Skip to main content
Glama
zendesk-mcp-server

zendesk-mcp-server

by wlaubernds

Zendesk MCP Server

Model Context Protocol server that connects AI assistants to Zendesk, enabling natural language queries over support tickets, help articles, and customer feedback.

Search your Zendesk data using AI in Cursor, Claude Desktop, or any MCP-compatible tool.

npm version License: MIT

Features

  • 🔍 Search Support Tickets - Query tickets with advanced filters (status, tags, dates, priority)

  • 📚 Search Knowledge Base - Find relevant help articles and documentation

  • 📊 Feature Feedback Analysis - Get comprehensive feedback about specific features including bug reports, feature requests, and support questions

  • 🎫 Ticket Details - Retrieve full ticket information including all comments

  • 📖 Article Content - Access complete help center articles

  • 🏷️ Tag-based Search - Find all tickets with specific tags

Related MCP server: zendesk-mcp

Installation

No installation required! Just configure and use:

{
  "mcpServers": {
    "zendesk": {
      "command": "npx",
      "args": ["-y", "zendesk-mcp-server"],
      "env": {
        "ZENDESK_SUBDOMAIN": "your-company",
        "ZENDESK_EMAIL": "your-email@company.com",
        "ZENDESK_API_TOKEN": "your-api-token-here"
      }
    }
  }
}

Option 2: Install Globally

npm install -g zendesk-mcp-server

Then configure:

{
  "mcpServers": {
    "zendesk": {
      "command": "zendesk-mcp-server",
      "env": {
        "ZENDESK_SUBDOMAIN": "your-company",
        "ZENDESK_EMAIL": "your-email@company.com",
        "ZENDESK_API_TOKEN": "your-api-token-here"
      }
    }
  }
}

Option 3: Local Development

git clone https://github.com/wlaubernds/zendesk-mcp-server.git
cd zendesk-mcp-server
npm install
npm run build

Then use the local path in your config:

{
  "mcpServers": {
    "zendesk": {
      "command": "node",
      "args": ["/path/to/zendesk-mcp-server/dist/index.js"],
      "env": {
        "ZENDESK_SUBDOMAIN": "your-company",
        "ZENDESK_EMAIL": "your-email@company.com",
        "ZENDESK_API_TOKEN": "your-api-token-here"
      }
    }
  }
}

Getting Your Zendesk API Credentials

  1. Log in to your Zendesk Admin Center

  2. Navigate to Apps and integrations > APIs > Zendesk API

  3. Click the Settings tab

  4. Under Token Access, click Add API token

  5. Give it a description (e.g., "MCP Server")

  6. Copy the token (you'll only see it once!)

  7. Your subdomain is the first part of your Zendesk URL: https://YOUR-SUBDOMAIN.zendesk.com

Configuration

Add the configuration to your MCP settings file:

  • Cursor: ~/.cursor/mcp.json

  • Claude Desktop:

    • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

    • Windows: %APPDATA%\Claude\claude_desktop_config.json

See examples/cursor-config.json for a complete example.

Usage

Once configured, you can use natural language to query your Zendesk data:

Example Queries

Analyze feature feedback:

"Show me all feedback about our mobile app from the last 3 months"

Search tickets:

"Find all high priority bug reports from this week"
"What are users saying about the new dashboard?"

Weekend summaries:

"Summarize all tickets from this weekend"

Specific ticket details:

"Get full details and comments for ticket #12345"

Search help articles:

"Find all help articles about password reset"

Available Tools

The MCP server provides these tools that AI assistants can use:

Tool

Description

search_tickets

Search tickets with advanced query syntax (status, priority, tags, dates)

get_ticket

Get full details of a specific ticket including all comments

search_articles

Search help center articles

get_article

Get full content of a specific article

search_feature_feedback

Comprehensive analysis of feedback for a specific feature

get_tickets_by_tag

Get all tickets with a specific tag

Zendesk Query Syntax

When searching tickets, you can use these filters:

  • status: - new, open, pending, hold, solved, closed

  • priority: - low, normal, high, urgent

  • tags: - ticket tags

  • subject: - search in subject line

  • created> or created< - date filters (format: YYYY-MM-DD)

  • updated> or updated< - last updated date

  • Plain text for full-text search

Query Examples:

status:open tags:bug priority:high
tags:mobile created>2024-01-01
subject:"cannot login" priority:urgent
password reset

Permissions

🔒 This MCP server is READ-ONLY - it has zero write permissions to Zendesk.

✅ What it CAN do:

  • Search and read support tickets

  • Retrieve ticket details and comments

  • Search knowledge base articles

  • Analyze feature feedback

  • Query tickets by tags, status, priority, dates

❌ What it CANNOT do:

  • Create tickets

  • Update or modify tickets

  • Add comments to tickets

  • Change ticket status

  • Modify tags, priority, or assignments

  • Create or edit help articles

  • Delete anything

This read-only design makes it safe to use with shared API tokens for analysis and reporting purposes.

