kokoro: persist custom voices across container recreate

Wrapper only enumerates one voice directory (settings.voices_dir,
default /app/api/src/voices/v1_0 — inside the container's writable
layer, not bind-mounted). Override via VOICES_DIR=/app/user_voices
(host bind mount) and add a command shim that cp -r's built-ins from
the in-image v1_0 into user_voices on every start. Built-ins re-seed
fresh from the image (so upgrades that add voices propagate); custom
.pt files in user_voices are preserved (cp -r is additive).

Also adds scripts/blend_kokoro_voice.py + a playbook around it that
mirrors the wrapper's request-time voice="a(w)+b(w)" math but writes
the result as a named .pt to user_voices, making it discoverable via
GET /v1/audio/voices and persistent across recreate. Defaults to
athena = af_bella(2)+af_aoede(1) normalized.

playbooks/deploy-kokoro.yaml

@@ -32,6 +32,17 @@ steps:
     shell: mkdir -p {{ user_voices_dir }}
     creates: "{{ user_voices_dir }}"
 
+  - name: Chown user-voices to 1001:1001 (appuser inside container)
+    # The compose `command:` shim runs as appuser (image USER) and
+    # cp's the in-image built-in voicepacks into this dir on every
+    # container start. appuser is uid 1001 inside the container; the
+    # host has no user with that uid, so we chown numerically. Once
+    # set, the shim can write; the blend script (run via docker exec
+    # -u 0) writes as root, producing 0644 files appuser can read.
+    shell: chown 1001:1001 {{ user_voices_dir }}
+    sudo: true
+    when: '[ "$(stat -c %u:%g {{ user_voices_dir }})" != "1001:1001" ]'
+
   - name: Ensure voices dir exists (used only if compose mount is enabled)
     shell: mkdir -p {{ voices_dir }}
     creates: "{{ voices_dir }}"