Zendesk MCP Server
This server provides AI-assisted CRUD and search over Zendesk Support, Help Center, Talk, and Chat, with configurable read-only and lite/full tool modes.
Tickets: Create, read, update, delete, list (full feed only), fetch comments, add public/internal comments, view/download attachments, and AI-analyze ticket images and documents (Claude Vision).
Search: Powerful Zendesk query language across tickets, users, organizations, groups, and articles with filters (recipient, assignee, status, tags, dates, etc.) and pagination.
Users: Create, list, update, soft-delete users; fetch profiles by ID (find by email via search).
Organizations: Create, list, update, delete organizations; manage domain_names and tags.
Automation & Workflows: Manage groups, macros, views, triggers, and automations (CRUD + list/get, with conditions/actions schemas).
Help Center: List, get, create, update, delete articles (HTML bodies, drafts, locales, labels).
Support/Account Info: Get authenticated agent identity, subdomain, brands, ticket forms (
support_info).Talk & Chat: Get Talk voice stats (
get_talk_stats) and list chat transcripts (list_chats).Deployment modes: Stdio (API token) or HTTP (OAuth 2.1/PKCE) with per-session auth;
MODE=litelimits to 10 essential tools;READ_ONLY/READ_ONLY_STRICTblock write operations.
Provides access to Zendesk API for managing tickets, users, organizations, automation, help center articles, and search, with AI-powered ticket analysis.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Zendesk MCP Server list tickets assigned to me"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Zendesk MCP Server
A Model Context Protocol server for Zendesk API integration with AI-powered ticket analysis
Quick Start • Configuration • Tools • Architecture • Development
Overview
Zendesk MCP Server provides comprehensive access to the Zendesk API through the Model Context Protocol. It supports two transport modes that are auto-detected from environment variables:
Stdio mode — API token auth, for CLI-based MCP clients (Claude Code, Cursor, etc.)
HTTP mode — OAuth 2.1 with PKCE, for web-based MCP clients
Both modes expose the same set of tools. No code changes are needed to switch between them.
Based on mattcoatsworth/zendesk-mcp-server with significant enhancements including AI-powered features, dual-mode authentication, improved error handling, and comprehensive retry logic.
Related MCP server: Zendesk MCP Server by Fruggr
Quick Start
Stdio Mode (API Token)
Best for CLI-based MCP clients like Claude Code or Cursor. Each user only needs their own email + API token.
1. Get your API token from Zendesk Admin Center → Apps and integrations → APIs → Zendesk API → Add API token.
2. Configure your MCP client:
{
"zendesk": {
"type": "stdio",
"command": "npx",
"args": ["@sshadows/zendesk-mcp-server"],
"env": {
"ZENDESK_SUBDOMAIN": "your-subdomain",
"ZENDESK_EMAIL": "you@example.com",
"ZENDESK_API_TOKEN": "your-api-token"
}
}
}That's it. The server auto-detects stdio mode and connects.
HTTP Mode (OAuth 2.1)
Best for web-based MCP clients or multi-user deployments with centralized OAuth.
1. Create an OAuth app in Zendesk Admin Center → Apps and integrations → APIs → OAuth Clients.
2. Create a .env file:
ZENDESK_SUBDOMAIN=your-subdomain
ZENDESK_OAUTH_CLIENT_ID=your_client_id
ZENDESK_OAUTH_CLIENT_SECRET=your_client_secret
ZENDESK_OAUTH_REDIRECT_URI=http://localhost:3030/zendesk/oauth/callback3. Start and authorize:
npm start
# Visit http://localhost:3030/oauth/authorize in your browser4. Use the token with your MCP client: Authorization: Bearer mcp_...
Installation
Requirement | Version | Notes |
Node.js | >= 18.0.0 | Required |
Zendesk Account | Any plan | Required |
Anthropic API Key | - | Only for AI analysis features |
# npm (recommended)
npm install -g @sshadows/zendesk-mcp-server
# Or from source
git clone https://github.com/SShadowS/zendesk-mcp-server.git
cd zendesk-mcp-server
npm installConfiguration
Environment Variables
The server auto-detects which mode to use based on which variables are set.
