feat: add Gremlin API health to server cron check

[nickmeinhold]

Context

The hourly server health check (/opt/scripts/health-check.sh) monitors disk, memory, swap, and container status — but doesn't check whether Gremlin can actually process messages. During the Mar 17 outage, the cron fired every hour with unrelated alerts while Gremlin was completely unable to respond.

Proposal

Add a check that curls the Gremlin health endpoint and alerts if the API component is degraded/down:

# Check Gremlin health endpoint
gremlin_health=$(curl -sf http://localhost:8080/health 2>/dev/null)
gremlin_status=$(echo "$gremlin_health" | jq -r '.status // "unreachable"')
if [ "$gremlin_status" != "healthy" ]; then
    issues+=("Gremlin: ${gremlin_status}")
fi

Depends on the /health endpoint having an API liveness check (#TBD).

In scope

• Add Gremlin health endpoint check to health-check.sh
• Filter out expected one-shot containers (outline_minio_setup) from container status check to reduce noise