First sign-in¶
The hub creates one administrator account at first start. Sign in with it, then change two things before you connect anything real.
The sign-in page. Each single sign-on provider you configure appears as a button under the password form; with none configured, there are none. The page shows nothing about your installation, because anyone who can reach it sees it before signing in.
The page comes in two styles — Aperture, the default, and Sentinel — chosen for the whole installation under Settings → Appearance.
Sign in¶
Open the hub's address using the ingress.host you installed with. The minimal install
uses http://argus.example.com; use https://argus.example.com if you enabled TLS through
a cert-manager ClusterIssuer or your own TLS Secret. Replace the example hostname with yours.
| Field | Value |
|---|---|
admin@localhost |
|
| Password | The bootstrap password from your installation |
The chart generates the bootstrap password and stores it in a Secret in the hub's namespace:
kubectl -n argus-system get secret argus-hub-bootstrap -o jsonpath='{.data.password}' | base64 -d; echo
Change these two things first¶
The bootstrap password. It exists to get you in once, and the hub asks you for a new one as soon as you sign in.
Your own account. The bootstrap account is a super-admin with access to every cluster. Create a named account for yourself with the narrowest role that covers your work, and use that day to day. See Users and roles.
What you will see¶
The cluster the hub runs in is already being watched: it appears on Clusters as local
within a minute of installing. A hub with no cluster connected yet — installed outside
argus-system, or with its own agent turned off — leads the overview with a four-step
onboarding card instead: create an enrolment token, install the agent, accept the cluster,
confirm the heartbeat. The card tracks the real state of the cluster you are enrolling and
disappears once one is connected.
Administrators also see Ready with a count in the top bar. It opens a short checklist: enrol your first cluster, choose an AI provider, add someone who can approve, route new incidents somewhere, and give your AI tools access. Each step is ticked off when ARGUS sees it done, not when it is clicked, and the checklist goes once every step is done or you hide it.