Stdio mode (set ZENDESK_EMAIL + ZENDESK_API_TOKEN, without ZENDESK_OAUTH_CLIENT_ID):
ZENDESK_SUBDOMAIN=mycompany
ZENDESK_EMAIL=user@example.com
ZENDESK_API_TOKEN=your-api-tokenHTTP mode (set ZENDESK_OAUTH_CLIENT_ID):
ZENDESK_SUBDOMAIN=mycompany
ZENDESK_OAUTH_CLIENT_ID=your_client_id
ZENDESK_OAUTH_CLIENT_SECRET=your_client_secret
ZENDESK_OAUTH_REDIRECT_URI=http://localhost:3030/zendesk/oauth/callbackCommon (both modes):
MODE=full # 'full' (all 55 tools) or 'lite' (10 essential tools)
READ_ONLY=false # 'true' blocks every write except internal ticket notes
READ_ONLY_STRICT=false # 'true' also blocks internal notes, leaving no write path
ANTHROPIC_API_KEY=sk-ant-... # Required for AI image/document analysis
ZENDESK_DEBUG=false # Enable debug loggingHTTP mode only:
PORT=3030
SERVER_BASE_URL=http://localhost:3030 # Use https:// in productionSee .env.example for the full list.
Tool Modes
Control which tools are exposed with the MODE environment variable:
full(default) — All 55 tools available.lite— 10 essential tools for reduced context usage:search,get_user,list_tickets,get_ticket,get_ticket_comments,add_ticket_comment,get_ticket_attachments,analyze_ticket_images,analyze_ticket_documents,get_document_summary.
MODE=lite npm startRead-Only Mode
Set READ_ONLY=true to prevent the server from modifying Zendesk data:
READ_ONLY=true npm startEvery
create_*,update_*anddelete_*tool is unregistered — 28 tools remain infullmode.add_ticket_commentstays available for internal notes only. Atype: 'public'call is refused rather than downgraded, so a caller is never left believing it replied to a customer who received nothing.Enforcement also sits in the HTTP client, which rejects any non-
GETrequest other than an internal-note payload. A write cannot reach the Zendesk API by another route.Composes with
MODE.MODE=lite READ_ONLY=truestill exposes all 10 lite tools, since onlyadd_ticket_commentmutates and it survives.
Recognized truthy values are true and 1 (case-insensitive); anything else, including unset, leaves the server writable. An unrecognized value is treated as false and logs a warning at startup, so a typo cannot silently reopen a write path.
Strict read-only
READ_ONLY still leaves one write path open: internal ticket notes. Where even that is unacceptable — for example a sidecar whose spec says only an orchestrator may write — add READ_ONLY_STRICT=true:
READ_ONLY=true READ_ONLY_STRICT=true npm startadd_ticket_commentis unregistered, leaving 27 tools infullmode and 9 underMODE=lite.The transport guard drops the internal-note exemption, so
request()permitsGETand nothing else. Both layers tighten together; unregistering the tool alone would be cosmetic, since anything holding the client object could still write.Strict is defined as the allowlist minus anything mutating, so it stays correct as tools are added rather than depending on a list someone has to remember to update.
Strict is a separate boolean rather than a third value of READ_ONLY on purpose. A single READ_ONLY=true is often shared across several MCP servers, and some treat an unrecognized value as false with only a warning — so READ_ONLY=strict would quietly make those servers writable.
READ_ONLY_STRICT=true implies read-only even if READ_ONLY is unset or false. A server told to be strict must never come up writable.
The active mode is logged at startup (mode=OFF, mode=STANDARD, or mode=STRICT).
