Freshdesk MCP Server
Provides a standard MCP interface for Freshdesk helpdesk operations, including ticket lifecycle management, draft/send replies and notes, contact lookup, and agent/group discovery.
README
Freshdesk MCP Server
A Model Context Protocol (MCP) server that exposes Freshdesk helpdesk operations as callable tools. It provides full ticket lifecycle management, two-phase draft/send for replies and notes, contact lookup, and agent/group discovery — all through a standard MCP interface compatible with any MCP client.
Features
- 14 tools covering tickets, replies, notes, contacts, groups, and agents
- Two-phase draft/send pattern for replies and notes — review before posting
- Custom fields support on ticket create and update
- Predefined filters and full query-syntax search for tickets
- Client-agnostic: works with any MCP-compatible host
Prerequisites
- Node.js 18 or later
- A Freshdesk account with API access
Getting Your API Key
- Log in to your Freshdesk account
- Click your avatar in the top-right corner and select Profile Settings
- Your API key is displayed in the right sidebar under Your API Key
Installation
git clone https://github.com/ArcaneSK/freshdesk-mcp-server.git
cd freshdesk-mcp-server
npm install
npm run build
Environment Setup
Copy the example environment file and fill in your credentials:
cp .env.example .env
Edit .env with your values:
FRESHDESK_API_KEY=your_api_key_here
FRESHDESK_DOMAIN=your_subdomain
| Variable | Description | Example |
|---|---|---|
FRESHDESK_API_KEY |
Your Freshdesk API key (see Getting Your API Key) | abcdef123456 |
FRESHDESK_DOMAIN |
Your Freshdesk subdomain — if your URL is https://acme.freshdesk.com, use acme |
acme |
Claude Desktop Configuration
Add the following to your Claude Desktop claude_desktop_config.json, replacing the path with your actual install location:
{
"mcpServers": {
"freshdesk": {
"command": "node",
"args": ["/path/to/freshdesk-mcp-server/dist/index.js"],
"env": {
"FRESHDESK_API_KEY": "your_api_key",
"FRESHDESK_DOMAIN": "your_subdomain"
}
}
}
}
Set FRESHDESK_DOMAIN to your Freshdesk subdomain only — for example, if your helpdesk URL is https://acme.freshdesk.com, use acme.
Tools Reference
| Tool | Description | Key Inputs |
|---|---|---|
list_tickets |
List tickets using predefined filters | filter, email, requester_id, page |
search_tickets |
Search tickets using Freshdesk query syntax | query (max 512 chars) |
get_ticket |
Retrieve a single ticket with all fields | ticket_id, include (conversations, requester, stats) |
create_ticket |
Create a new ticket | subject, description, email or requester_id |
update_ticket |
Update ticket fields | ticket_id, status, priority, responder_id, custom_fields |
delete_ticket |
Move a ticket to trash | ticket_id |
draft_reply |
Stage a reply for review (does not send) | ticket_id, body, cc_emails |
send_reply |
Send a previously drafted reply | draft_id |
draft_note |
Stage a note for review (does not post) | ticket_id, body, private |
send_note |
Post a previously drafted note | draft_id |
list_contacts |
List contacts by email or search term | email, search_term, page |
get_contact |
Get full contact details by ID | contact_id |
list_groups |
List all agent groups | page, per_page |
list_agents |
List agents, optionally filtered by group | group_id, page |
Two-Phase Draft/Send
Replies and notes use a two-step confirmation pattern to prevent accidental sends:
- Call
draft_replyordraft_notewith the content. The tool returns adraft_idand a preview of the content — nothing is sent to Freshdesk. - Review the draft, then call
send_replyorsend_notewith thedraft_idto post it.
Drafts are held in memory with a 10-minute expiration window. Each draft can only be sent once. If a draft expires, create a new one.
Development
# Clone and install
git clone https://github.com/ArcaneSK/freshdesk-mcp-server.git
cd freshdesk-mcp-server
npm install
# Build
npm run build
# Run tests
npm test
# Watch mode (TypeScript)
npm run dev
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。