feat(dashboard): an Alerts widget beside the Overview; the Overview's overflow links to Alerts and Monitors (#5510)

The Alerts widget (/alerts/widget) is shaped like every widget: a verdict
("2 firing, 1 silent." / "Nothing firing: all 12 rules are quiet."), then
what is firing or silent with its reading now, what the rule fires on and
since when, then what someone acknowledged and who; a line when alert
email failed in the last 7 days or is retrying (outbox.alert_trouble),
linking admins to Settings -> System; the quiet rules counted, a covering
rule once. It sits beside the Overview in the recommended layout (6x4
each).

The Overview's "+N more" line pointed at the removed /status/; alerts now
link to /alerts/ and monitors to /monitors/.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

steward/alerts/routes.py

@@ -178,6 +178,19 @@ async def active_fragment():
         rules_off=sum(1 for r in rules if not r.enabled))
 
 
+@alerts_bp.get("/widget")
+@require_role(UserRole.viewer)
+async def widget():
+    """The dashboard Alerts widget (#5510): a verdict, what is firing or
+    silent, then what someone has taken, a line when alert email is failing,
+    and the quiet rules counted."""
+    async with current_app.db_sessionmaker() as db:
+        active = await alert_svc.active_rows(db, datetime.now(timezone.utc))
+        view = alert_svc.widget_view(active, await alert_svc.list_rules(db))
+        email = await outbox.alert_trouble(db)
+    return await render_template("alerts/widget.html", **view, email=email)
+
+
 def _delivery(label: str, sent: bool | None, error: str | None) -> dict | None:
     """A channel's outcome as recorded on the event; None when not tried."""
     if sent is None:

steward/alerts/service.py

@@ -231,11 +231,13 @@ async def active_rows(db, now: datetime) -> list[dict]:
             since = started.get(rule.id, state.last_transition_at)
         else:
             since = state.last_transition_at
+        value = await latest_value(db, rule)
         rows.append({
             "state": state, "rule": rule, "label": label, "since": since,
             "breach_since": started.get(rule.id) if label != "no_data" else None,
             "acked_by": names.get(state.acknowledged_by, state.acknowledged_by),
-            "says": alert_sentence(rule, label, await latest_value(db, rule), since, now),
+            "value": value,
+            "says": alert_sentence(rule, label, value, since, now),
             "url": await resource_url(db, rule),
         })
     epoch = datetime.max.replace(tzinfo=timezone.utc)
@@ -243,6 +245,19 @@ async def active_rows(db, now: datetime) -> list[dict]:
     return rows
 
 
+def widget_view(active: list[dict], rules: list[AlertRule]) -> dict:
+    """The dashboard Alerts widget (#5510) from active_rows and every rule:
+    the alerts in that order, each with what its rule fires on, and how many
+    enabled rules, as people made them, are quiet (a covering rule counts
+    once, and is quiet only while every host under it is).
+    {"rows", "quiet", "total"}."""
+    busy = {r["rule"].id for r in active}
+    busy |= {r["rule"].parent_id for r in active if r["rule"].parent_id}
+    made = [r for r in rules if r.enabled and not r.parent_id]
+    return {"rows": [{**r, "condition": rule_condition(r["rule"])} for r in active],
+            "quiet": sum(1 for r in made if r.id not in busy), "total": len(made)}
+
+
 def watches_host(rule: AlertRule, host_name: str, monitor_keys: set) -> bool:
     """Whether a rule watches something on this host: its agent series
     ("<host>" or "<host>:<part>"), one of its containers ("<host>/<name>"),

steward/core/outbox.py

@@ -35,7 +35,7 @@ from datetime import datetime, timedelta, timezone
 from email.message import EmailMessage
 from typing import TYPE_CHECKING, Callable
 
-from sqlalchemy import insert, or_, select, text, update
+from sqlalchemy import and_, insert, or_, select, text, update
 