Available Tools
Tool | Description |
| List tickets with filters (status, assignee, etc.) |
| Get ticket details with optional comments |
| Create a new ticket |
| Update ticket fields |
| Get all comments on a ticket |
| Add public or internal comment |
| Get ticket attachments |
| AI-powered image analysis with Claude Vision |
| AI-powered document analysis |
| Quick document summary |
Tool | Description |
| List all users |
| Get user details |
| Create new user |
| Update user info |
| Delete a user |
Tool | Description |
| List all organizations |
| Get organization details |
| Create new organization |
| Update organization |
| Delete organization |
Category | Tools |
Groups |
|
Macros |
|
Views |
|
Triggers |
|
Automations |
|
Category | Tools |
Search |
|
Help Center |
|
Talk |
|
Chat |
|
Architecture
Transport Modes
src/index.js (auto-detection)
├── ZENDESK_EMAIL + ZENDESK_API_TOKEN → Stdio mode
│ ├── ZendeskClient.setApiTokenAuth()
│ ├── setDefaultZendeskClient(client)
│ └── StdioServerTransport (stdin/stdout)
│
└── ZENDESK_OAUTH_CLIENT_ID → HTTP mode
├── Express server (src/http-server.js)
├── OAuth 2.1 with PKCE (src/auth/)
├── Per-session ZendeskClient instances
└── StreamableHTTPServerTransportTools are identical in both modes. They call getZendeskClient() which resolves to:
HTTP mode: Per-session client via AsyncLocalStorage
Stdio mode: Singleton default client
Project Structure
zendesk-mcp-server/
├── src/
│ ├── index.js # Entry point (auto-detects mode)
│ ├── http-server.js # Express server with OAuth (HTTP mode only)
│ ├── server.js # MCP server setup and tool registration
│ ├── request-context.js # Per-session + default client context
│ ├── auth/
│ │ ├── oauth-handler.js # OAuth 2.1 with PKCE
│ │ ├── session-store.js # Session management
│ │ └── middleware.js # Bearer token auth middleware
│ ├── zendesk-client/
│ │ ├── base.js # Auth, HTTP requests, retry logic
│ │ ├── index.js # Mixin composition
│ │ ├── tickets.js # Ticket API methods
│ │ ├── users.js # User API methods
│ │ └── ... # Other API domain mixins
│ ├── tools/ # MCP tool implementations
│ ├── config/
│ │ └── tool-modes.js # Full/lite mode filtering
│ └── utils/
│ ├── errors.js # Classified error types
│ ├── retry.js # Exponential backoff
│ ├── ticket-context.js # AI prompt context builder
│ ├── document-handler.js # Document routing
│ └── converter-client.js # Office-to-PDF conversion
├── tests/ # Vitest test suite
├── .env.example # Environment variable template
└── CLAUDE.md # AI assistant project guideKey Design Decisions
Dual auth in one client:
ZendeskClientBasesupports bothsetApiTokenAuth()(Basic) andsetAccessToken()(Bearer). The_authModefield determines which headergetAuthHeader()returns.Default client fallback: AsyncLocalStorage doesn't propagate through StdioServerTransport's event callbacks. Instead of fighting that,
getZendeskClient()falls back to a module-level default client in stdio mode. Zero changes needed in any tool file.Console.error everywhere: In stdio mode, stdout is the MCP transport. All diagnostic logging in shared code paths uses
console.error.HTTP mode is unchanged:
src/http-server.jsandsrc/auth/*are only imported in HTTP mode. No changes were needed.
Development
npm start # Start server (auto-detects mode)
npm run dev # Start with auto-reload
npm test # Run all tests
npm run test:watch # Run tests in watch mode
npm run inspect # Launch MCP InspectorTesting
Tests use Vitest and are in tests/ mirroring the src/ directory:
npm test # Run all tests
npm run test:watch # Watch modeIntegration tests (against real Zendesk + Anthropic APIs) require .env credentials and are automatically skipped when credentials are missing.
HTTP Mode Endpoints
Endpoint | Description |
| Main MCP endpoint (requires Bearer token) |
| Start OAuth flow |
| OAuth callback |
| Token exchange |
| Dynamic client registration (RFC 7591) |
| OAuth metadata (RFC 8414) |
| Protected resource metadata (RFC 9728) |
| Health check |
Troubleshooting
The server needs either API token or OAuth credentials. Set one of:
# Stdio mode
ZENDESK_SUBDOMAIN=... ZENDESK_EMAIL=... ZENDESK_API_TOKEN=...
