This MCP server provides a Zendesk integration for Model Context Protocol clients. It exposes tools to fetch and manage Zendesk tickets and comments, analyze tickets with specialized prompts, and access Zendesk Help Center articles as knowledge base content. Typical usage involves starting the server with uv in a designated directory, ensuring that Zendesk credentials are configured via environment variables or a .env file, and then connecting to the MCP server through Claude Code/Desktop or another MCP client. Core tools include get_tickets to list tickets with pagination, get_ticket to retrieve a specific ticket, get_ticket_comments to fetch a ticket’s comments, create_ticket_comment to add comments, create_ticket to open new tickets, and update_ticket to modify existing tickets. There are also prompts like analyze-ticket for deep analysis and draft-ticket-response to generate responses for tickets. These capabilities collectively let you manage Zendesk workflows directly from the MCP interface.

• Prerequisites:
  • Python 3.8+ and uvicorn/uv tooling as required by the project
  • Access to a Zendesk account with API credentials
  • Node.js is not required for this server (uses Python/uv by default in this setup)
  • Git to clone the repository
• Installation steps:
  1. Clone the repository: git clone https://github.com/reminia/zendesk-mcp-server.git cd zendesk-mcp-server
  2. Create and activate a Python virtual environment, then install dependencies: uv venv uv pip install -e .

     or if you have a build command in this project

     uv build
  3. Configure credentials:
     • Copy .env.example to .env and fill in your Zendesk credentials (subdomain, email, API token).
     • Ensure .env is excluded from version control.
  4. Run the server locally (example): uv run zendesk
  5. If using Docker, refer to the Docker section in the README for containerized setup.
• Notes:
  • The exact run command may vary depending on your environment and the setup in the README. Adjust paths as needed.

Environment variables are required to authenticate with Zendesk (subdomain, user email, and API token). If running via Docker, pass an env-file with these credentials. Some hosting environments may require network access to Zendesk APIs and proper permission scopes for API tokens. When wiring to Claude MCP, use the provided mcpServers entry structure and ensure the command path to uv and the working directory are correct. If you encounter permission errors in Docker, ensure the container has access to the .env file and that the container user has appropriate privileges. The Prompts section (analyze-ticket, draft-ticket-response) can help generate insights and draft replies based on ticket content. For production deployments, consider setting up proper secret management for credentials and restricting access to the MCP server endpoints.