 from steward.models.outbox import EmailOutbox, email_outbox_events
 
@@ -434,3 +434,27 @@ async def outgoing(db, now: datetime | None = None, limit: int = 20) -> dict:
     shown = sum(counts.get(s, 0) for s in open_states)
     return {"rows": rows, "more": max(shown - len(rows), 0), "sent": counts.get("sent", 0),
             "problems": problems}
+
+
+async def alert_trouble(db, now: datetime | None = None) -> dict:
+    """The dashboard Alerts widget's delivery line (#5510): alert emails of
+    the last SHOWN_DAYS that failed, and those still retrying (tried, not yet
+    accepted), and the line that says so ("" when neither).
+    {"failed": n, "retrying": n, "said": str}."""
+    from sqlalchemy import func
+    now = now or datetime.now(timezone.utc)
+    rows = dict((await db.execute(
+        select(EmailOutbox.state, func.count())
+        .where(EmailOutbox.kind == "alert", EmailOutbox.created_at >= now - timedelta(days=SHOWN_DAYS),
+               or_(EmailOutbox.state == "failed",
+                   and_(EmailOutbox.state.in_(("pending", "sending")), EmailOutbox.attempts > 0)))
+        .group_by(EmailOutbox.state))).all())
+    failed = rows.get("failed", 0)
+    retrying = rows.get("pending", 0) + rows.get("sending", 0)
+    said = []
+    if failed:
+        said.append(f"{failed} alert email{'s' if failed != 1 else ''} failed in the last {SHOWN_DAYS} days")
+    if retrying:
+        said.append(f"{retrying} still retrying" if failed
+                    else f"{retrying} alert email{'s' if retrying != 1 else ''} still retrying")
+    return {"failed": failed, "retrying": retrying, "said": "; ".join(said)}

steward/core/widgets.py

@@ -70,6 +70,16 @@ WIDGET_REGISTRY: dict[str, dict] = {
         "poll": True,
         "params": [],
     },
+    "alerts": {
+        "key": "alerts",
+        "label": "Alerts",
+        "description": "What is firing or silent, its reading and since when, then what someone has acknowledged; says when alert email is failing",
+        "hx_url": "/alerts/widget",
+        "detail_url": "/alerts/",
+        "plugin": None,
+        "poll": True,
+        "params": [],
+    },
     "hosts_overview": {
         "key": "hosts_overview",
         "label": "Hosts — Overview",
@@ -305,13 +315,15 @@ RETIRED = {
     "uptime_summary": "monitors",         # Monitors shows 24h/7d/30d uptime per monitor (#5250)
 }
 
-# The recommended dashboard (#5213): what needs attention first, then the
+# The recommended dashboard (#5213): what needs attention first (the
+# Overview, and beside it the Alerts, #5510), then the
 # fleet, then each integration. (key, width, height, config) on the 12-column
 # grid, one cell 70px tall; a widget whose plugin is off is skipped and the
 # rest close up. Seeded into the first dashboard and offered by
 # "Reset to the recommended layout" in edit mode.
 RECOMMENDED: list[tuple[str, int, int, dict]] = [
-    ("status_overview", 12, 3, {}),
+    ("status_overview", 6, 4, {}),
+    ("alerts", 6, 4, {}),
     ("hosts_overview", 6, 7, {}),
     ("monitors", 6, 7, {"show_down_only": "no", "limit": 10}),
     ("docker_containers", 6, 6, {}),
@@ -342,7 +354,7 @@ def recommended_layout(plugins_cfg: dict) -> list[dict]:
 # Group each widget for the picker (Core monitors / Monitoring capabilities /
 # Integrations), mirroring the Settings → Plugins taxonomy. Capability plugins
 # are host/monitoring facets; traefik/unifi are discrete vendor integrations.
-_CORE_WIDGETS = {"monitors", "status_overview", "hosts_overview"}
+_CORE_WIDGETS = {"monitors", "status_overview", "alerts", "hosts_overview"}
 _INTEGRATION_PLUGINS = {"traefik", "unifi"}
 GROUP_LABELS = {
     "core": "Core monitors",

