N-central MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| MCP_PORT | No | Port number for HTTP mode (e.g., '3100'). Omit to use stdio. | |
| MCP_API_KEY | No | Bearer token for HTTP mode. Required unless MCP_ALLOW_UNAUTHENTICATED=1. | |
| NC_JWT_TOKEN | No | User-API JWT from the N-central UI. Required in single-tenant mode. | |
| NC_SERVER_URL | No | Your N-central URL, e.g. https://ncentral.example.com. Required in single-tenant mode. | |
| NC_WRITE_MODE | No | One of 'read-only', 'write', or 'full'. | write |
| MCP_CORS_ORIGIN | No | Comma-separated allowed origins for CORS. | |
| NC_MULTI_TENANT | No | Set to '1' to enable multi-tenant mode (requires HTTP mode). | |
| MCP_BIND_ADDRESS | No | Interface to bind in HTTP mode. | 127.0.0.1 |
| NC_FQDN_ALLOWLIST | No | Comma-separated host suffixes for SSRF guard in multi-tenant mode. | |
| NC_RESOURCE_CACHE_TTL_MS | No | Cache TTL in milliseconds for resources. Default 60000. | |
| MCP_ALLOW_UNAUTHENTICATED | No | Set to '1' to allow HTTP mode without API key (not recommended). |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| prompts | {
"listChanged": true
} |
| resources | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| get_server_statusA | Check N-central service health and API version, optionally including the server clock. |
| validate_sessionA | Confirm that the current authenticated N-central API session is valid without returning token material. |
| get_current_userA | Return compact current-user identity and authorization context, or explicit full profile detail. |
| search_organizationsA | Find service organizations, customers, sites, or organization units by human name or bounded filtering and pagination. Customer-scoped site listing is PREVIEW and may change. |
| get_organization_contextA | Load compact organization details plus selected bounded children, limits, and custom-property context. Request full/all explicitly. |
| search_devicesA | Search the global or organization-scoped N-central device inventory by human name or bounded filters and pagination. |
| get_device_contextA | Load a device plus selected monitoring, asset, lifecycle, note, custom-property, and task context. |
| list_active_issuesA | List current active monitoring issues for a customer or site (service-org scope is unsupported), with bounded pagination and a compact default projection. Use detailLevel=full only when complete upstream records are required. |
| list_device_scheduled_tasksA | List scheduled tasks associated with one N-central device. |
| get_scheduled_task_contextA | Load a scheduled-task definition and optionally its aggregate or per-device execution status. |
| run_reportB | Run one reviewed read-only report adapter; arbitrary REST methods or paths are not accepted. |
| list_job_statusesA | List asynchronous N-central job statuses with local filtering, bounded pagination, and a compact default projection. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| full-customer-report | Comprehensive customer/site report with org custom properties. Optional customerId to scope to a single customer. |
| device-health-audit | Audit device health — active issues, monitoring status. Optional orgUnitId to scope. |
| agent-deployment-status | Check agent deployment coverage — find sites with missing or low device counts |
| custom-property-audit | Audit custom property values across all customers for consistency |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| org-tree | Bounded org hierarchy: Service Orgs → Customers → Sites with IDs and names. |
| status | Server health and version info. |
TDQS
Scored across 12 tools
Each tool targets a distinct resource and action: server status, session validation, user profile, organization search/context, device search/context, scheduled tasks, active issues, reports, and jobs. The closest pair is list_device_scheduled_tasks versus get_scheduled_task_context, but one is a device-scoped listing and the other is a definition/status context, so they remain clearly separable.
All tool names follow a consistent lower_snake_case verb_noun pattern, mixing list_, get_, search_, validate_, and run_. There is no camelCase or inconsistent verb style, making the naming predictable and easy to navigate.
Twelve tools is a well-scoped size for an N-central monitoring and reporting server. Each tool covers a distinct capability, and none feels redundant or like filler.
The tool set covers the main read-only workflows: authentication, server health, organizations, devices, scheduled tasks, active issues, reports, and async job status. A broader management server might also expect create/update/delete actions, but given the clearly read-only report stance, the only real gap is the absence of any mutation/lifecycle operations.