Initial commit

source/mindbot/config/extension.toml

@@ -0,0 +1,35 @@
+[package]
+
+# Semantic Versioning is used: https://semver.org/
+version = "0.1.0"
+
+# Description
+category = "isaaclab"
+readme  = "README.md"
+
+title = "Extension Template"
+author = "Isaac Lab Project Developers"
+maintainer = "Isaac Lab Project Developers"
+description="Extension Template for Isaac Lab"
+repository = "https://github.com/isaac-sim/IsaacLab.git"
+keywords = ["extension", "template", "isaaclab"]
+
+[dependencies]
+"isaaclab" = {}
+"isaaclab_assets" = {}
+"isaaclab_mimic" = {}
+"isaaclab_rl" = {}
+"isaaclab_tasks" = {}
+# NOTE: Add additional dependencies here
+
+[[python.module]]
+name = "mindbot"
+
+[isaac_lab_settings]
+# TODO: Uncomment and list any apt dependencies here.
+#       If none, leave it commented out.
+# apt_deps = ["example_package"]
+# TODO: Uncomment and provide path to a ros_ws
+#       with rosdeps to be installed. If none,
+#       leave it commented out.
+# ros_ws = "path/from/extension_root/to/ros_ws"

source/mindbot/docs/CHANGELOG.rst

@@ -0,0 +1,10 @@
+Changelog
+---------
+
+0.1.0 (2025-11-13)
+~~~~~~~~~~~~~~~~~~
+
+Added
+^^^^^
+
+* Created an initial template for building an extension or project based on Isaac Lab

source/mindbot/mindbot/__init__.py

@@ -0,0 +1,14 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+"""
+Python module serving as a project/extension template.
+"""
+
+# Register Gym environments.
+from .tasks import *
+
+# Register UI extensions.
+from .ui_extension_example import *

source/mindbot/mindbot/tasks/__init__.py

@@ -0,0 +1,17 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+"""Package containing task implementations for the extension."""
+
+##
+# Register Gym environments.
+##
+
+from isaaclab_tasks.utils import import_packages
+
+# The blacklist is used to prevent importing configs from sub-packages
+_BLACKLIST_PKGS = ["utils", ".mdp"]
+# Import all configs in this package
+import_packages(__name__, _BLACKLIST_PKGS)

source/mindbot/mindbot/tasks/direct/__init__.py

@@ -0,0 +1,6 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+import gymnasium as gym  # noqa: F401

source/mindbot/mindbot/tasks/direct/mindbot/__init__.py

@@ -0,0 +1,29 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+import gymnasium as gym
+
+from . import agents
+
+##
+# Register Gym environments.
+##
+
+
+gym.register(
+    id="Template-Mindbot-Direct-v0",
+    entry_point=f"{__name__}.mindbot_env:MindbotEnv",
+    disable_env_checker=True,
+    kwargs={
+        "env_cfg_entry_point": f"{__name__}.mindbot_env_cfg:MindbotEnvCfg",
+        "rl_games_cfg_entry_point": f"{agents.__name__}:rl_games_ppo_cfg.yaml",
+        "rsl_rl_cfg_entry_point": f"{agents.__name__}.rsl_rl_ppo_cfg:PPORunnerCfg",
+        "skrl_amp_cfg_entry_point": f"{agents.__name__}:skrl_amp_cfg.yaml",
+        "skrl_ippo_cfg_entry_point": f"{agents.__name__}:skrl_ippo_cfg.yaml",
+        "skrl_mappo_cfg_entry_point": f"{agents.__name__}:skrl_mappo_cfg.yaml",
+        "skrl_cfg_entry_point": f"{agents.__name__}:skrl_ppo_cfg.yaml",
+        "sb3_cfg_entry_point": f"{agents.__name__}:sb3_ppo_cfg.yaml",
+    },
+)

