System info

CPU, memory, disk, temperature, and uptime for the machine Athena itself runs on - the one integration in this app with nothing to configure at all.

No setup, always on

Unlike every other integration, there's no address, credential, or pairing step - System info is a single always-on poll loop (every 10 seconds) started once when Athena boots, reading via gopsutil rather than connecting to anything external.

What it reports

CPU - overall usage percentage.
Memory - usage percentage, used, and total (GB).
Disk - usage percentage per mounted filesystem.
Temperature - one reading per hardware sensor the host exposes.
Uptime - shown as days/hours rather than a raw decimal.
Athena version, Athena latest version, and Athena update available - this build's own current version, the latest published version, and whether a newer release exists, each a real entity in its own right (not just fields on a special-cased card) - usable on a Dashboard card or in an automation exactly like any hardware reading, e.g. "notify me when an Athena update is available".

Every reading works with Dashboard cards and Automations exactly like any other integration's entities - useful for a "warn me if this box is overheating or nearly out of disk" automation without needing a separate monitoring tool.

Restart and self-update

The "Athena" card on this page (and its Dashboard-card counterpart) has two buttons, each also a real button entity you can wire into an automation the same way:

Restart - always available, no setup needed. Just exits the process and relies on your docker-compose.yml's own restart: unless-stopped policy (or your orchestrator's equivalent) to bring a fresh container straight back up.
Update - only shown once you've opted in (see below). Shells out to the real docker compose CLI, exactly as if you'd typed docker compose pull && docker compose up -d by hand against your own real compose file - not an Athena-specific partial update, the whole stack.

Either action shows a full-screen overlay with a spinner that polls in the background and jumps back to the app automatically once Athena is reachable again - the same "stay on the restart screen, come back on its own" behavior a lot of self-hosted server dashboards use for their own reboots.

Update needs two things you enable yourself, since handing a container root-equivalent access to the host's Docker socket is a real security boundary crossing this app never does without asking: in docker-compose.yml, uncomment the two commented-out lines under the athena service's own volumes: ("Self-update" comment block) - the host's Docker socket, and the deploy directory itself bind-mounted at the exact same absolute path on both sides - then set ATHENA_DEPLOY_DIR in .env to that directory's real absolute host path. Leave it unset to keep Update hidden entirely (Restart still works either way). There is no automatic rollback if an update goes wrong - the overlay says so plainly if Athena doesn't come back online within a few minutes.