# HTTP mode
ZENDESK_SUBDOMAIN=... ZENDESK_OAUTH_CLIENT_ID=... ZENDESK_OAUTH_CLIENT_SECRET=...Complete OAuth flow: visit
http://localhost:3030/oauth/authorizeCheck if token expired (24-hour TTL) — re-authorize if needed
Ensure Bearer token is included:
Authorization: Bearer mcp_xxxToken format should start with
mcp_
Verify
ZENDESK_EMAILis correctVerify
ZENDESK_API_TOKENis a valid API token (not a password)Verify
ZENDESK_SUBDOMAINis correctCheck the connection test output in stderr on startup
In-memory sessions are cleared on restart. Re-authorize to get a new token. For production, implement a Redis-based session store (see src/auth/session-store.js).
Set ANTHROPIC_API_KEY in your environment. This is only needed for analyze_ticket_images, analyze_ticket_documents, and get_document_summary.
The server includes exponential backoff retry logic. If you hit rate limits frequently, consider using MODE=lite to reduce API calls, or check if multiple clients share the same credentials.
Contributing
Fork the repository
Create your feature branch (
git checkout -b feature/my-feature)Run tests (
npm test)Commit your changes
Open a Pull Request
License
MIT License — see LICENSE for details.
Acknowledgments
Original implementation by @mattcoatsworth
Built with Model Context Protocol
AI features powered by Anthropic Claude
Made with care by SShadowS
Available Tools
55 toolsadd_ticket_commentA
Append a comment to an existing ticket. Default visibility is internal (agent-only note) — pass type:'public' to send a reply visible to the requester. Use this rather than update_ticket when you only want to add a comment without changing other ticket fields.
Inputs: id (required), body (required), type, author_id.
analyze_ticket_documentsA
Comprehensively analyze all document attachments from a ticket (PDF, DOCX, TXT, CSV, etc.) using AI. Long documents are truncated to fit the analysis budget. Note: this may take 30-60 seconds for multiple documents.
Inputs: id (required), max_tokens, quick_mode, document_types, +2 more.
analyze_ticket_imagesA
Download and analyze images from a ticket using AI vision with comprehensive analysis. Includes both file attachments and inline images embedded in comment bodies. Optionally scope to a single comment/post via comment_id, or to specific images via attachment_ids (discover ids with get_ticket_attachments).
Inputs: id (required), comment_id, max_tokens, attachment_ids, +2 more.
create_articleB
Create a Help Center article in a specific section. Body accepts HTML. Specify locale if you support multiple languages (defaults to instance default).
Inputs: body (required), title (required), section_id (required), draft, +4 more.
create_automationA
Create a new automation (runs hourly on tickets matching the conditions). Requires title, conditions, and actions. Conditions should include a time-based field (e.g. hours_since_update) — otherwise consider a trigger instead.
Inputs: title (required), actions (required), conditions (required), description.
create_groupA
Create a new agent group. Group members are added separately via Zendesk's group_memberships endpoint (not exposed by this tool).
Inputs: name (required), description.
create_macroB
Create a new macro. Provide title and an actions array. Each action is {field, value} describing what the macro changes when applied.
Inputs: title (required), actions (required), description.
create_organizationA
Create a new organization. Name must be unique within the instance. Use domain_names (array) to auto-associate end-users whose email domain matches.
Inputs: name (required), tags, notes, details, +1 more.
create_ticketC
Create a new ticket. Supports named_custom_fields (e.g. ado_work_item_id) and raw custom_fields.
Inputs: comment (required), subject (required), tags, type, +7 more.
create_triggerB
Create a new trigger. Requires title, conditions (when it fires), and actions (what it does). Triggers run on every ticket create/update — be conservative with conditions to avoid performance impact.
Inputs: title (required), actions (required), conditions (required), description.
create_userA
Create a new Zendesk user. Email must be unique across the instance — duplicates return 422. Role defaults to end-user; specify agent or admin for staff accounts. Returns the created user's full profile including the auto-assigned ID.
Inputs: name (required), email (required), role, tags, +3 more.
create_viewA
Create a new ticket view. Pass conditions as an object with optional all and any arrays (Zendesk evaluates all as AND-logic, any as OR-logic). Each condition is {field, operator, value} per Zendesk's view conditions schema. Use output to control which columns appear, plus grouping and sort.
