0.31.1: make dashboard metrics honest

- Container card now compares compose-only counts across hosts: agents
  report compose_running in /agent/system (pre-0.31.1 agents fall back to
  the all-containers number); card retitled, ResourceBar stat labelled
  'Containers (all)'.
- Uptime is sampled every 5 min (background loop + opportunistic on read)
  and charted as daily averages instead of a once-a-day snapshot; no
  sample is written when no compose containers exist (was: fake 100%).
  Legacy daily entries in uptime.jsonl still count; file pruned at startup.
- Funnel stage 'monitored' is now per-stack and real: stacks with an
  enabled local auto-update policy (was: global webhook-exists toggle).
  Frontend label renamed to 'Auto-managed'.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

README.md

@@ -172,14 +172,16 @@ as intuitive as Dockge, as capable as Portainer for Compose workflows.
   (`bg`/`card`/`accent`) are remapped onto the tokens so all pages reskin
   consistently. Typeface: Schibsted Grotesk (bundled, offline-friendly).
 - **Stack Health funnel** — the dashboard centrepiece. Five stages
-  (`discovered → running → healthy → updated → monitored`) from
-  `GET /api/dashboard/funnel` (30 s server cache, `?refresh=true` to bust).
-  SVG waterfall with alternating gradient / diagonal-hatch bars, value chips
-  and hover conversion/drop-off tooltips.
-- **Summary widgets** from `GET /api/dashboard/summary`: containers-running
-  card with per-host breakdown and an "Insights" chip (healthy-rate %), a
-  30-day uptime line chart (daily samples appended to `DATA_DIR/uptime.jsonl`),
-  and an ops contribution grid from audit-log activity with the peak weekday.
+  (`discovered → running → healthy → updated → auto-managed`) from
+  `GET /api/dashboard/funnel` (30 s server cache, `?refresh=true` to bust;
+  the last stage — API key `monitored` — counts stacks with an enabled
+  auto-update policy). SVG waterfall with alternating gradient /
+  diagonal-hatch bars, value chips and hover conversion/drop-off tooltips.
+- **Summary widgets** from `GET /api/dashboard/summary`: compose-containers
+  card with per-host breakdown (agents report a compose-only count) and an
+  "Insights" chip (healthy-rate %), a 30-day uptime line chart (sampled every
+  5 min into `DATA_DIR/uptime.jsonl`, charted as daily averages), and an ops
+  contribution grid from audit-log activity with the peak weekday.
   A 30/7-day range selector slices both series client-side.
 - **Explore prompt bar** under the funnel: typed queries or `/running`,
   `/stopped`, `/attention` tags deep-link to the Stacks page, which now

backend/agent_app.py

@@ -66,7 +66,7 @@ def _map_docker(exc: DockerError):
         raise HTTPException(status_code=code, detail=exc.detail or exc.error)
     raise exc  # falls through to the global 502 DockerError handler
 
-AGENT_VERSION = "0.31.0"
+AGENT_VERSION = "0.31.1"
 
 
 # --------------------------------------------------------------------------- #
@@ -210,7 +210,7 @@ def _disk_info() -> tuple[int, int]:
 def _system_info() -> dict:
     docker_version = ""
     host_os = ""
-    running = total = 0
+    running = total = compose_running = 0
     try:
         client = get_client()
         docker_version = safe_call(client.version).get("Version", "")
@@ -218,6 +218,12 @@ def _system_info() -> dict:
         host_os = info.get("OperatingSystem", "")
         running = info.get("ContainersRunning", 0)
         total = info.get("Containers", 0)
+        # Running compose-managed containers — the dashboard's container card
+        # compares this across hosts; counting everything would include this
+        # agent itself and skew the bars.
+        compose_running = len(
+            safe_call(client.api.containers, filters={"label": compose_service.COMPOSE_LABEL})
+        )
     except DockerError as exc:
         docker_version = f"unavailable ({exc.error})"
     mem_total, mem_used = _mem_info()
@@ -233,6 +239,7 @@ def _system_info() -> dict:
         "disk_used": disk_used,
         "containers_running": running,
         "containers_total": total,
+        "compose_running": compose_running,
     }
 