source/mindbot/mindbot/tasks/direct/mindbot/agents/__init__.py

@@ -0,0 +1,4 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause

source/mindbot/mindbot/tasks/direct/mindbot/agents/rl_games_ppo_cfg.yaml

@@ -0,0 +1,78 @@
+params:
+  seed: 42
+
+  # environment wrapper clipping
+  env:
+    # added to the wrapper
+    clip_observations: 5.0
+    # can make custom wrapper?
+    clip_actions: 1.0
+
+  algo:
+    name: a2c_continuous
+
+  model:
+    name: continuous_a2c_logstd
+
+  # doesn't have this fine grained control but made it close
+  network:
+    name: actor_critic
+    separate: False
+    space:
+      continuous:
+        mu_activation: None
+        sigma_activation: None
+
+        mu_init:
+          name: default
+        sigma_init:
+          name: const_initializer
+          val: 0
+        fixed_sigma: True
+    mlp:
+      units: [32, 32]
+      activation: elu
+      d2rl: False
+
+      initializer:
+        name: default
+      regularizer:
+        name: None
+
+  load_checkpoint: False # flag which sets whether to load the checkpoint
+  load_path: '' # path to the checkpoint to load
+
+  config:
+    name: cartpole_direct
+    env_name: rlgpu
+    device: 'cuda:0'
+    device_name: 'cuda:0'
+    multi_gpu: False
+    ppo: True
+    mixed_precision: False
+    normalize_input: True
+    normalize_value: True
+    num_actors: -1  # configured from the script (based on num_envs)
+    reward_shaper:
+      scale_value: 0.1
+    normalize_advantage: True
+    gamma: 0.99
+    tau : 0.95
+    learning_rate: 5e-4
+    lr_schedule: adaptive
+    kl_threshold: 0.008
+    score_to_win: 20000
+    max_epochs: 150
+    save_best_after: 50
+    save_frequency: 25
+    grad_norm: 1.0
+    entropy_coef: 0.0
+    truncate_grads: True
+    e_clip: 0.2
+    horizon_length: 32
+    minibatch_size: 16384
+    mini_epochs: 8
+    critic_coef: 4
+    clip_value: True
+    seq_length: 4
+    bounds_loss_coef: 0.0001

source/mindbot/mindbot/tasks/direct/mindbot/agents/rsl_rl_ppo_cfg.py

@@ -0,0 +1,38 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+from isaaclab.utils import configclass
+
+from isaaclab_rl.rsl_rl import RslRlOnPolicyRunnerCfg, RslRlPpoActorCriticCfg, RslRlPpoAlgorithmCfg
+
+
+@configclass
+class PPORunnerCfg(RslRlOnPolicyRunnerCfg):
+    num_steps_per_env = 16
+    max_iterations = 150
+    save_interval = 50
+    experiment_name = "cartpole_direct"
+    policy = RslRlPpoActorCriticCfg(
+        init_noise_std=1.0,
+        actor_obs_normalization=False,
+        critic_obs_normalization=False,
+        actor_hidden_dims=[32, 32],
+        critic_hidden_dims=[32, 32],
+        activation="elu",
+    )
+    algorithm = RslRlPpoAlgorithmCfg(
+        value_loss_coef=1.0,
+        use_clipped_value_loss=True,
+        clip_param=0.2,
+        entropy_coef=0.005,
+        num_learning_epochs=5,
+        num_mini_batches=4,
+        learning_rate=1.0e-3,
+        schedule="adaptive",
+        gamma=0.99,
+        lam=0.95,
+        desired_kl=0.01,
+        max_grad_norm=1.0,
+    )

source/mindbot/mindbot/tasks/direct/mindbot/agents/sb3_ppo_cfg.yaml

@@ -0,0 +1,20 @@
+# Reference: https://github.com/DLR-RM/rl-baselines3-zoo/blob/master/hyperparams/ppo.yml#L32
+seed: 42
+
+n_timesteps: !!float 1e6
+policy: 'MlpPolicy'
+n_steps: 16
+batch_size: 4096
+gae_lambda: 0.95
+gamma: 0.99
+n_epochs: 20
+ent_coef: 0.01
+learning_rate: !!float 3e-4
+clip_range: !!float 0.2
+policy_kwargs:
+  activation_fn: nn.ELU
+  net_arch: [32, 32]
+  squash_output: False
+vf_coef: 1.0
+max_grad_norm: 1.0
+device: "cuda:0"

source/mindbot/mindbot/tasks/direct/mindbot/agents/skrl_amp_cfg.yaml

@@ -0,0 +1,111 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: True
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: -2.9
+    fixed_log_std: True
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [1024, 512]
+        activations: relu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [1024, 512]
+        activations: relu
+    output: ONE
+  discriminator:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [1024, 512]
+        activations: relu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+# AMP memory (reference motion dataset)
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+motion_dataset:
+  class: RandomMemory
+  memory_size: 200000
+
+# AMP memory (preventing discriminator overfitting)
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+reply_buffer:
+  class: RandomMemory
+  memory_size: 1000000
+
+
+# AMP agent configuration (field names are from AMP_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/agents/amp.html
+agent:
+  class: AMP
+  rollouts: 16
+  learning_epochs: 6
+  mini_batches: 2
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 5.0e-05
+  learning_rate_scheduler: null
+  learning_rate_scheduler_kwargs: null
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  amp_state_preprocessor: RunningStandardScaler
+  amp_state_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 0.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.5
+  discriminator_loss_scale: 5.0
+  amp_batch_size: 512
+  task_reward_weight: 0.0
+  style_reward_weight: 1.0
+  discriminator_batch_size: 4096
+  discriminator_reward_scale: 2.0
+  discriminator_logit_regularization_scale: 0.05
+  discriminator_gradient_penalty_scale: 5.0
+  discriminator_weight_decay_scale: 1.0e-04
+  # rewards_shaper_scale: 1.0
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "humanoid_amp_run"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 80000
+  environment_info: log

