feat(music-lab): cover 16:9 landscape 생성 + 메타데이터 프로페셔널화

- cover.py: DALL·E 3 → 1792x1024, gpt-image-1 → 1536x1024 (모델별 자동),
  prompt에 'cinematic landscape composition' 명시. OPENAI_IMAGE_SIZE env로 override 가능.
- metadata.py: prompt를 list+join 패턴으로 재구성 (인접 문자열/+ 충돌 해결)
  + lofi 채널 카피라이터 페르소나 부여. description 5-7섹션 구조 명시:
  후크/분위기/사용시나리오/챕터/시청권장/콜투액션/해시태그.
  mix vs single 분기 + tags 가이드 + 출력 JSON schema 명시.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

music-lab/app/pipeline/cover.py

@@ -113,6 +113,25 @@ async def generate(*, pipeline_id: int, genre: str, prompt_template: str,
     }
 
 
+def _get_image_size(model: str) -> str:
+    """모델별 16:9에 가장 가까운 landscape 사이즈.
+
+    OPENAI_IMAGE_SIZE 환경변수로 override 가능.
+    - gpt-image-1: 1536x1024 (3:2)
+    - dall-e-3: 1792x1024 (7:4)
+    - 기타: 1024x1024 (square 폴백)
+    """
+    override = os.getenv("OPENAI_IMAGE_SIZE", "")
+    if override:
+        return override
+    m = (model or "").lower()
+    if "dall-e-3" in m or "dalle3" in m:
+        return "1792x1024"
+    if "gpt-image" in m:
+        return "1536x1024"
+    return "1024x1024"
+
+
 async def _generate_with_dalle(prompt_template: str, mood: str,
                                 feedback: str, out_path: str,
                                 *, api_key: str, model: str,
@@ -124,13 +143,15 @@ async def _generate_with_dalle(prompt_template: str, mood: str,
         prompt = f"{prompt}, {mood} mood"
     if feedback:
         prompt = f"{prompt}. 추가 지시: {feedback}"
-    prompt = f"{prompt}, no text, high quality"
+    # cinematic landscape 명시 — 16:9 영상에 시각적으로 fit하도록 구도 유도
+    prompt = f"{prompt}, no text, high quality, cinematic landscape composition, wide aspect"
 
+    image_size = _get_image_size(model)
     async with httpx.AsyncClient(timeout=DALLE_TIMEOUT_S) as client:
         resp = await client.post(
             "https://api.openai.com/v1/images/generations",
             headers={"Authorization": f"Bearer {api_key}"},
-            json={"model": model, "prompt": prompt, "size": "1024x1024", "n": 1},
+            json={"model": model, "prompt": prompt, "size": image_size, "n": 1},
         )
         resp.raise_for_status()
         data = resp.json()["data"][0]