 

backend/main.py

@@ -35,7 +35,7 @@ from routers import (
     volumes,
     ws,
 )
-from services import schedule_service, template_service, update_service
+from services import dashboard_service, schedule_service, template_service, update_service
 
 logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger("stackpilot")
@@ -58,13 +58,15 @@ async def lifespan(app: FastAPI):
         logger.warning("Legacy template migration failed: %s", exc)
     update_task = asyncio.create_task(update_service.background_loop())
     schedule_task = asyncio.create_task(schedule_service.scheduler_loop())
+    uptime_task = asyncio.create_task(dashboard_service.uptime_sampler_loop())
     logger.info("StackPilot backend ready on port %s", settings.PORT)
     yield
     update_task.cancel()
     schedule_task.cancel()
+    uptime_task.cancel()
 
 
-app = FastAPI(title="StackPilot", version="0.31.0", lifespan=lifespan)
+app = FastAPI(title="StackPilot", version="0.31.1", lifespan=lifespan)
 
 app.add_middleware(
     CORSMiddleware,

backend/services/dashboard_service.py

@@ -19,7 +19,7 @@ from sqlmodel import Session, select
 from config import settings
 from docker_client import DockerError, get_client, safe_call
 from models.audit import AuditLog
-from models.setting import Webhook
+from models.auto_update import AutoUpdate
 from services import compose_service, update_service
 
 COMPOSE_LABEL = compose_service.COMPOSE_LABEL
@@ -28,6 +28,7 @@ FUNNEL_TTL = 30.0
 
 UPTIME_FILE = os.path.join(settings.DATA_DIR, "uptime.jsonl")
 UPTIME_DAYS = 30
+UPTIME_SAMPLE_INTERVAL = 300.0  # seconds between uptime samples
 
 _funnel_cache: dict = {"data": None, "ts": 0.0}
 
@@ -77,13 +78,15 @@ def _is_updated(containers: list[dict], cache: dict[str, dict]) -> bool:
     return True
 
 
-def _has_notify_target(session: Session) -> bool:
-    if settings.NOTIFY_WEBHOOKS:
-        return True
-    for wh in session.exec(select(Webhook)).all():
-        if wh.enabled:
-            return True
-    return False
+def _auto_managed_ids(session: Session) -> set[str]:
+    """Local stacks with an enabled auto-update policy."""
+    rows = session.exec(
+        select(AutoUpdate.stack_id).where(
+            AutoUpdate.enabled == True,  # noqa: E712 — SQL expression, not identity
+            AutoUpdate.agent_id == None,  # noqa: E711
+        )
+    ).all()
+    return set(rows)
 
 
 async def compute_funnel(session: Session, refresh: bool = False) -> dict:
@@ -98,7 +101,7 @@ async def compute_funnel(session: Session, refresh: bool = False) -> dict:
         raw = []
     by_project = _group_by_project(raw)
     update_cache = update_service.get_cache()
-    notify_configured = _has_notify_target(session)
+    auto_managed = _auto_managed_ids(session)
 
     running = healthy = updated = monitored = 0
     for stack_id in discovered_ids:
@@ -113,7 +116,8 @@ async def compute_funnel(session: Session, refresh: bool = False) -> dict:
         if not _is_updated(containers, update_cache):
             continue
         updated += 1
-        if notify_configured:
+        # Final stage: the stack also keeps itself fresh (auto-update enabled).
+        if stack_id in auto_managed:
             monitored += 1
 
     data = {
@@ -130,7 +134,12 @@ async def compute_funnel(session: Session, refresh: bool = False) -> dict:
 