source/mindbot/mindbot/tasks/direct/mindbot/agents/skrl_ippo_cfg.yaml

@@ -0,0 +1,80 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: False
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: 0.0
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+
+# IPPO agent configuration (field names are from IPPO_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/multi_agents/ippo.html
+agent:
+  class: IPPO
+  rollouts: 16
+  learning_epochs: 8
+  mini_batches: 1
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 3.0e-04
+  learning_rate_scheduler: KLAdaptiveLR
+  learning_rate_scheduler_kwargs:
+    kl_threshold: 0.008
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 1.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.0
+  kl_threshold: 0.0
+  rewards_shaper_scale: 1.0
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "cart_double_pendulum_direct"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 4800
+  environment_info: log

source/mindbot/mindbot/tasks/direct/mindbot/agents/skrl_mappo_cfg.yaml

@@ -0,0 +1,82 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: True
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: 0.0
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+
+# MAPPO agent configuration (field names are from MAPPO_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/multi_agents/mappo.html
+agent:
+  class: MAPPO
+  rollouts: 16
+  learning_epochs: 8
+  mini_batches: 1
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 3.0e-04
+  learning_rate_scheduler: KLAdaptiveLR
+  learning_rate_scheduler_kwargs:
+    kl_threshold: 0.008
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  shared_state_preprocessor: RunningStandardScaler
+  shared_state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 1.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.0
+  kl_threshold: 0.0
+  rewards_shaper_scale: 1.0
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "cart_double_pendulum_direct"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 4800
+  environment_info: log

source/mindbot/mindbot/tasks/direct/mindbot/agents/skrl_ppo_cfg.yaml

@@ -0,0 +1,80 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: False
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: 0.0
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+
+# PPO agent configuration (field names are from PPO_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/agents/ppo.html
+agent:
+  class: PPO
+  rollouts: 32
+  learning_epochs: 8
+  mini_batches: 8
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 5.0e-04
+  learning_rate_scheduler: KLAdaptiveLR
+  learning_rate_scheduler_kwargs:
+    kl_threshold: 0.008
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 1.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.0
+  kl_threshold: 0.0
+  rewards_shaper_scale: 0.1
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "cartpole_direct"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 4800
+  environment_info: log

