Nebula — AI Infrastructure Console

Workspace

Settings

Connection and integration preferences for the Nebula Core console.

Backend connection

Endpoints used by the API client. Driven by environment variables.

API base URL

NEXT_PUBLIC_API_BASE_URL

https://nebula.privatecore.de

API key (X-API-Key)

NEXT_PUBLIC_API_KEY — sent on every request.

3fdb••••6b28

Health

GET /api/health

Connecting

Realtime telemetry

Polling cadence for the API client.

Polling interval

Defined in lib/config.js (API_POLL_MS).

5 seconds

Ollama

Local LLM runtime preferences (saved in this browser).

Host

http(s)://host:port of the Ollama daemon.

Keep-alive

How long Ollama keeps idle models loaded in VRAM.

Auto-pull missing models

ComfyUI

Diffusion workflow runtime preferences.

Endpoint

ComfyUI HTTP server URL.

Default workflow

Absolute path to a saved workflow JSON.

GPU monitoring thresholds

Local thresholds, applied client-side once GPU telemetry is exposed.

Temperature warning

°C

Temperature critical

°C

Sustained utilization warning

% over a 60s window.

Docker connectivity

Container metrics are proxied through the Nebula Core API. The Docker socket is never exposed to the browser.

Source

GET /api/containers

docker-runtime via Nebula Core

Security note. Container actions (start / stop / restart) are not yet exposed. When they land, they will go through the same X-API-Key authenticated REST surface.

Debug & cache

Runtime helpers for troubleshooting.

Debug logging

Enables verbose console output from the API client.