feat: grouped server cards with progress bars and sparklines

- Agent sends group name (defaults to hostname) and richer metric notes
  (memory/disk include absolute GB values; CPU note includes system uptime)
- push() accepts group param; server stores it as agent_group on deployments
- Upsert key is now (company_id, agent_group, name) so two agents on the
  same company but different servers don't overwrite each other's CPU/Memory
- Add widget config packs: linux-base, web-nginx, database-postgres/redis,
  docker-host, custom-cmd examples

ethica-agent.example.yml

@@ -4,9 +4,13 @@
 # Get it from: admin panel → Company → API tab → Company token
 #
 # token: REPLACE_ME          ← or set ETHICA_TOKEN in .env / environment
-# home: https://ethica.no   ← or set ETHICA_HOME  (default: https://ethica.no)
+# home: https://ethica.no   ← or set ETHICA_HOME   (default: https://ethica.no)
 # interval: 60               ← or set ETHICA_INTERVAL (seconds, default: 60)
+# group: my-server           ← or set ETHICA_GROUP  (default: system hostname)
+#                              All metrics from this agent are grouped under this
+#                              name in the Ethica dashboard (shows as a server card).
 
+group: my-server
 interval: 60
 
 widgets:

ethica-agent.py

@@ -211,19 +211,48 @@ def run_widget(w: dict) -> tuple:
 
     if wtype == "system_cpu":
         pct = cpu_percent()
+        uptime_note = ""
+        if SYSTEM == "Linux":
+            try:
+                with open("/proc/uptime") as f:
+                    secs = float(f.readline().split()[0])
+                days = int(secs // 86400)
+                hours = int((secs % 86400) // 3600)
+                uptime_note = f" up {days}d{hours}h"
+            except Exception:
+                pass
         status = _pct_status(pct, w.get("degrade_threshold", 80), w.get("offline_threshold", 95))
-        return name, kind, status, None, "", f"cpu {pct}%"
+        return name, kind, status, None, "", f"cpu {pct}%{uptime_note}"
 
     if wtype == "system_memory":
         pct = mem_percent()
+        abs_note = ""
+        if SYSTEM == "Linux":
+            try:
+                info: dict = {}
+                with open("/proc/meminfo") as f:
+                    for line in f:
+                        k, _, v = line.partition(":")
+                        info[k.strip()] = int(v.split()[0])
+                total_kb = info.get("MemTotal", 0)
+                avail_kb = info.get("MemAvailable", 0)
+                used_gb = round((total_kb - avail_kb) / 1048576, 0)
+                total_gb = round(total_kb / 1048576, 0)
+                abs_note = f" {int(used_gb)}/{int(total_gb)}GB"
+            except Exception:
+                pass
         status = _pct_status(pct, w.get("degrade_threshold", 85), w.get("offline_threshold", 97))
-        return name, kind, status, None, "", f"mem {pct}%"
+        return name, kind, status, None, "", f"mem {pct}%{abs_note}"
 
     if wtype == "system_disk":
         path = w.get("path", "/")
-        pct = disk_percent(path)
+        u = shutil.disk_usage(path)
+        pct = round(100.0 * u.used / u.total, 1)
+        total_gb = round(u.total / 1e9, 0)
+        used_gb = round(u.used / 1e9, 0)
+        abs_note = f" {int(used_gb)}/{int(total_gb)}GB"
         status = _pct_status(pct, w.get("degrade_threshold", 80), w.get("offline_threshold", 95))
-        return name, "system", status, None, "", f"disk {pct}% ({path})"
+        return name, "system", status, None, "", f"disk {pct}%{abs_note} ({path})"
 
     if wtype == "http_check":
         url = w["url"]
@@ -276,12 +305,14 @@ def run_widget(w: dict) -> tuple:
 # ── Push ───────────────────────────────────────────────────────────────────────
 
 def push(home: str, token: str, name: str, kind: str, status: str,
-         latency_ms, version: str, note: str) -> int:
+         latency_ms, version: str, note: str, group: str = "") -> int:
     url = home.rstrip("/") + "/api/company/metrics"
     payload: dict = {"name": name, "kind": kind, "status": status,
                      "version": version, "note": note}
     if latency_ms is not None:
         payload["latency_ms"] = latency_ms
+    if group:
+        payload["group"] = group
     data = json.dumps(payload).encode()
     req = urllib.request.Request(
         url, data=data, method="POST",
@@ -313,10 +344,12 @@ def main():
     # ETHICA_TOKEN  : company API token  (required)
     # ETHICA_HOME   : home URL           (optional, default https://ethica.no)
     # ETHICA_INTERVAL: collection interval in seconds (optional)
-    token    = os.environ.get("ETHICA_TOKEN")    or cfg.get("token", "")
+    # ETHICA_GROUP  : server/group label shown in the dashboard card (default: hostname)
-    home     = os.environ.get("ETHICA_HOME")     or cfg.get("home", "https://ethica.no")
+    token       = os.environ.get("ETHICA_TOKEN")    or cfg.get("token", "")
-    interval = int(os.environ.get("ETHICA_INTERVAL") or cfg.get("interval", 60))
+    home        = os.environ.get("ETHICA_HOME")     or cfg.get("home", "https://ethica.no")
-    widgets  = cfg.get("widgets", [])
+    interval    = int(os.environ.get("ETHICA_INTERVAL") or cfg.get("interval", 60))
+    agent_group = os.environ.get("ETHICA_GROUP")    or cfg.get("group", platform.node())
+    widgets     = cfg.get("widgets", [])
 