source/mindbot/mindbot/tasks/direct/mindbot/mindbot_env.py

@@ -0,0 +1,135 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+from __future__ import annotations
+
+import math
+import torch
+from collections.abc import Sequence
+
+import isaaclab.sim as sim_utils
+from isaaclab.assets import Articulation
+from isaaclab.envs import DirectRLEnv
+from isaaclab.sim.spawners.from_files import GroundPlaneCfg, spawn_ground_plane
+from isaaclab.utils.math import sample_uniform
+
+from .mindbot_env_cfg import MindbotEnvCfg
+
+
+class MindbotEnv(DirectRLEnv):
+    cfg: MindbotEnvCfg
+
+    def __init__(self, cfg: MindbotEnvCfg, render_mode: str | None = None, **kwargs):
+        super().__init__(cfg, render_mode, **kwargs)
+
+        self._cart_dof_idx, _ = self.robot.find_joints(self.cfg.cart_dof_name)
+        self._pole_dof_idx, _ = self.robot.find_joints(self.cfg.pole_dof_name)
+
+        self.joint_pos = self.robot.data.joint_pos
+        self.joint_vel = self.robot.data.joint_vel
+
+    def _setup_scene(self):
+        self.robot = Articulation(self.cfg.robot_cfg)
+        # add ground plane
+        spawn_ground_plane(prim_path="/World/ground", cfg=GroundPlaneCfg())
+        # clone and replicate
+        self.scene.clone_environments(copy_from_source=False)
+        # we need to explicitly filter collisions for CPU simulation
+        if self.device == "cpu":
+            self.scene.filter_collisions(global_prim_paths=[])
+        # add articulation to scene
+        self.scene.articulations["robot"] = self.robot
+        # add lights
+        light_cfg = sim_utils.DomeLightCfg(intensity=2000.0, color=(0.75, 0.75, 0.75))
+        light_cfg.func("/World/Light", light_cfg)
+
+    def _pre_physics_step(self, actions: torch.Tensor) -> None:
+        self.actions = actions.clone()
+
+    def _apply_action(self) -> None:
+        self.robot.set_joint_effort_target(self.actions * self.cfg.action_scale, joint_ids=self._cart_dof_idx)
+
+    def _get_observations(self) -> dict:
+        obs = torch.cat(
+            (
+                self.joint_pos[:, self._pole_dof_idx[0]].unsqueeze(dim=1),
+                self.joint_vel[:, self._pole_dof_idx[0]].unsqueeze(dim=1),
+                self.joint_pos[:, self._cart_dof_idx[0]].unsqueeze(dim=1),
+                self.joint_vel[:, self._cart_dof_idx[0]].unsqueeze(dim=1),
+            ),
+            dim=-1,
+        )
+        observations = {"policy": obs}
+        return observations
+
+    def _get_rewards(self) -> torch.Tensor:
+        total_reward = compute_rewards(
+            self.cfg.rew_scale_alive,
+            self.cfg.rew_scale_terminated,
+            self.cfg.rew_scale_pole_pos,
+            self.cfg.rew_scale_cart_vel,
+            self.cfg.rew_scale_pole_vel,
+            self.joint_pos[:, self._pole_dof_idx[0]],
+            self.joint_vel[:, self._pole_dof_idx[0]],
+            self.joint_pos[:, self._cart_dof_idx[0]],
+            self.joint_vel[:, self._cart_dof_idx[0]],
+            self.reset_terminated,
+        )
+        return total_reward
+
+    def _get_dones(self) -> tuple[torch.Tensor, torch.Tensor]:
+        self.joint_pos = self.robot.data.joint_pos
+        self.joint_vel = self.robot.data.joint_vel
+
+        time_out = self.episode_length_buf >= self.max_episode_length - 1
+        out_of_bounds = torch.any(torch.abs(self.joint_pos[:, self._cart_dof_idx]) > self.cfg.max_cart_pos, dim=1)
+        out_of_bounds = out_of_bounds | torch.any(torch.abs(self.joint_pos[:, self._pole_dof_idx]) > math.pi / 2, dim=1)
+        return out_of_bounds, time_out
+
+    def _reset_idx(self, env_ids: Sequence[int] | None):
+        if env_ids is None:
+            env_ids = self.robot._ALL_INDICES
+        super()._reset_idx(env_ids)
+
+        joint_pos = self.robot.data.default_joint_pos[env_ids]
+        joint_pos[:, self._pole_dof_idx] += sample_uniform(
+            self.cfg.initial_pole_angle_range[0] * math.pi,
+            self.cfg.initial_pole_angle_range[1] * math.pi,
+            joint_pos[:, self._pole_dof_idx].shape,
+            joint_pos.device,
+        )
+        joint_vel = self.robot.data.default_joint_vel[env_ids]
+
+        default_root_state = self.robot.data.default_root_state[env_ids]
+        default_root_state[:, :3] += self.scene.env_origins[env_ids]
+
+        self.joint_pos[env_ids] = joint_pos
+        self.joint_vel[env_ids] = joint_vel
+
+        self.robot.write_root_pose_to_sim(default_root_state[:, :7], env_ids)
+        self.robot.write_root_velocity_to_sim(default_root_state[:, 7:], env_ids)
+        self.robot.write_joint_state_to_sim(joint_pos, joint_vel, None, env_ids)
+
+
+@torch.jit.script
+def compute_rewards(
+    rew_scale_alive: float,
+    rew_scale_terminated: float,
+    rew_scale_pole_pos: float,
+    rew_scale_cart_vel: float,
+    rew_scale_pole_vel: float,
+    pole_pos: torch.Tensor,
+    pole_vel: torch.Tensor,
+    cart_pos: torch.Tensor,
+    cart_vel: torch.Tensor,
+    reset_terminated: torch.Tensor,
+):
+    rew_alive = rew_scale_alive * (1.0 - reset_terminated.float())
+    rew_termination = rew_scale_terminated * reset_terminated.float()
+    rew_pole_pos = rew_scale_pole_pos * torch.sum(torch.square(pole_pos).unsqueeze(dim=1), dim=-1)
+    rew_cart_vel = rew_scale_cart_vel * torch.sum(torch.abs(cart_vel).unsqueeze(dim=1), dim=-1)
+    rew_pole_vel = rew_scale_pole_vel * torch.sum(torch.abs(pole_vel).unsqueeze(dim=1), dim=-1)
+    total_reward = rew_alive + rew_termination + rew_pole_pos + rew_cart_vel + rew_pole_vel
+    return total_reward