steward/templates/alerts/widget.html

@@ -0,0 +1,58 @@
+{# alerts/widget.html — the dashboard Alerts widget (#5510), in the widget
+   shape (snippet #5259): a verdict; what is firing or silent, its reading
+   now against what the rule fires on, and since when; then what someone has
+   acknowledged, by whom and when; a line when alert email is failing or
+   retrying; the quiet rules counted. Rows come from alert_svc.active_rows,
+   so they read in the Alerts page's order. #}
+{% from "_macros.html" import status_dot %}
+{% set firing = rows|selectattr("label", "equalto", "firing")|list|length %}
+{% if not total and not rows %}
+<div class="w-verdict" style="color:var(--text-muted);">
+  No alert rules yet.{% if session.user_role in ('operator', 'admin') %} <a href="/alerts/rules/new">Add a rule →</a>{% endif %}
+</div>
+{% else %}
+<div class="w-verdict {{ 'tone-crit' if firing else ('tone-warn' if rows else 'tone-ok') }}">
+  {%- if firing %}{{ firing }} firing{% else %}Nothing firing{% endif %}
+  {%- for key, word in (("no_data", "silent"), ("acknowledged", "acknowledged"), ("pending", "pending")) %}
+  {%- set n = rows|selectattr("label", "equalto", key)|list|length %}
+  {%- if n %}, {{ n }} {{ word }}{% endif %}
+  {%- endfor %}
+  {%- if not rows %}: {{ "the rule is" if total == 1 else "all " ~ total ~ " rules are" }} quiet{% endif %}.
+</div>
+{% if rows %}
+<table class="w-table">
+  <colgroup><col><col style="width:5rem"><col style="width:5.5rem"></colgroup>
+  <thead><tr><th>Alert</th><th>Now</th><th>Since</th></tr></thead>
+  <tbody>
+  {% for r in rows %}
+  {% set rule = r.rule %}
+  <tr>
+    <td class="w-name" title="{{ rule.name }} · {{ rule.resource_name }}">
+      {{- status_dot(r.label) -}}
+      {% if session.user_id %}<a href="/alerts/history?rule={{ rule.id }}">{{ rule.name }}</a>{% else %}{{ rule.name }}{% endif %}
+      {%- if r.label == "acknowledged" %}
+      <span class="w-sub">{{ rule.resource_name }} · acknowledged{% if r.acked_by %} by {{ r.acked_by }}{% endif %}{% if r.state.acknowledged_at %} {{ r.state.acknowledged_at|ago }}{% endif %}</span>
+      {%- elif r.label == "no_data" %}
+      <span class="w-sub tone-warn">{{ rule.resource_name }} · no reading</span>
+      {%- else %}
+      <span class="w-sub">{{ rule.resource_name }} · fires on {{ r.condition }}</span>
+      {%- endif %}
+    </td>
+    <td class="{{ 'tone-crit' if r.label == 'firing' else ('tone-warn' if r.label == 'pending') }}">
+      {%- if r.label == "no_data" %}—{% else %}{{ metric_value(rule.metric_name, r.value) }}{% endif -%}
+    </td>
+    <td>{% if r.since %}{{ r.since|ago }}{% else %}—{% endif %}</td>
+  </tr>
+  {% endfor %}
+  </tbody>
+</table>
+{% endif %}
+{% if email.said %}
+<div class="{{ 'tone-crit' if email.failed else 'tone-warn' }}" style="font-size:0.75rem;margin-top:0.4rem;">
+  {{- email.said }}{% if session.user_role == "admin" %} — <a href="/settings/system/">Outgoing email</a>{% endif %}
+</div>
+{% endif %}
+{% if rows and quiet %}
+<div class="w-more">+{{ quiet }} rule{{ "s" if quiet != 1 }} quiet{% if session.user_id %} — <a href="/alerts/">Alerts</a>{% endif %}</div>
+{% endif %}
+{% endif %}

steward/templates/status/widget.html

@@ -54,11 +54,13 @@
     {% if e.since %}<span style="color:var(--text-dim);font-size:0.75rem;white-space:nowrap;">{% if e.since_at_least %}over {% endif %}{{ e.since|ago }}</span>{% endif %}
   </div>
   {% endfor %}
-  {% if alerts|length > 6 or problems|length > 8 %}
+  {# What didn't fit, each linking to its own list (#5510). #}
+  {% set more_alerts, more_monitors = alerts|length - 6, problems|length - 8 %}
+  {% if more_alerts > 0 or more_monitors > 0 %}
   <div style="color:var(--text-dim);font-size:0.75rem;padding-top:0.2rem;">
-    {% if alerts|length > 6 %}+{{ alerts|length - 6 }} more alerts{% endif %}
-    {% if problems|length > 8 %}{% if alerts|length > 6 %} · {% endif %}+{{ problems|length - 8 }} more monitors{% endif %}
-    {% if session.user_id %} — <a href="/status/">Status</a>{% endif %}
+    {%- if more_alerts > 0 %}{% if session.user_id %}<a href="/alerts/">+{{ more_alerts }} more alerts</a>{% else %}+{{ more_alerts }} more alerts{% endif %}{% endif %}
+    {%- if more_alerts > 0 and more_monitors > 0 %} · {% endif %}
+    {%- if more_monitors > 0 %}{% if session.user_id %}<a href="/monitors/">+{{ more_monitors }} more monitors</a>{% else %}+{{ more_monitors }} more monitors{% endif %}{% endif -%}
   </div>
   {% endif %}
 </div>