Use Redis for cache indexes - enables multi-worker scaling

The cache_manager now uses Redis hashes for the content_index and
ipfs_cids mappings. This allows multiple uvicorn workers to share
state, so files added by one worker are immediately visible to all
others.

- Added redis_client parameter to L1CacheManager
- Index lookups check Redis first, then fall back to in-memory
- Index updates go to both Redis and JSON file (backup)
- Migrates existing JSON indexes to Redis on first load
- Re-enabled workers=4 in uvicorn

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

server.py

@@ -57,10 +57,7 @@ IPFS_GATEWAY_URL = os.environ.get("IPFS_GATEWAY_URL", "")
 CACHE_DIR = Path(os.environ.get("CACHE_DIR", str(Path.home() / ".artdag" / "cache")))
 CACHE_DIR.mkdir(parents=True, exist_ok=True)
 
-# Initialize L1 cache manager (no L2 config - determined dynamically from token)
-cache_manager = L1CacheManager(cache_dir=CACHE_DIR)
-
-# Redis for persistent run storage
+# Redis for persistent run storage and shared cache index (multi-worker support)
 REDIS_URL = os.environ.get('REDIS_URL', 'redis://localhost:6379/5')
 parsed = urlparse(REDIS_URL)
 redis_client = redis.Redis(
@@ -73,6 +70,9 @@ redis_client = redis.Redis(
 RUNS_KEY_PREFIX = "artdag:run:"
 RECIPES_KEY_PREFIX = "artdag:recipe:"
 
+# Initialize L1 cache manager with Redis for shared state between workers
+cache_manager = L1CacheManager(cache_dir=CACHE_DIR, redis_client=redis_client)
+
 
 def save_run(run: "RunStatus"):
     """Save run to Redis."""
@@ -4014,6 +4014,5 @@ async def ui_run_partial(run_id: str, request: Request):
 
 if __name__ == "__main__":
     import uvicorn
-    # Note: workers disabled because cache_manager uses in-memory state
-    # that isn't shared between worker processes
-    uvicorn.run("server:app", host="0.0.0.0", port=8100)
+    # Workers enabled - cache indexes shared via Redis
+    uvicorn.run("server:app", host="0.0.0.0", port=8100, workers=4)