slack MCP server via Claude Code, your agents can send Slack messages, search conversations, post updates, and more directly from the terminal.

To help you start using the server in your terminal, we'll cover each implementation step in detail.

Prerequisites

Before getting started, you’ll need:

• Merge Agent Handler account
• Claude Code installed and configured
• Slack workspace with appropriate permissions
• Tool Pack in Merge Agent Handler with Slack connector enabled
• Registered User in Merge Agent Handler                                                                                                                                                                           
• API key from Merge Agent Handler

How It Works

Merge Agent Handler acts as a secure bridge between your agent and Slack's API. You'll configure a Tool Pack (a scoped collection of connectors) and a Registered User (your agent's identity), which together generate a unique MCP URL that Claude Code uses over HTTP.

This URL, coupled with an auth token you'd generate in Agent Handler, lets you connect to our Slack MCP. Here’s what the full code block can look like:

{
 "mcpServers": {
   "agent-handler": {
     "command": "npx",
     "args": [
       "-y",
       "mcp-remote@latest",
       "https://ah-api-develop.merge.dev/api/v1/tool-packs/{TOOL_PACK_ID}/registered-users/{REGISTERED_USER_ID}/mcp",
       "--header",
       "Authorization: Bearer ${AUTH_TOKEN}"
     ],
     "env": {
       "AUTH_TOKEN": "yMt*****"
     }
   }
 }
}

Here’s how you can get the two custom items (auth token and MCP URL) and complete the connection flow.

Steps for connecting to the Slack MCP via Claude Code

Here's every step you'll need to take.

1. Create a Tool Pack

Log into the Merge Agent Handler dashboard, navigate to Tool Packs, and click Create Tool Pack. Give it a name like "Communications Agent," add a description, and select Slack as your connector. 

2. Add a Registered User

Go to Registered Users and click Add Registered User. This creates the identity context for your agent's operations and generates the unique MCP URL Claude Code will use.

3. Authenticate Slack

From the Registered User detail page, click Add Connector and complete the Slack OAuth flow. This grants your agent access to the workspace channels and conversations you specify.

4. Gather your credentials 

You'll need two things from Merge Agent Handler:

• MCP URL (from the Tool Pack detail page): It should look like the following but have your unique Registered User ID and Tool Pack ID

https://ah-api.merge.dev/api/v1/tool-packs/{TOOL_PACK_ID}/registered-users/{REGISTERED_USER_ID}/mcp

• Access token (from Settings > API Keys > Create new key)

5. Configure Claude Code

Run the following command in your terminal, replacing the placeholders with your actual values: 

Verify the connection by running: <code class="blog_inline-code">claude mcp list</code>. You should see agent-handler listed with a connected status.

Once connected, you can prompt Claude Code to take action in Slack. For example: “Search #engineering for any recent discussion about the auth service timeout issue and summarize what the team has already tried.” Or “Post a message to #deployments letting the team know the v2.4.1 release is live and link to the changelog.” 

{{this-blog-only-cta}}

Slack MCP FAQ 

In case you have any more questions on setting up and using a Slack MCP server in Claude Code, we’ve addressed several more questions below.

What can agents do with a Slack MCP server?

Here are just a few common use cases:

• Support triage in shared channels: An agent monitors/reads new posts in #support-style channels, classifies urgency/category, pulls the right SME into the thread, and posts a suggested first response (and/or creates the downstream ticket and drops the link back in-thread)
  

• Incident comms + stakeholder updates: In an incident channel, an agent aggregates latest context from the thread, produces periodic exec-ready status updates, and posts them to #leadership/#status while keeping a running timeline for the postmortem
  

• Sales deal desk / approvals in Slack: Reps ask for discount/terms/exception approval in a thread; the agent collects required fields, routes for approval, and once approved posts the final decision + next steps (and records it to the system of record)
  

• Recruiting coordination: In a hiring channel, an agent pulls candidate context, prompts interviewers for feedback if missing, summarizes feedback for the hiring manager, and posts an “inclined/declined + rationale” recap to keep loops moving
  

• IT/helpdesk requests via Slack → tracked workflow: An employee requests access/new software in Slack; the agent gathers details, checks policy, creates the request in the appropriate system, and posts status changes back to the original Slack thread until completion     

Why should I use Merge Agent Handler's Slack MCP server?

Merge Agent Handler's Slack connector includes 50+ tools with enterprise-grade security controls, data loss prevention, and real-time observability over every tool call. 

You can scope exactly which channels and actions your agent can access, and apply custom guardrails so agents never post to the wrong place.

Why connect Slack MCP to Claude Code?

Engineers already live in the terminal. Connecting Slack to Claude Code means your agent can pull conversation context before writing code, post status updates when tasks complete, and surface team knowledge from Slack history, all without leaving your development environment.