API Rate Limits

Zendesk enforces API rate limits:

  • Professional plans: 700 requests per minute

  • Team plans: 400 requests per minute

The server doesn't implement rate limiting internally, so be mindful when making large queries.

Troubleshooting

"Error: Missing required environment variables"

Make sure all three environment variables are set in your MCP config:

  • ZENDESK_SUBDOMAIN

  • ZENDESK_EMAIL

  • ZENDESK_API_TOKEN

"Zendesk API error (401)"

  • Check that your email and API token are correct

  • Verify the API token hasn't been revoked in Zendesk

  • Ensure you're using the correct authentication format

"Zendesk API error (404)"

  • Verify your subdomain is correct

  • Check that the ticket/article ID exists

Server not appearing in Cursor

  • Make sure you've restarted Cursor completely after adding the config

  • Check that the path in mcp.json is correct (if using local installation)

  • Verify the JSON syntax in your config file is valid

Development

Building

npm run build

Watch Mode

npm run watch

Running Locally

npm run dev

Make sure to set the environment variables:

export ZENDESK_SUBDOMAIN=your-company
export ZENDESK_EMAIL=your-email@company.com
export ZENDESK_API_TOKEN=your-api-token
npm run dev

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

  1. Fork the repository

  2. Create your feature branch (git checkout -b feature/amazing-feature)

  3. Commit your changes (git commit -m 'Add some amazing feature')

  4. Push to the branch (git push origin feature/amazing-feature)

  5. Open a Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Security

⚠️ Never commit your API tokens to version control!

Your MCP configuration file should remain local to your machine and not be shared or committed to git.

Acknowledgments

Built with the Model Context Protocol SDK by Anthropic.

Support

If you encounter any issues or have questions:

Available Tools

6 tools
get_articleB

Get full content of a specific Help Center article by ID

Inputs: article_id (required).

get_ticketA

Get full details of a specific ticket by ID, including all comments and metadata

Inputs: ticket_id (required), include_comments.

get_tickets_by_tagB

Get all tickets with a specific tag

Inputs: tag (required), page, per_page.

search_articlesA

Search Zendesk Help Center articles (knowledge base).

Examples:

  • "collections feature" - Search for articles about collections

  • "route planning guide" - Find articles about route planning

  • "label:feature_name" - Search by label

Inputs: query (required), page, per_page.

search_feature_feedbackB

Search for all tickets and articles related to a specific feature. This is a convenience tool that searches for feature-related content including: - Bug reports - Feature requests - User feedback - Support questions

Example: "collections" will find all tickets and articles mentioning collections

Inputs: feature_name (required), days_back, include_solved.

search_ticketsB

Search Zendesk support tickets with advanced query syntax.

Examples:

  • "status:open tags:collections" - Find open tickets tagged with "collections"

  • "subject:bug priority:high" - Find high priority bug tickets

  • "tags:feature_request created>2024-01-01" - Find feature requests created after Jan 1, 2024

  • "collections route" - Full-text search for "collections route"

Query syntax supports:

  • status: (new, open, pending, hold, solved, closed)

  • priority: (low, normal, high, urgent)

  • tags: (tag names)

  • created: (date comparisons like >2024-01-01, <2024-12-31)

  • subject: (search in subject)

  • Full text search without prefix

Inputs: query (required), page, per_page.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.5/5.0

    Scored across 6 tools

    Disambiguation4/5

    Most tools have distinct purposes: search_tickets (query-based search), get_ticket (by ID), search_articles (KB search), get_article (by ID). However, search_feature_feedback and get_tickets_by_tag overlap with search_tickets, as they are essentially specialized searches. search_feature_feedback is broad and could be redundant, while get_tickets_by_tag could be achieved with search_tickets using a tag: query. Minor ambiguity exists but descriptions clarify usage.

    Naming Consistency5/5

    All tool names follow a consistent verb_noun pattern: search_tickets, get_ticket, search_articles, get_article, search_feature_feedback, get_tickets_by_tag. The pattern is predictable and aids in understanding tool functionality.

    Tool Count4/5

    With 6 tools, the count is well within the typical range for a focused MCP server. Each tool addresses a specific need, though some (like search_feature_feedback and get_tickets_by_tag) could be seen as convenience wrappers that overlap with search_tickets, slightly inflating the count.

    Completeness3/5

    The server covers search and retrieval for tickets and articles, but lacks write operations such as creating, updating, or deleting tickets and articles. For a Zendesk support tool, common actions like adding comments or updating ticket status are missing, limiting agent workflows.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      C
      maintenance
      Enables AI assistants to search tickets, manage tags, create tickets, inspect automations, and more in Zendesk.
      MIT
    • A
      license
      A
      quality
      D
      maintenance
      Connects AI assistants to your Zendesk account with 14 curated tools for managing tickets, users, organizations, and help center articles, running locally with no third-party services.
      14
      MIT
    • F
      license
      Not graded
      quality
      A
      maintenance
      Enables AI assistants to retrieve and manage support tickets and comments, analyze tickets and draft responses via built-in prompts, and search Help Center articles as a knowledge base, with per-operator OAuth so each user acts under their own Zendesk permissions.
      -