source/mindbot/mindbot/tasks/direct/mindbot/mindbot_env_cfg.py

@@ -0,0 +1,48 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+from isaaclab_assets.robots.cartpole import CARTPOLE_CFG
+
+from isaaclab.assets import ArticulationCfg
+from isaaclab.envs import DirectRLEnvCfg
+from isaaclab.scene import InteractiveSceneCfg
+from isaaclab.sim import SimulationCfg
+from isaaclab.utils import configclass
+
+
+@configclass
+class MindbotEnvCfg(DirectRLEnvCfg):
+    # env
+    decimation = 2
+    episode_length_s = 5.0
+    # - spaces definition
+    action_space = 1
+    observation_space = 4
+    state_space = 0
+
+    # simulation
+    sim: SimulationCfg = SimulationCfg(dt=1 / 120, render_interval=decimation)
+
+    # robot(s)
+    robot_cfg: ArticulationCfg = CARTPOLE_CFG.replace(prim_path="/World/envs/env_.*/Robot")
+
+    # scene
+    scene: InteractiveSceneCfg = InteractiveSceneCfg(num_envs=4096, env_spacing=4.0, replicate_physics=True)
+
+    # custom parameters/scales
+    # - controllable joint
+    cart_dof_name = "slider_to_cart"
+    pole_dof_name = "cart_to_pole"
+    # - action scale
+    action_scale = 100.0  # [N]
+    # - reward scales
+    rew_scale_alive = 1.0
+    rew_scale_terminated = -2.0
+    rew_scale_pole_pos = -1.0
+    rew_scale_cart_vel = -0.01
+    rew_scale_pole_vel = -0.005
+    # - reset states/conditions
+    initial_pole_angle_range = [-0.25, 0.25]  # pole angle sample range on reset [rad]
+    max_cart_pos = 3.0  # reset if cart exceeds this position [m]

source/mindbot/mindbot/tasks/direct/mindbot_marl/__init__.py

@@ -0,0 +1,29 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+import gymnasium as gym
+
+from . import agents
+
+##
+# Register Gym environments.
+##
+
+
+gym.register(
+    id="Template-Mindbot-Marl-Direct-v0",
+    entry_point=f"{__name__}.mindbot_marl_env:MindbotMarlEnv",
+    disable_env_checker=True,
+    kwargs={
+        "env_cfg_entry_point": f"{__name__}.mindbot_marl_env_cfg:MindbotMarlEnvCfg",
+        "rl_games_cfg_entry_point": f"{agents.__name__}:rl_games_ppo_cfg.yaml",
+        "rsl_rl_cfg_entry_point": f"{agents.__name__}.rsl_rl_ppo_cfg:PPORunnerCfg",
+        "skrl_amp_cfg_entry_point": f"{agents.__name__}:skrl_amp_cfg.yaml",
+        "skrl_ippo_cfg_entry_point": f"{agents.__name__}:skrl_ippo_cfg.yaml",
+        "skrl_mappo_cfg_entry_point": f"{agents.__name__}:skrl_mappo_cfg.yaml",
+        "skrl_cfg_entry_point": f"{agents.__name__}:skrl_ppo_cfg.yaml",
+        "sb3_cfg_entry_point": f"{agents.__name__}:sb3_ppo_cfg.yaml",
+    },
+)

source/mindbot/mindbot/tasks/direct/mindbot_marl/agents/__init__.py

@@ -0,0 +1,4 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause

source/mindbot/mindbot/tasks/direct/mindbot_marl/agents/rl_games_ppo_cfg.yaml

@@ -0,0 +1,78 @@
+params:
+  seed: 42
+
+  # environment wrapper clipping
+  env:
+    # added to the wrapper
+    clip_observations: 5.0
+    # can make custom wrapper?
+    clip_actions: 1.0
+
+  algo:
+    name: a2c_continuous
+
+  model:
+    name: continuous_a2c_logstd
+
+  # doesn't have this fine grained control but made it close
+  network:
+    name: actor_critic
+    separate: False
+    space:
+      continuous:
+        mu_activation: None
+        sigma_activation: None
+
+        mu_init:
+          name: default
+        sigma_init:
+          name: const_initializer
+          val: 0
+        fixed_sigma: True
+    mlp:
+      units: [32, 32]
+      activation: elu
+      d2rl: False
+
+      initializer:
+        name: default
+      regularizer:
+        name: None
+
+  load_checkpoint: False # flag which sets whether to load the checkpoint
+  load_path: '' # path to the checkpoint to load
+
+  config:
+    name: cartpole_direct
+    env_name: rlgpu
+    device: 'cuda:0'
+    device_name: 'cuda:0'
+    multi_gpu: False
+    ppo: True
+    mixed_precision: False
+    normalize_input: True
+    normalize_value: True
+    num_actors: -1  # configured from the script (based on num_envs)
+    reward_shaper:
+      scale_value: 0.1
+    normalize_advantage: True
+    gamma: 0.99
+    tau : 0.95
+    learning_rate: 5e-4
+    lr_schedule: adaptive
+    kl_threshold: 0.008
+    score_to_win: 20000
+    max_epochs: 150
+    save_best_after: 50
+    save_frequency: 25
+    grad_norm: 1.0
+    entropy_coef: 0.0
+    truncate_grads: True
+    e_clip: 0.2
+    horizon_length: 32
+    minibatch_size: 16384
+    mini_epochs: 8
+    critic_coef: 4
+    clip_value: True
+    seq_length: 4
+    bounds_loss_coef: 0.0001