Inputs: title (required), conditions (required), output, description.
delete_articleA
Delete a Help Center article. Soft-deletable — Zendesk retains the record briefly for undo. Existing links to the article will 404.
Inputs: id (required).
delete_automationA
Delete an automation. Tickets previously modified by it are unaffected.
Inputs: id (required).
delete_groupA
Delete an agent group. Tickets currently assigned to this group will be unassigned (group_id set to null).
Inputs: id (required).
delete_macroA
Delete a macro. Tickets previously modified by it are unaffected.
Inputs: id (required).
delete_organizationA
Delete an organization. Associated users are not deleted but are unlinked from the org.
Inputs: id (required).
delete_ticketB
Delete a ticket
Inputs: id (required).
delete_triggerA
Delete a trigger. Tickets previously modified by it are unaffected.
Inputs: id (required).
delete_userA
Soft-delete a Zendesk user (sets active:false; the record is retained for ticket history). To permanently purge, use Zendesk's GDPR delete endpoint (not exposed by this tool). Cannot delete the account owner.
Inputs: id (required).
delete_viewA
Delete a saved view. Tickets are unaffected.
Inputs: id (required).
get_articleA
Fetch one Help Center article by numeric ID, returning title, body (HTML), section_id, labels, and locale.
Inputs: id (required).
get_automationA
Fetch one automation's definition by numeric ID, including its conditions and actions array.
Inputs: id (required).
get_document_summaryB
Get a quick summary of all documents attached to a ticket
Inputs: id (required).
get_groupA
Fetch one agent group by numeric ID, returning name, description, and metadata.
Inputs: id (required).
get_macroA
Fetch one macro's definition by numeric ID, including the title and the array of actions it performs (status change, comment, tag adds, etc.).
Inputs: id (required).
get_organizationA
Fetch one organization by numeric ID, returning name, domain_names, tags, custom_fields, and notes. If you only have a name or domain, use search first to find the ID.
Inputs: id (required).
get_talk_statsA
Return Zendesk Talk (voice) aggregate stats: total calls, average wait, abandoned rate, agent availability. Requires Zendesk Talk add-on. Returns the rolling window Zendesk publishes (typically last 30 days).
Inputs: no named parameters.
get_ticketA
Fetch one ticket by numeric ID, returning the full ticket object plus named_custom_fields (e.g. ado_work_item_id). Pass include_comments:true to also pull the comment thread inline (otherwise comments are omitted to save tokens — fetch them separately with get_ticket_comments if needed). If you have an email/subject/tag rather than an ID, use search first to find the ID.
Inputs: id (required), include_comments.
get_ticket_attachmentsA
List every attachment across a ticket's comment thread (files and inline images). Use this to discover what's attached before deciding to call analyze_ticket_images or analyze_ticket_documents. Each attachment includes content_type, size, filename, and a content_url for downloading.
Inputs: id (required).
get_ticket_commentsA
List the comment thread for a ticket (both public replies and internal agent notes). Useful when you need conversation history but already used get_ticket without include_comments:true. Comments are paginated — large tickets may have 50+ comments across multiple pages.
Inputs: id (required), page, per_page, sort_order.
get_triggerA
Fetch one trigger's definition by numeric ID, returning its conditions (all_conditions/any_conditions) and actions array.
Inputs: id (required).
get_userA
Fetch one user by numeric ID, returning the full profile (name, email, role, organization, tags, custom_fields). If you only have an email or name, use search with type:user email:<email> first to find the ID.
Inputs: id (required).
get_viewA
Fetch one view's definition by numeric ID, returning the filter conditions, title, and metadata. Useful for understanding how an agent's existing dashboard is built.
Inputs: id (required).
list_articlesA
List Help Center (knowledge-base) articles. To find articles by title/body/section, use search with type:article title:<text> instead — list_articles has no filtering parameters and the catalog can be large.
Inputs: page, sort_by, per_page, sort_order.
list_automationsA
List ALL automations (time-based rules that run hourly — e.g. closing pending tickets after 7 days). Unlike triggers, automations fire on a schedule, not on events. Use this to audit existing time-based workflows.
