feat(docker): retention + hourly rollup for metrics/events with Settings windows

Bounds Docker time-series growth (the main scaling concern). New
docker_metrics_hourly table + docker_006 migration; a plugin retention module
(docker.run_retention capability) rolls raw docker_metrics older than the raw
window into hourly averages (idempotent upsert), deletes the rolled raw rows,
then prunes stale rollups + lifecycle events. Core cleanup.py drives it each
hourly run via the capability (no plugin-model import), reading the three
retention windows fresh from settings so changes apply without restart (rule 25).

Settings → "Thresholds & Retention" gains a Docker retention card (raw /
rolled-up / events windows, working defaults 7/90/30 days). Unit tests cover the
hour-aligned cutoff/bucketing helpers; integration test exercises the real
rollup-average + prune across both windows.

Milestone 77 task #941.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016Jg27rgypiW2efULXJDtMC

steward/templates/settings/_tabs.html

@@ -3,7 +3,7 @@
 <div style="display:flex;gap:0;border-bottom:1px solid var(--border-mid);margin-bottom:1.5rem;">
   {% set tabs = [
     ("general",       "General",        "/settings/general/"),
-    ("thresholds",    "Thresholds",     "/settings/thresholds/"),
+    ("thresholds",    "Thresholds & Retention", "/settings/thresholds/"),
     ("notifications", "Notifications",  "/settings/notifications/"),
     ("reports",       "Reports",        "/settings/reports/"),
     ("auth",          "Auth",           "/settings/auth/"),

steward/templates/settings/thresholds.html

@@ -1,6 +1,6 @@
 {# steward/templates/settings/thresholds.html — tunable degraded/critical cutoffs #}
 {% extends "base.html" %}
-{% block title %}Settings — Thresholds — Steward{% endblock %}
+{% block title %}Settings — Thresholds & Retention — Steward{% endblock %}
 {% block content %}
 {% set active_tab = "thresholds" %}
 {% include "settings/_tabs.html" %}
@@ -42,6 +42,34 @@
   {{ pair("Uptime / SLA", "uptime_warn", "thresholds.uptime_warn", "uptime_crit", "thresholds.uptime_crit", "%", "Lower is worse — amber below warn, red below crit.", step="0.1") }}
 </div>
 
+{% macro days(label, field, key, hint) %}
+<div class="form-group" style="margin-bottom:1.1rem;">
+  <label>{{ label }} <span style="color:var(--text-muted);font-size:0.8rem;">(days)</span></label>
+  <div style="margin-top:0.25rem;">
+    <input type="number" name="{{ field }}" min="1" step="1"
+           value="{{ settings[key] }}" style="max-width:110px;">
+  </div>
+  {% if hint %}<div style="font-size:0.78rem;color:var(--text-muted);margin-top:0.3rem;">{{ hint }}</div>{% endif %}
+</div>
+{% endmacro %}
+
+<div class="card" style="max-width:640px;margin-top:1.5rem;">
+  <h2 class="section-title" style="margin-bottom:0.5rem;">Docker data retention</h2>
+  <p style="font-size:0.82rem;color:var(--text-muted);margin-bottom:1.25rem;">
+    Bounds how much Docker history is stored. Raw per-sample container metrics are
+    kept for the raw window, then rolled up into hourly averages kept for the
+    rollup window; lifecycle events are kept for the events window. Applied by the
+    hourly cleanup task — a change takes effect on its next run.
+  </p>
+
+  {{ days("Raw metrics", "docker_metrics_raw_days", "docker.retention.metrics_raw_days",
+          "Keep per-sample container CPU/memory points this long, then roll up to hourly.") }}
+  {{ days("Rolled-up metrics", "docker_metrics_rollup_days", "docker.retention.metrics_rollup_days",
+          "Keep the hourly-averaged series this long for multi-day history.") }}
+  {{ days("Lifecycle events", "docker_events_days", "docker.retention.events_days",
+          "Keep container start/stop/die/health history this long.") }}
+</div>
+
 <div style="margin-top:1rem;display:flex;align-items:center;gap:1rem;">
   <button type="submit" class="btn">Save</button>
   <span style="font-size:0.82rem;color:var(--text-muted);">Takes effect immediately — no restart.</span>