atera-mcp
MCP server for the Atera RMM API. Implements a decision tree architecture for efficient tool discovery, enabling Claude to manage devices, alerts, tickets, and customers in Atera.
README
Atera MCP Server
Model Context Protocol (MCP) server for interacting with the Atera RMM API. Implements a decision tree architecture for efficient tool discovery and reduced context overhead.
One-Click Deployment
Features
- Decision Tree Navigation: Tools are organized by domain (customers, agents, tickets, alerts, contacts). Navigate to a domain first, then use domain-specific tools.
- Lazy Client Loading: The Atera client is only instantiated when first needed, reducing startup time.
- Full API Coverage: Supports customer management, device/agent monitoring, ticket operations, alert handling, and contact management.
- Rate Limit Handling: Built-in rate limiting via the node-atera client (700 req/min).
Installation
npm install @wyre-technology/atera-mcp
Or build from source:
git clone https://github.com/wyre-technology/atera-mcp.git
cd atera-mcp
npm install
npm run build
Configuration
Set the following environment variable:
| Variable | Required | Description |
|---|---|---|
ATERA_API_KEY |
Yes | Your Atera API key from Admin > API |
Getting Your API Key
- Log into Atera as an admin
- Go to Admin > API
- Generate or copy your API key
Usage
With Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"atera": {
"command": "npx",
"args": ["@wyre-technology/atera-mcp"],
"env": {
"ATERA_API_KEY": "your-api-key-here"
}
}
}
}
With MCP Gateway
Configure in the gateway registry:
{
"name": "atera-mcp",
"command": "node",
"args": ["/path/to/atera-mcp/dist/index.js"],
"env": {
"ATERA_API_KEY": "${ATERA_API_KEY}"
}
}
Docker
docker build -t atera-mcp .
docker run -e ATERA_API_KEY=your-key atera-mcp
Decision Tree Architecture
This server uses a navigation-based approach to tool discovery:
- Start: Only
atera_navigatetool is available - Navigate: Call
atera_navigatewith a domain (customers, agents, tickets, alerts, contacts) - Domain Tools: After navigation, domain-specific tools become available
- Back: Use
atera_backto return to domain selection
This architecture:
- Reduces tool list size for better LLM performance
- Groups related operations logically
- Minimizes context window usage
Available Domains
Customers
Manage customer (company) records.
atera_customers_list- List customers with paginationatera_customers_get- Get customer by IDatera_customers_create- Create new customer
Agents
Manage devices/endpoints with the Atera agent installed.
atera_agents_list- List agents with optional customer filteratera_agents_get- Get agent by IDatera_agents_get_by_machine- Get agent by machine name
Tickets
Manage service tickets.
atera_tickets_list- List tickets with filtersatera_tickets_get- Get ticket by IDatera_tickets_create- Create new ticketatera_tickets_update- Update existing ticket
Alerts
Monitor alerts from devices and agents.
atera_alerts_list- List alerts with filtersatera_alerts_get- Get alert by IDatera_alerts_by_agent- List alerts for an agentatera_alerts_by_device- List alerts for a device
Contacts
Manage customer contacts.
atera_contacts_list- List all contactsatera_contacts_get- Get contact by IDatera_contacts_by_customer- List contacts for a customer
Example Conversation
User: List all open tickets
Claude: I'll navigate to the tickets domain and list open tickets.
[Calls atera_navigate with domain: "tickets"]
[Calls atera_tickets_list with ticketStatus: "Open"]
Result: Found 15 open tickets...
Development
# Install dependencies
npm install
# Build
npm run build
# Run in development
npm run dev
# Type check
npm run typecheck
# Lint
npm run lint
# Test
npm run test
API Rate Limits
Atera API allows 700 requests per minute. The underlying node-atera client handles rate limiting automatically with request queuing.
License
Apache-2.0
Contributing
Contributions welcome! Please read our contributing guidelines and submit PRs to the main branch.
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。