Overview

The Zendesk MCP Server enables AI assistants to interact with the Zendesk customer support platform. It provides tools for retrieving and managing tickets, creating and reading comments, and accessing help center articles as a knowledge base — bringing support workflows directly into AI conversations.

Community Server:

Developed by reminia. Python-based, installable via uv/pip or Docker.

Key Features

🎫

Ticket Management

List, retrieve, create, and update Zendesk tickets with full control over status, priority, assignee, and custom fields.

💬

Comments & Responses

Retrieve all comments on a ticket and create new public or private responses.

📖

Knowledge Base Access

Access help center articles as a knowledge base resource for research and response drafting.

🧠

AI-Powered Prompts

Built-in prompts for ticket analysis and response drafting that leverage AI to understand and resolve issues.

Available Tools

Quick Reference

Tool	Purpose	Category
`get_tickets`	List tickets with pagination and sorting	Read
`get_ticket`	Retrieve a single ticket by ID	Read
`get_ticket_comments`	Get all comments on a ticket	Read
`create_ticket_comment`	Add a comment to a ticket	Write
`create_ticket`	Create a new support ticket	Write
`update_ticket`	Update ticket fields (status, priority, assignee)	Write

Detailed Usage

get_tickets▶

Fetch tickets with pagination support, sorting by created_at, updated_at, priority, or status.

use_mcp_tool({
  server_name: "zendesk",
  tool_name: "get_tickets",
  arguments: {
    page: 1,
    per_page: 25,
    sort_by: "created_at",
    sort_order: "desc"
  }
});

get_ticket▶

Retrieve a specific ticket by its ID, with full details including subject, status, priority, description, and timestamps.

use_mcp_tool({
  server_name: "zendesk",
  tool_name: "get_ticket",
  arguments: {
    ticket_id: 42
  }
});

create_ticket▶

Create a new ticket with subject, description, priority, type, tags, and custom fields.

use_mcp_tool({
  server_name: "zendesk",
  tool_name: "create_ticket",
  arguments: {
    subject: "Login issue on production",
    description: "Users unable to log in after latest deployment",
    priority: "high",
    type: "incident",
    tags: ["production", "login", "urgent"]
  }
});

update_ticket▶

Update ticket fields including status, priority, assignee, type, tags, and custom fields.

use_mcp_tool({
  server_name: "zendesk",
  tool_name: "update_ticket",
  arguments: {
    ticket_id: 42,
    status: "solved",
    priority: "high",
    assignee_id: 123
  }
});

create_ticket_comment▶

Add a public or private comment to an existing ticket.

use_mcp_tool({
  server_name: "zendesk",
  tool_name: "create_ticket_comment",
  arguments: {
    ticket_id: 42,
    comment: "We've identified the root cause and deployed a fix.",
    public: true
  }
});

get_ticket_comments▶

Retrieve all comments for a ticket, including public and internal notes.

use_mcp_tool({
  server_name: "zendesk",
  tool_name: "get_ticket_comments",
  arguments: {
    ticket_id: 42
  }
});

Prompts

The server includes two built-in prompts for AI-assisted support workflows:

analyze-ticket — Analyze a Zendesk ticket and provide a detailed breakdown of the issue
draft-ticket-response — Draft a response to a Zendesk ticket based on its context and available knowledge base articles

Resources

zendesk://knowledge-base — Access help center articles as a searchable knowledge base

Installation

Clone the repo and install:

git clone https://github.com/reminia/zendesk-mcp-server.git
cd zendesk-mcp-server
uv venv && uv pip install -e .

Configure in your MCP client:

{
  "mcpServers": {
    "zendesk": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/zendesk-mcp-server",
        "run",
        "zendesk"
      ]
    }
  }
}

Configuration

The server requires Zendesk credentials in a .env file. Copy .env.example and fill in your details:

ZENDESK_SUBDOMAIN=your-subdomain
ZENDESK_EMAIL=your-email@example.com
ZENDESK_API_TOKEN=your-api-token

Zendesk API Token:

Generate an API token in Zendesk under Admin > Apps and Integrations > Zendesk API.

Common Use Cases

Ticket Triage: List and prioritize tickets by status, priority, or creation date
Response Drafting: Use the built-in prompt to draft responses with knowledge base context
Ticket Resolution: Update ticket statuses, assignees, and priorities as issues are resolved
Customer Communication: Add comments and responses to tickets, both public and internal
Support Analytics: Analyze ticket patterns, volumes, and resolution times

Zendesk MCP Server