     if not token or token == "REPLACE_ME":
         print("Set ETHICA_TOKEN env var or 'token' in config "
@@ -334,7 +367,7 @@ def main():
             label = w.get("name", w.get("type", "?"))
             try:
                 name, kind, status, lat, ver, note = run_widget(w)
-                code = push(home, token, name, kind, status, lat, ver, note)
+                code = push(home, token, name, kind, status, lat, ver, note, agent_group)
                 print(f"  [{status:<9}] {name}  → {code}")
             except NotImplementedError as e:
                 print(f"  [SKIP     ] {label}: {e}", file=sys.stderr)

widgets/custom-cmd.yml

@@ -0,0 +1,33 @@
+# Custom command examples.
+# Exit code convention: 0 = online, 1 = degraded, anything else = offline.
+
+widgets:
+  # Health endpoint that returns HTTP 200
+  - type: custom_cmd
+    name: App health
+    cmd: curl -sf https://myapp.example.com/health > /dev/null && exit 0 || exit 2
+
+  # Disk space on a non-root mount
+  - type: custom_cmd
+    name: Disk /backup
+    cmd: |
+      PCT=$(df /backup | awk 'NR==2{gsub(/%/,"",$5); print $5}')
+      [ "$PCT" -gt 90 ] && exit 2
+      [ "$PCT" -gt 80 ] && exit 1 || exit 0
+
+  # SSL certificate expiry (degraded < 14 days, offline < 3 days)
+  - type: custom_cmd
+    name: SSL expiry
+    cmd: |
+      DAYS=$(echo | openssl s_client -connect myapp.example.com:443 2>/dev/null \
+        | openssl x509 -noout -enddate 2>/dev/null \
+        | sed 's/notAfter=//' \
+        | xargs -I{} python3 -c "import datetime; e=datetime.datetime.strptime('{}','%b %d %H:%M:%S %Y %Z'); print((e-datetime.datetime.utcnow()).days)" 2>/dev/null)
+      [ -z "$DAYS" ] && exit 2
+      [ "$DAYS" -lt 3 ] && exit 2
+      [ "$DAYS" -lt 14 ] && exit 1 || exit 0
+
+  # Systemd service check
+  - type: custom_cmd
+    name: my-service.service
+    cmd: systemctl is-active --quiet my-service && exit 0 || exit 2

widgets/database-postgres.yml

@@ -0,0 +1,21 @@
+# PostgreSQL — process check + port + optional replication lag custom check.
+
+widgets:
+  - type: process_check
+    name: PostgreSQL
+    process: postgres
+
+  - type: port_check
+    name: PG Port
+    host: localhost
+    port: 5432
+    timeout: 5
+
+  # Optional: check replication lag (requires psql access).
+  # Exit 0 = ok, exit 1 = lag > threshold, exit 2 = cannot connect.
+  # - type: custom_cmd
+  #   name: PG Replication lag
+  #   cmd: |
+  #     LAG=$(psql -U postgres -tAc "SELECT EXTRACT(EPOCH FROM (now() - pg_last_xact_replay_timestamp()))::int" 2>/dev/null)
+  #     [ -z "$LAG" ] && exit 2
+  #     [ "$LAG" -gt 300 ] && exit 1 || exit 0

widgets/database-redis.yml

@@ -0,0 +1,17 @@
+# Redis — process check + port + optional ping.
+
+widgets:
+  - type: process_check
+    name: Redis
+    process: redis-server
+
+  - type: port_check
+    name: Redis Port
+    host: localhost
+    port: 6379
+    timeout: 3
+
+  # Optional: PING check (requires redis-cli).
+  # - type: custom_cmd
+  #   name: Redis PING
+  #   cmd: redis-cli ping | grep -q PONG && exit 0 || exit 2

widgets/docker-host.yml

@@ -0,0 +1,20 @@
+# Docker host — disk for docker data dir + Docker daemon check.
+
+widgets:
+  - type: system_disk
+    name: Docker disk
+    path: /var/lib/docker
+    degrade_threshold: 75
+    offline_threshold: 90
+
+  - type: custom_cmd
+    name: Docker daemon
+    cmd: docker info > /dev/null 2>&1 && exit 0 || exit 2
+
+  # Optional: count running containers (degraded if 0, offline if docker is down).
+  # - type: custom_cmd
+  #   name: Containers running
+  #   cmd: |
+  #     COUNT=$(docker ps -q 2>/dev/null | wc -l)
+  #     [ $? -ne 0 ] && exit 2
+  #     [ "$COUNT" -eq 0 ] && exit 1 || exit 0

widgets/linux-base.yml

@@ -0,0 +1,19 @@
+# Linux base — CPU, memory and disk.
+# Paste into your ethica-agent.yml widgets section.
+
+widgets:
+  - type: system_cpu
+    name: CPU
+    degrade_threshold: 80
+    offline_threshold: 95
+
+  - type: system_memory
+    name: Memory
+    degrade_threshold: 85
+    offline_threshold: 97
+
+  - type: system_disk
+    name: Disk /
+    path: /
+    degrade_threshold: 80
+    offline_threshold: 95

widgets/web-nginx.yml

@@ -0,0 +1,22 @@
+# Nginx web server — process check + HTTP/HTTPS endpoint + ports.
+# Replace MY_DOMAIN with the actual domain.
+
+widgets:
+  - type: process_check
+    name: nginx
+    process: nginx
+
+  - type: http_check
+    name: HTTPS
+    url: https://MY_DOMAIN
+    timeout: 10
+
+  - type: port_check
+    name: Port 80
+    host: localhost
+    port: 80
+
+  - type: port_check
+    name: Port 443
+    host: localhost
+    port: 443