source/mindbot/mindbot/tasks/direct/mindbot_marl/agents/rsl_rl_ppo_cfg.py

@@ -0,0 +1,38 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+from isaaclab.utils import configclass
+
+from isaaclab_rl.rsl_rl import RslRlOnPolicyRunnerCfg, RslRlPpoActorCriticCfg, RslRlPpoAlgorithmCfg
+
+
+@configclass
+class PPORunnerCfg(RslRlOnPolicyRunnerCfg):
+    num_steps_per_env = 16
+    max_iterations = 150
+    save_interval = 50
+    experiment_name = "cartpole_direct"
+    policy = RslRlPpoActorCriticCfg(
+        init_noise_std=1.0,
+        actor_obs_normalization=False,
+        critic_obs_normalization=False,
+        actor_hidden_dims=[32, 32],
+        critic_hidden_dims=[32, 32],
+        activation="elu",
+    )
+    algorithm = RslRlPpoAlgorithmCfg(
+        value_loss_coef=1.0,
+        use_clipped_value_loss=True,
+        clip_param=0.2,
+        entropy_coef=0.005,
+        num_learning_epochs=5,
+        num_mini_batches=4,
+        learning_rate=1.0e-3,
+        schedule="adaptive",
+        gamma=0.99,
+        lam=0.95,
+        desired_kl=0.01,
+        max_grad_norm=1.0,
+    )

source/mindbot/mindbot/tasks/direct/mindbot_marl/agents/sb3_ppo_cfg.yaml

@@ -0,0 +1,20 @@
+# Reference: https://github.com/DLR-RM/rl-baselines3-zoo/blob/master/hyperparams/ppo.yml#L32
+seed: 42
+
+n_timesteps: !!float 1e6
+policy: 'MlpPolicy'
+n_steps: 16
+batch_size: 4096
+gae_lambda: 0.95
+gamma: 0.99
+n_epochs: 20
+ent_coef: 0.01
+learning_rate: !!float 3e-4
+clip_range: !!float 0.2
+policy_kwargs:
+  activation_fn: nn.ELU
+  net_arch: [32, 32]
+  squash_output: False
+vf_coef: 1.0
+max_grad_norm: 1.0
+device: "cuda:0"

source/mindbot/mindbot/tasks/direct/mindbot_marl/agents/skrl_amp_cfg.yaml

@@ -0,0 +1,111 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: True
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: -2.9
+    fixed_log_std: True
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [1024, 512]
+        activations: relu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [1024, 512]
+        activations: relu
+    output: ONE
+  discriminator:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [1024, 512]
+        activations: relu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+# AMP memory (reference motion dataset)
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+motion_dataset:
+  class: RandomMemory
+  memory_size: 200000
+
+# AMP memory (preventing discriminator overfitting)
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+reply_buffer:
+  class: RandomMemory
+  memory_size: 1000000
+
+
+# AMP agent configuration (field names are from AMP_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/agents/amp.html
+agent:
+  class: AMP
+  rollouts: 16
+  learning_epochs: 6
+  mini_batches: 2
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 5.0e-05
+  learning_rate_scheduler: null
+  learning_rate_scheduler_kwargs: null
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  amp_state_preprocessor: RunningStandardScaler
+  amp_state_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 0.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.5
+  discriminator_loss_scale: 5.0
+  amp_batch_size: 512
+  task_reward_weight: 0.0
+  style_reward_weight: 1.0
+  discriminator_batch_size: 4096
+  discriminator_reward_scale: 2.0
+  discriminator_logit_regularization_scale: 0.05
+  discriminator_gradient_penalty_scale: 5.0
+  discriminator_weight_decay_scale: 1.0e-04
+  # rewards_shaper_scale: 1.0
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "humanoid_amp_run"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 80000
+  environment_info: log

source/mindbot/mindbot/tasks/direct/mindbot_marl/agents/skrl_ippo_cfg.yaml

@@ -0,0 +1,80 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: False
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: 0.0
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+
+# IPPO agent configuration (field names are from IPPO_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/multi_agents/ippo.html
+agent:
+  class: IPPO
+  rollouts: 16
+  learning_epochs: 8
+  mini_batches: 1
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 3.0e-04
+  learning_rate_scheduler: KLAdaptiveLR
+  learning_rate_scheduler_kwargs:
+    kl_threshold: 0.008
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 1.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.0
+  kl_threshold: 0.0
+  rewards_shaper_scale: 1.0
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "cart_double_pendulum_direct"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 4800
+  environment_info: log

