Split the worker into a distribution of its own

A cluster or GPU host installs `pip install fluksio-worker` and gets the
agent and the runner, not psycopg, numpy and the MCP SDK. The engine
depends on it as a workspace member, so the file it launches node code
with is the same file a remote worker runs — which is what keeps a node
unable to tell the difference. Copying the two files by hand still works.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

backend/fluksio/flow/controller.py

@@ -20,6 +20,7 @@ from typing import Any, cast
 
 from fastapi import FastAPI
 from fastapi.concurrency import run_in_threadpool
+from fluksio_worker.worker_main import load_function
 
 from fluksio.core.config import settings
 from fluksio.flow.alerts import AlertManager
@@ -60,7 +61,6 @@ from fluksio.flow.secrets import SecretNotFound, resolve_params
 from fluksio.flow.state import MemoryState, StateBackend
 from fluksio.flow.store import LIB_DIR, FlowNotFound, FlowStore, LibNotFound
 from fluksio.flow.supervision import Supervisor
-from fluksio.flow.worker_main import load_function
 from fluksio.flow.workers import PythonWorkerPool
 
 logger = logging.getLogger(__name__)