 
 # --------------------------------------------------------------------------- #
-# Uptime series (one sample per day, JSONL on disk)
+# Uptime series (sampled every few minutes, aggregated to a daily mean)
+#
+# Each sample is one JSONL line {"ts": iso, "value": pct}; a background loop
+# samples every UPTIME_SAMPLE_INTERVAL and summary reads top up opportunistically,
+# so the daily value approximates real uptime instead of a once-a-day snapshot.
+# Legacy pre-0.31.1 lines {"date": d, "value": pct} still count as one sample.
 # --------------------------------------------------------------------------- #
 
 
@@ -156,40 +165,98 @@ def _append_uptime(entry: dict) -> None:
         fh.write(json.dumps(entry) + "\n")
 
 
+def _entry_date(e: dict) -> Optional[str]:
+    if "date" in e:
+        return e["date"]
+    ts = e.get("ts")
+    return ts[:10] if isinstance(ts, str) and len(ts) >= 10 else None
+
+
+def _latest_sample_ts(entries: list[dict]) -> float:
+    latest = 0.0
+    for e in entries:
+        ts = e.get("ts")
+        if not isinstance(ts, str):
+            continue
+        try:
+            latest = max(latest, datetime.fromisoformat(ts).timestamp())
+        except ValueError:
+            continue
+    return latest
+
+
 def _sample_uptime(raw: list[dict]) -> Optional[dict]:
-    """Append today's sample if not yet recorded. Uptime% = share of compose
-    containers currently running."""
-    today = datetime.now(timezone.utc).date().isoformat()
+    """Record a sample if the last one is older than the sample interval.
+
+    Uptime% = share of compose containers currently running. With no compose
+    containers at all there is nothing to be up — skip rather than fake 100%.
+    """
     entries = _read_uptime()
-    if any(e.get("date") == today for e in entries):
+    now = datetime.now(timezone.utc)
+    if now.timestamp() - _latest_sample_ts(entries) < UPTIME_SAMPLE_INTERVAL:
         return None
     labelled = [c for c in raw if (c.get("Labels") or {}).get(COMPOSE_LABEL)]
-    total = len(labelled)
+    if not labelled:
+        return None
     running = sum(1 for c in labelled if c.get("State") == "running")
-    value = round(running / total * 100, 1) if total else 100.0
-    entry = {"date": today, "value": value}
+    value = round(running / len(labelled) * 100, 1)
+    entry = {"ts": now.isoformat(timespec="seconds"), "value": value}
     _append_uptime(entry)
     return entry
 
 
+def prune_uptime_file() -> None:
+    """Drop samples older than twice the chart window (called at startup)."""
+    entries = _read_uptime()
+    if not entries:
+        return
+    cutoff = (datetime.now(timezone.utc).date() - timedelta(days=UPTIME_DAYS * 2)).isoformat()
+    kept = [e for e in entries if (_entry_date(e) or cutoff) >= cutoff]
+    if len(kept) == len(entries):
+        return
+    os.makedirs(settings.DATA_DIR, exist_ok=True)
+    tmp = UPTIME_FILE + ".tmp"
+    with open(tmp, "w", encoding="utf-8") as fh:
+        for e in kept:
+            fh.write(json.dumps(e) + "\n")
+    os.replace(tmp, UPTIME_FILE)
+
+
 def uptime_series(raw: list[dict]) -> list[dict]:
     _sample_uptime(raw)
-    entries = _read_uptime()
-    by_date = {e["date"]: e for e in entries if "date" in e}
+    by_date: dict[str, list[float]] = {}
+    for e in _read_uptime():
+        day = _entry_date(e)
+        value = e.get("value")
+        if day and isinstance(value, (int, float)):
+            by_date.setdefault(day, []).append(float(value))
     series = []
     today = datetime.now(timezone.utc).date()
     last_value: Optional[float] = None
     for i in range(UPTIME_DAYS - 1, -1, -1):
         day = (today - timedelta(days=i)).isoformat()
-        e = by_date.get(day)
-        if e is not None:
-            last_value = e.get("value")
+        samples = by_date.get(day)
+        if samples:
+            last_value = round(sum(samples) / len(samples), 1)
         # Days before monitoring started (or gaps) reuse the last known value
         # so the chart doesn't show artificial dips.
         series.append({"date": day, "value": last_value})
     return series
 
