feat(tts): Phase 4 — Read Page Aloud with OS-native voices

Platform-dispatched TTS (spd-say + espeak-ng fallback on Linux, say on
macOS, PowerShell System.Speech on Windows) with a shared SpeakerButton
component. Tap to speak, tap again to stop; only one button speaks at
a time so two surfaces don't talk over each other. Text always travels
via argv (or a PowerShell here-string delivered through -EncodedCommand
on Windows) so user content never enters a shell string.

Mount points: DictationPage transcript footer, transcript viewer header,
per-step in MicroSteps. Settings gains a "Read aloud" accordion with
voice picker (lazy-loaded from the OS synth), rate slider 0.5-2.0x,
and a British-English test utterance.

Rust tests cover rate mapping, NaN handling, and Windows here-string
terminator safety. No pause/resume, no SSML, no cloud voices — that
stays out of scope per the Layer-1 roadmap.

src/lib/pages/SettingsPage.svelte

@@ -633,6 +633,47 @@
     }
   }
 
+  // Phase 4 Read Page Aloud. Voices are lazy-loaded on first open so
+  // Settings doesn't pay the `say -v ?` / PowerShell cost on every
+  // mount. An empty list renders as "System default" only.
+  let ttsVoices = $state([]);
+  let ttsVoicesLoaded = $state(false);
+  let ttsVoicesLoading = $state(false);
+  let ttsVoicesError = $state("");
+
+  async function refreshTtsVoices() {
+    if (ttsVoicesLoading) return;
+    ttsVoicesLoading = true;
+    ttsVoicesError = "";
+    try {
+      ttsVoices = await invoke("tts_list_voices");
+      ttsVoicesLoaded = true;
+    } catch (err) {
+      ttsVoicesError = String(err);
+    } finally {
+      ttsVoicesLoading = false;
+    }
+  }
+
+  async function toggleReadAloudSection() {
+    openSection = openSection === 'readAloud' ? null : 'readAloud';
+    if (openSection === 'readAloud' && !ttsVoicesLoaded) {
+      await refreshTtsVoices();
+    }
+  }
+
+  async function testReadAloudVoice() {
+    try {
+      await invoke("tts_speak", {
+        text: "This is Corbie reading aloud.",
+        rate: settings.ttsRate,
+        voice: settings.ttsVoice ?? null,
+      });
+    } catch (err) {
+      toasts.warn("Could not read aloud", String(err));
+    }
+  }
+
   onMount(async () => {
     try {
       await refreshRuntimeCapabilities();
@@ -1355,6 +1396,77 @@
         {/if}
       </div>
 
+      <!-- Read aloud (Phase 4 roadmap) -->
+      <div class="border-b border-border-subtle">
+        <button
+          class="flex items-center justify-between w-full py-4 px-5 text-left"
+          onclick={toggleReadAloudSection}
+        >
+          <h3 class="font-display text-[18px] italic text-text">Read aloud</h3>
+          <span class="text-text-tertiary text-[16px] leading-none">{openSection === 'readAloud' ? '−' : '+'}</span>
+        </button>
+        {#if openSection === 'readAloud'}
+          <div class="px-5 pb-5 animate-fade-in">
+            <p class="text-[11px] text-text-tertiary mb-4">
+              Uses your operating system's built-in voices. No audio leaves the machine.
+            </p>
+
+            <div class="mb-4">
+              <label for="tts-voice" class="block text-[10px] font-medium text-text-tertiary uppercase tracking-wider mb-2">Voice</label>
+              <select
+                id="tts-voice"
+                class="w-full bg-bg-input border border-border rounded-lg px-3 py-2 text-[12px] text-text focus:border-accent focus:outline-none"
+                value={settings.ttsVoice ?? ""}
+                onchange={(e) => { settings.ttsVoice = e.currentTarget.value || null; }}
+                disabled={ttsVoicesLoading}
+              >
+                <option value="">System default</option>
+                {#each ttsVoices as voice (voice.id)}
+                  <option value={voice.id}>
+                    {voice.name}{#if voice.language} · {voice.language}{/if}
+                  </option>
+                {/each}
+              </select>
+              {#if ttsVoicesError}
+                <p class="text-[11px] text-danger mt-2">{ttsVoicesError}</p>
+              {:else if ttsVoicesLoaded && ttsVoices.length === 0}
+                <p class="text-[11px] text-text-tertiary mt-2">
+                  No additional voices reported by the system synth. Install extra voices through your OS accessibility settings.
+                </p>
+              {/if}
+            </div>
+
+            <div class="mb-4">
+              <label for="tts-rate" class="block text-[10px] font-medium text-text-tertiary uppercase tracking-wider mb-2">
+                Rate · {settings.ttsRate.toFixed(1)}×
+              </label>
+              <input
+                id="tts-rate"
+                type="range"
+                min="0.5"
+                max="2.0"
+                step="0.1"
+                class="w-full accent-accent"
+                bind:value={settings.ttsRate}
+              />
+              <div class="flex justify-between text-[10px] text-text-tertiary mt-1">
+                <span>Slower</span>
+                <span>Normal</span>
+                <span>Faster</span>
+              </div>
+            </div>
+
+            <button
+              type="button"
+              class="px-3 py-2 rounded-lg bg-bg-elevated border border-border text-[12px] text-text hover:border-accent"
+              onclick={testReadAloudVoice}
+            >
+              Test voice
+            </button>
+          </div>
+        {/if}
+      </div>
+
       <!-- AI Assistant -->
       <div class="border-b border-border-subtle">
         <button