Inputs: page, per_page.
list_chatsA
List Zendesk Chat conversations (live chat transcripts), most recent first. Requires Zendesk Chat to be enabled on the account. Use this for end-of-day chat reviews or to find a recent chat by visitor name.
Inputs: page, per_page.
list_groupsA
List ALL agent groups (teams that own tickets) in the Zendesk instance. Each group has an ID used in ticket routing — use this when you need to discover group IDs for update_ticket group_id:<id> or search type:ticket group:<id>.
Inputs: page, per_page.
list_macrosA
List ALL macros (predefined ticket actions agents can apply) accessible to the API user. Use this to discover macro IDs and inspect what canned responses or ticket updates your team has built.
Inputs: page, per_page.
list_organizationsA
List ALL organizations in the Zendesk instance. No filter parameters — to find an organization by name, domain, external_id, or tag use search with type:organization name:<name> or type:organization tags:<tag>. Supports pagination.
Inputs: page, per_page.
list_ticketsA
List ALL tickets in the Zendesk instance (no filtering). Has no parameters for filtering by recipient, assignee, requester, status, tags, or dates — for any filtered query use search instead with operators like type:ticket recipient:<email> or type:ticket assignee:me status<solved. Use list_tickets only when you genuinely want the full chronological feed across all queues (e.g. for a global activity report). Supports pagination and sorting.
Inputs: page, sort_by, per_page, sort_order.
list_triggersA
List ALL triggers (event-driven rules that fire on ticket create/update). Use this to audit existing triggers before adding new ones, or to discover trigger IDs for inspection. For time-based rules use list_automations instead.
Inputs: page, per_page.
list_usersA
List Zendesk users with optional role filter (end-user, agent, admin). For finding a user by email/name/organization/external_id, use search with type:user email:<email> — that's far cheaper than paginating the full directory. list_users is appropriate when you need the full directory feed (e.g. building a snapshot).
Inputs: page, role, per_page.
list_viewsA
List ALL saved ticket views (filtered ticket lists agents use as dashboards) accessible to the API user. Use this to discover view IDs, then call Zendesk's /api/v2/views/{id}/tickets for the actual ticket list (this MCP doesn't yet expose execute_view; use search with the equivalent filters as a workaround).
Inputs: page, per_page.
searchA
Search Zendesk tickets, users, organizations, groups, and Help Center articles using Zendesk's full search query language. The query parameter supports operators (combined with spaces, implicit AND): type:ticket|user|organization|group|article, recipient:<email>, assignee:<email|me|none>, requester:<email>, submitter:<email>, status<solved (also >, <=, >=), tags:<tag>, brand:<id>, group:<id>, created>2026-05-01 (also updated, solved_at, due_at), priority:high|normal|low|urgent, via:web|email|chat|phone, plus full-text on subject/description. Combine for precision (e.g. type:ticket recipient:support@acme.com status<solved created>2026-05-01). IMPORTANT: default sort is relevance, not date — for chronological results pass sort_by:created_at (or updated_at) with sort_order:desc. Use this instead of list_tickets whenever you need filtering by recipient/assignee/tag/status/dates or full-text search; list_tickets has no filter parameters. Reference: https://support.zendesk.com/hc/en-us/articles/4408886879258
Inputs: query (required), page, sort_by, per_page, +1 more.
support_infoA
Return the authenticated agent's identity (id, name, email, role) plus account-level Support metadata: subdomain, account settings, brands, and ticket forms. Use this to confirm which Zendesk instance the session is connected to and which ticket forms / brands are available before constructing tickets.
Inputs: no named parameters.
update_articleA
Update an existing Help Center article's title, body, labels, or section. Pass only the fields you want to change.
Inputs: id (required), body, draft, title, +4 more.
update_automationA
Update an existing automation. Pass conditions and/or actions to replace those sections (not merge).
Inputs: id (required), title, actions, conditions, +1 more.
update_groupC
Update an existing group's name or description.
Inputs: id (required), name, description.
update_macroA
Update an existing macro's title or actions. Pass actions to replace the full action set (not merge).
Inputs: id (required), title, actions, description.
update_organizationA
Update an existing organization. Pass only the fields you want to change. To clear domain_names or tags, pass an empty array.