source/mindbot/mindbot/tasks/direct/mindbot_marl/agents/skrl_mappo_cfg.yaml

@@ -0,0 +1,82 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: True
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: 0.0
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+
+# MAPPO agent configuration (field names are from MAPPO_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/multi_agents/mappo.html
+agent:
+  class: MAPPO
+  rollouts: 16
+  learning_epochs: 8
+  mini_batches: 1
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 3.0e-04
+  learning_rate_scheduler: KLAdaptiveLR
+  learning_rate_scheduler_kwargs:
+    kl_threshold: 0.008
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  shared_state_preprocessor: RunningStandardScaler
+  shared_state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 1.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.0
+  kl_threshold: 0.0
+  rewards_shaper_scale: 1.0
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "cart_double_pendulum_direct"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 4800
+  environment_info: log

source/mindbot/mindbot/tasks/direct/mindbot_marl/agents/skrl_ppo_cfg.yaml

@@ -0,0 +1,80 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: False
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: 0.0
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+
+# PPO agent configuration (field names are from PPO_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/agents/ppo.html
+agent:
+  class: PPO
+  rollouts: 32
+  learning_epochs: 8
+  mini_batches: 8
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 5.0e-04
+  learning_rate_scheduler: KLAdaptiveLR
+  learning_rate_scheduler_kwargs:
+    kl_threshold: 0.008
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 1.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.0
+  kl_threshold: 0.0
+  rewards_shaper_scale: 0.1
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "cartpole_direct"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 4800
+  environment_info: log

