Fix Lyria API response parsing — audio in parts[1], not parts[0]

The Lyria API returns a candidates[].content.parts[] array where:
- parts[0] = text metadata (e.g. '<instrumental>')
- parts[1] = inlineData with the actual audio payload

Previous code only checked parts[0] for inlineData, which always failed
with 'No audio data in response'. Fixed by iterating all parts to find
the inlineData entry.

Successfully generated first real audio clip (lo-fi hip hop, 30s, 728KB MP3).

Also includes: .env.example, requirements updates, docker-compose fixes,
frontend vite config, logger setup.

backend/src/main.py

@@ -1,11 +1,16 @@
-from fastapi import FastAPI, HTTPException
+from fastapi import FastAPI, HTTPException, BackgroundTasks
 from fastapi.middleware.cors import CORSMiddleware
-from fastapi.responses import FileResponse, StreamingResponse
+from fastapi.responses import FileResponse
 from pathlib import Path
 import os
+from dotenv import load_dotenv
+
+# Load environment variables BEFORE importing lyria_client (it reads env at module level)
+load_dotenv(Path(__file__).parent.parent / ".env")
 
 from models import GenerateRequest, GenerateResponse, ClipList
 from database import init_db, create_clip, get_clips, get_clip, delete_clip, update_clip_status
+from lyria_client import generate_clip, save_clip_file
 
 app = FastAPI(title="Lyria Studio", version="0.1.0")
 
@@ -28,10 +33,44 @@ async def startup():
 async def health():
     return {"status": "ok", "version": "0.1.0"}
 
+async def _generate_and_save(clip_id: str, prompt: str, model: str, duration: int):
+    """Background task: call Lyria API and save the result."""
+    try:
+        result = await generate_clip(prompt, model, duration)
+        
+        if result["success"]:
+            file_info = await save_clip_file(clip_id, result["audio_bytes"])
+            await update_clip_status(
+                clip_id, 
+                status="ready", 
+                file_size=file_info["file_size"]
+            )
+        else:
+            await update_clip_status(
+                clip_id, 
+                status="failed", 
+                error=result.get("error", "Unknown error")
+            )
+    except Exception as e:
+        await update_clip_status(
+            clip_id, 
+            status="failed", 
+            error=str(e)
+        )
+
 @app.post("/api/generate", response_model=GenerateResponse)
-async def generate(request: GenerateRequest):
+async def generate(request: GenerateRequest, background_tasks: BackgroundTasks):
     clip_id = await create_clip(request.prompt, request.model, request.duration)
-    # TODO: T-003 — integrate Lyria API client
+    
+    # Kick off generation in background
+    background_tasks.add_task(
+        _generate_and_save,
+        clip_id,
+        request.prompt,
+        request.model,
+        request.duration
+    )
+    
     return GenerateResponse(
         id=clip_id,
         status="generating",