Inputs: id (required), name, tags, notes, +2 more.
update_ticketA
Update an existing ticket. Supports named_custom_fields (e.g. ado_work_item_id, pass null to clear) and raw custom_fields. Pass macro_id to apply a macro's field changes and comment as part of the update — explicit fields you also pass will override the macro's values.
Inputs: id (required), tags, type, status, +8 more.
update_triggerA
Update an existing trigger. Pass conditions and/or actions to replace those sections (not merge).
Inputs: id (required), title, actions, conditions, +1 more.
update_userA
Update an existing user. Pass only the fields you want to change; omitted fields are preserved. Use get_user first to confirm the ID and current state. Pass organization_id: null to detach the user from their current organization. String fields can be cleared by passing an empty string.
Inputs: id (required), name, role, tags, +4 more.
update_viewA
Update an existing view's title, description, conditions, and/or output (columns, grouping, sort). Use get_view first to retrieve and modify the current structure. Only fields you pass are changed — omitted fields are preserved.
Inputs: id (required), title, output, conditions, +1 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 55 tools
Tools are organized by resource (groups, organizations, tickets, users, macros, views, automations, triggers, articles) and each has a clear responsibility. The only mild overlaps are list_tickets vs search and the three document-analysis tools (analyze_ticket_documents, get_document_summary, analyze_ticket_images), but descriptions clarify when to use which.
The vast majority follow a consistent verb_noun pattern (create_organization, update_ticket, delete_macro, list_views). A few exceptions like add_ticket_comment, support_info, get_talk_stats, list_chats, and get_document_summary deviate slightly but are still intuitive and readable.
55 tools is far above the 25-tool threshold and spans Support, Help Center, Talk, Chat, and AI-analysis features. While each tool has a purpose, the surface is large enough that agents will spend significant effort just discovering and selecting the right tool.
Core CRUD is covered for tickets, users, organizations, groups, macros, views, automations, triggers, and articles, plus search and AI analysis. Gaps exist: group memberships are explicitly not exposed, article sections cannot be listed/created, execute_view is missing, and there are no tools for satisfaction ratings or ticket metrics.
Maintenance
Related MCP Connectors
Zendesk MCP server for ChatGPT, Claude and any AI agent: list tickets, reply to ticket, update ticket, apply macro, add internal note and search tickets in your own Zendesk account. Cotlet works in the account through your Cloud Browser, a real Chrome where you sign in to Zendesk yourself; Cotlet does not store your password and you need no Zendesk API key. About the official Zendesk API: Zendesk has a full REST API, the Support API (also called the Ticketing API), but no documented Zendesk-hosted MCP server for working tickets (checked 16 Oct 2026). Zendesk's MCP documentation covers the MCP client, where Zendesk calls other servers. Setup and FAQ: https://cotlet.dev/integrations/zendesk-api-mcp-server
Connect Claude, ChatGPT or any AI assistant to Zendesk tickets, CSAT, macros and Help Center.
Helpdesk tickets from your AI: find, reply, take, hold, complete — within your own role. OAuth.
Zendesk configuration management for AI agents: propose, review, apply and roll back changes.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceEnables interaction with Zendesk Support for managing tickets, comments, and filtering via the Model Context Protocol. It allows LLMs to perform actions like creating, updating, and listing support tickets through natural language in a Heroku-native environment.-
- AlicenseAqualityAmaintenanceA Model Context Protocol (MCP) server that connects LLMs to the Zendesk Support & Help Center APIs — with per-user OAuth 2.1 PKCE authentication and fine-grained tool visibility controls.631,561 npm15MIT
- FlicenseNot gradedqualityAmaintenanceEnables LLM clients to retrieve, create, and update Zendesk tickets and comments, and to search Help Center articles as a knowledge base, with per-operator OAuth so each caller acts under their own Zendesk permissions. It also ships prompts for analyzing tickets and drafting responses.-
- FlicenseNot gradedqualityAmaintenanceEnables users to retrieve and manage support tickets and comments, analyze tickets, draft responses, and access help center articles through a Model Context Protocol integration with OAuth-backed per-user permissions.-