fix: reject unprovenanced ML artifacts

tests/test_scoring_engine_ml.py

@@ -59,5 +59,6 @@ def test_score_all_ml_preserves_classic_fallback_when_weights_missing(monkeypatc
     scored = engine.score_all_ml(_complete_metrics())
 
     assert scored["ml_mode"] is False
-    assert scored["ml_error"] == "ML weights not found — run ml/optimizer.py"
+    assert scored["ml_error"]
+    assert scored["ml_artifact"]["valid"] is False
     assert "classic_score" not in scored