Default zero agent to Newton visualizer

Author: StafaH

source/isaaclab_rl/changelog.d/fix-zero-agent-newton-visualizer.rst

@@ -0,0 +1,5 @@
+Changed
+^^^^^^^
+
+* Changed the zero agent to use the Newton GL visualizer by default. Pass ``--viz kit`` to keep using the Kit
+  visualizer.

source/isaaclab_rl/isaaclab_rl/entrypoints/simple_agents.py

@@ -245,8 +245,8 @@ def _parse_args(argv: list[str] | None, policy: PolicyName) -> argparse.Namespac
     )
     # append AppLauncher cli args
     add_launcher_args(parser)
-    # simple agents should open Kit visualizer by default
-    parser.set_defaults(visualizer=["kit"])
+    # Keep the zero agent on the kitless default path while preserving the random-agent behavior.
+    parser.set_defaults(visualizer=["newton_gl"] if policy == "zero" else ["kit"])
     args_cli, hydra_args = setup_preset_cli(parser, argv)
     sys.argv = [sys.argv[0]] + hydra_args
     return args_cli

source/isaaclab_rl/test/test_entrypoints.py

@@ -184,6 +184,20 @@ def test_zero_agent_supports_direct_multi_agent_action_spaces() -> None:
     assert torch.equal(actions["object"], torch.zeros(3, 1, dtype=torch.int64))
 
 
+@pytest.mark.parametrize(("policy", "expected_visualizer"), [("zero", ["newton_gl"]), ("random", ["kit"])])
+def test_simple_agent_default_visualizer_depends_on_policy(
+    policy: _simple_agents.PolicyName,
+    expected_visualizer: list[str],
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    """The zero agent defaults to Newton visualization without changing the random agent."""
+    monkeypatch.setattr(sys, "argv", ["pytest"])
+
+    args = _simple_agents._parse_args([], policy)
+
+    assert args.visualizer == expected_visualizer
+
+
 def test_zero_agent_rejects_invalid_config_before_launch(monkeypatch: pytest.MonkeyPatch) -> None:
     """Unsupported task presets fail cleanly before a simulator backend is initialized."""