Install guide and config at curatedmcp.com

Zendesk MCP: Give Your AI Agent Direct Access to Your Support Tickets

If you're running a support team and tired of manually feeding AI context about tickets, Zendesk MCP is the bridge you've been waiting for. This official Zendesk integration lets Claude, Cursor, or Windsurf read and modify tickets, pull customer history, and execute support workflows—all without leaving your AI agent's conversation.

What It Does

Zendesk MCP plugs your support system directly into AI agents. Instead of copy-pasting ticket details or manually searching your knowledge base, your AI can:

• Search and retrieve tickets with full context about the customer, conversation history, and current status
• Create and update tickets programmatically—set priorities, assign groups, add tags, and update SLA targets
• Access customer profiles to understand account history, organization details, and past interactions
• Add internal notes and public replies so the AI can help draft responses or document troubleshooting steps
• Query automation status—inspect macros, triggers, and SLA metrics to understand your workflows
• Generate summaries of long support conversations for faster handoffs

The real value: your AI agent can now reason about your actual support queue in real time, not hypothetically. That means smarter ticket routing suggestions, faster customer context retrieval, and the ability to prototype support automation without custom scripts.

How to Install

Install the server via npm:

npx -y @zendesk/mcp

Then add it to your Claude Desktop config:

{
  "mcpServers": {
    "zendesk-mcp": {
      "command": "npx -y @zendesk/mcp",
      "env": {
        "ZENDESK_SUBDOMAIN": "your-subdomain",
        "ZENDESK_EMAIL": "[e-mail]",
        "ZENDESK_API_TOKEN": "your-api-token"
      }
    }
  }
}

Grab your API token from Zendesk Admin → Apps and integrations → APIs.

Real-World Use Cases

• Ticket triage automation: Claude reads incoming tickets, searches similar past issues, and suggests priority levels and groups based on your SLA policies.
• Customer context in seconds: A support agent asks Claude about a customer's account. The agent retrieves full conversation history, organization details, and previous ticket outcomes—no tab switching.
• Draft responses at scale: Have Claude analyze a backlog of tickets, draft public responses for common issues, and flag ones that need human judgment before posting.

Full install guides for Claude Desktop, Cursor, Windsurf, and more at CuratedMCP.