feat(p1): faster-whisper engine + audio ingest + transcribe (CPU verified)

- engine/: FasterWhisperEngine 래퍼 + model_registry (turbo→CT2 repo)
- audio/ingest.py: ffprobe duration/size probe + 413 상한 훅
- cli transcribe: device-auto, model 오버라이드, 413 가드, model_used 출력
- 단위 테스트 3 (resolve_model, probe_media); README 갱신

검증(CPU): JFK 11s 클립 → 정확 전사, detected_lang=en. 10 tests pass, ruff clean.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

tests/test_engine_audio.py

@@ -0,0 +1,23 @@
+"""engine.model_registry / audio.ingest 경량 단위 테스트 (모델 로드 불요)."""
+from __future__ import annotations
+
+import pytest
+
+from luke_scribe.audio.ingest import probe_media
+from luke_scribe.engine.model_registry import resolve_model
+
+
+def test_resolve_model_turbo_maps_to_ct2_repo():
+    expected = "deepdml/faster-whisper-large-v3-turbo-ct2"
+    assert resolve_model("large-v3-turbo") == expected
+    assert resolve_model("turbo") == expected
+
+
+def test_resolve_model_standard_passthrough():
+    assert resolve_model("tiny") == "tiny"
+    assert resolve_model("large-v3") == "large-v3"
+
+
+def test_probe_media_missing_raises():
+    with pytest.raises(FileNotFoundError):
+        probe_media("/no/such/file.wav")