 
+async def uptime_sampler_loop() -> None:
+    """Background task: keep the uptime series fed even when nobody is
+    looking at the dashboard."""
+    prune_uptime_file()
+    while True:
+        try:
+            raw = await _containers_with_timeout()
+            await asyncio.to_thread(_sample_uptime, raw)
+        except (DockerError, asyncio.TimeoutError, OSError):
+            pass
+        await asyncio.sleep(UPTIME_SAMPLE_INTERVAL)
+
+
 # --------------------------------------------------------------------------- #
 # Ops (audit-log) activity
 # --------------------------------------------------------------------------- #

frontend/package.json

@@ -1,7 +1,7 @@
 {
   "name": "stackpilot-frontend",
   "private": true,
-  "version": "0.31.0",
+  "version": "0.31.1",
   "type": "module",
   "scripts": {
     "dev": "vite",

frontend/src/api/agents.ts

@@ -15,6 +15,7 @@ export interface AgentSystem {
   disk_used: number;
   containers_running: number;
   containers_total: number;
+  compose_running?: number; // agents < 0.31.1 don't report it
 }
 
 export const agentsApi = {

frontend/src/pages/Dashboard.tsx

@@ -207,7 +207,7 @@ const FUNNEL_STAGES = [
   { key: "running", label: "Running" },
   { key: "healthy", label: "Healthy" },
   { key: "updated", label: "Up to date" },
-  { key: "monitored", label: "Monitored" },
+  { key: "monitored", label: "Auto-managed" },
 ] as const;
 
 function StackHealthCard({
@@ -302,14 +302,19 @@ function ContainerCountCard({
 
   const hosts: { name: string; count: number }[] = [
     { name: "local", count: localRunning ?? 0 },
-    ...online.map((a, i) => ({ name: a.name, count: remote[i].data?.containers_running ?? 0 })),
+    // compose_running keeps the bars comparable with the local compose-only
+    // count; pre-0.31.1 agents only report the all-containers number.
+    ...online.map((a, i) => ({
+      name: a.name,
+      count: remote[i].data?.compose_running ?? remote[i].data?.containers_running ?? 0,
+    })),
   ];
   const total = hosts.reduce((s, h) => s + h.count, 0);
   const max = Math.max(...hosts.map((h) => h.count), 1);
 
   return (
     <div className="sp-card sp-rise flex flex-col p-5 sm:p-6" style={{ animationDelay: "60ms" }}>
-      <h2 className="sp-label">Containers running</h2>
+      <h2 className="sp-label">Compose containers running</h2>
       {loading ? (
         <div className="mt-3 space-y-3">
           <div className="sp-skeleton h-14 w-28" />
@@ -379,7 +384,9 @@ function UptimeCard({
           <UptimeChart data={values} />
         )}
       </div>
-      <p className="mt-2 text-xs text-sp-text-3">Daily share of compose containers running, this host.</p>
+      <p className="mt-2 text-xs text-sp-text-3">
+        Share of compose containers running on this host, sampled every 5 min (daily average).
+      </p>
     </div>
   );
 }
@@ -552,7 +559,7 @@ function ResourceBar({
       />
       <Stat
         icon={<Container className="h-5 w-5" />}
-        label="Containers"
+        label="Containers (all)"
         value={`${containersRunning} / ${containersTotal}`}
       />
       <Stat icon={<Server className="h-5 w-5" />} label="Docker" value={dockerVersion || "—"} />