source/mindbot/mindbot/tasks/direct/mindbot_marl/mindbot_marl_env.py

@@ -0,0 +1,184 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+from __future__ import annotations
+
+import math
+import torch
+from collections.abc import Sequence
+
+import isaaclab.sim as sim_utils
+from isaaclab.assets import Articulation
+from isaaclab.envs import DirectMARLEnv
+from isaaclab.sim.spawners.from_files import GroundPlaneCfg, spawn_ground_plane
+from isaaclab.utils.math import sample_uniform
+
+from .mindbot_marl_env_cfg import MindbotMarlEnvCfg
+
+
+class MindbotMarlEnv(DirectMARLEnv):
+    cfg: MindbotMarlEnvCfg
+
+    def __init__(self, cfg: MindbotMarlEnvCfg, render_mode: str | None = None, **kwargs):
+        super().__init__(cfg, render_mode, **kwargs)
+
+        self._cart_dof_idx, _ = self.robot.find_joints(self.cfg.cart_dof_name)
+        self._pole_dof_idx, _ = self.robot.find_joints(self.cfg.pole_dof_name)
+        self._pendulum_dof_idx, _ = self.robot.find_joints(self.cfg.pendulum_dof_name)
+
+        self.joint_pos = self.robot.data.joint_pos
+        self.joint_vel = self.robot.data.joint_vel
+
+    def _setup_scene(self):
+        self.robot = Articulation(self.cfg.robot_cfg)
+        # add ground plane
+        spawn_ground_plane(prim_path="/World/ground", cfg=GroundPlaneCfg())
+        # clone and replicate
+        self.scene.clone_environments(copy_from_source=False)
+        # we need to explicitly filter collisions for CPU simulation
+        if self.device == "cpu":
+            self.scene.filter_collisions(global_prim_paths=[])
+        # add articulation to scene
+        self.scene.articulations["robot"] = self.robot
+        # add lights
+        light_cfg = sim_utils.DomeLightCfg(intensity=2000.0, color=(0.75, 0.75, 0.75))
+        light_cfg.func("/World/Light", light_cfg)
+
+    def _pre_physics_step(self, actions: dict[str, torch.Tensor]) -> None:
+        self.actions = actions
+
+    def _apply_action(self) -> None:
+        self.robot.set_joint_effort_target(
+            self.actions["cart"] * self.cfg.cart_action_scale, joint_ids=self._cart_dof_idx
+        )
+        self.robot.set_joint_effort_target(
+            self.actions["pendulum"] * self.cfg.pendulum_action_scale, joint_ids=self._pendulum_dof_idx
+        )
+
+    def _get_observations(self) -> dict[str, torch.Tensor]:
+        pole_joint_pos = normalize_angle(self.joint_pos[:, self._pole_dof_idx[0]].unsqueeze(dim=1))
+        pendulum_joint_pos = normalize_angle(self.joint_pos[:, self._pendulum_dof_idx[0]].unsqueeze(dim=1))
+        observations = {
+            "cart": torch.cat(
+                (
+                    self.joint_pos[:, self._cart_dof_idx[0]].unsqueeze(dim=1),
+                    self.joint_vel[:, self._cart_dof_idx[0]].unsqueeze(dim=1),
+                    pole_joint_pos,
+                    self.joint_vel[:, self._pole_dof_idx[0]].unsqueeze(dim=1),
+                ),
+                dim=-1,
+            ),
+            "pendulum": torch.cat(
+                (
+                    pole_joint_pos + pendulum_joint_pos,
+                    pendulum_joint_pos,
+                    self.joint_vel[:, self._pendulum_dof_idx[0]].unsqueeze(dim=1),
+                ),
+                dim=-1,
+            ),
+        }
+        return observations
+
+    def _get_rewards(self) -> dict[str, torch.Tensor]:
+        total_reward = compute_rewards(
+            self.cfg.rew_scale_alive,
+            self.cfg.rew_scale_terminated,
+            self.cfg.rew_scale_cart_pos,
+            self.cfg.rew_scale_cart_vel,
+            self.cfg.rew_scale_pole_pos,
+            self.cfg.rew_scale_pole_vel,
+            self.cfg.rew_scale_pendulum_pos,
+            self.cfg.rew_scale_pendulum_vel,
+            self.joint_pos[:, self._cart_dof_idx[0]],
+            self.joint_vel[:, self._cart_dof_idx[0]],
+            normalize_angle(self.joint_pos[:, self._pole_dof_idx[0]]),
+            self.joint_vel[:, self._pole_dof_idx[0]],
+            normalize_angle(self.joint_pos[:, self._pendulum_dof_idx[0]]),
+            self.joint_vel[:, self._pendulum_dof_idx[0]],
+            math.prod(self.terminated_dict.values()),
+        )
+        return total_reward
+
+    def _get_dones(self) -> tuple[dict[str, torch.Tensor], dict[str, torch.Tensor]]:
+        self.joint_pos = self.robot.data.joint_pos
+        self.joint_vel = self.robot.data.joint_vel
+
+        time_out = self.episode_length_buf >= self.max_episode_length - 1
+        out_of_bounds = torch.any(torch.abs(self.joint_pos[:, self._cart_dof_idx]) > self.cfg.max_cart_pos, dim=1)
+        out_of_bounds = out_of_bounds | torch.any(torch.abs(self.joint_pos[:, self._pole_dof_idx]) > math.pi / 2, dim=1)
+
+        terminated = {agent: out_of_bounds for agent in self.cfg.possible_agents}
+        time_outs = {agent: time_out for agent in self.cfg.possible_agents}
+        return terminated, time_outs
+
+    def _reset_idx(self, env_ids: Sequence[int] | None):
+        if env_ids is None:
+            env_ids = self.robot._ALL_INDICES
+        super()._reset_idx(env_ids)
+
+        joint_pos = self.robot.data.default_joint_pos[env_ids]
+        joint_pos[:, self._pole_dof_idx] += sample_uniform(
+            self.cfg.initial_pole_angle_range[0] * math.pi,
+            self.cfg.initial_pole_angle_range[1] * math.pi,
+            joint_pos[:, self._pole_dof_idx].shape,
+            joint_pos.device,
+        )
+        joint_pos[:, self._pendulum_dof_idx] += sample_uniform(
+            self.cfg.initial_pendulum_angle_range[0] * math.pi,
+            self.cfg.initial_pendulum_angle_range[1] * math.pi,
+            joint_pos[:, self._pendulum_dof_idx].shape,
+            joint_pos.device,
+        )
+        joint_vel = self.robot.data.default_joint_vel[env_ids]
+
+        default_root_state = self.robot.data.default_root_state[env_ids]
+        default_root_state[:, :3] += self.scene.env_origins[env_ids]
+
+        self.joint_pos[env_ids] = joint_pos
+        self.joint_vel[env_ids] = joint_vel
+
+        self.robot.write_root_pose_to_sim(default_root_state[:, :7], env_ids)
+        self.robot.write_root_velocity_to_sim(default_root_state[:, 7:], env_ids)
+        self.robot.write_joint_state_to_sim(joint_pos, joint_vel, None, env_ids)
+
+
+@torch.jit.script
+def normalize_angle(angle):
+    return (angle + math.pi) % (2 * math.pi) - math.pi
+
+
+@torch.jit.script
+def compute_rewards(
+    rew_scale_alive: float,
+    rew_scale_terminated: float,
+    rew_scale_cart_pos: float,
+    rew_scale_cart_vel: float,
+    rew_scale_pole_pos: float,
+    rew_scale_pole_vel: float,
+    rew_scale_pendulum_pos: float,
+    rew_scale_pendulum_vel: float,
+    cart_pos: torch.Tensor,
+    cart_vel: torch.Tensor,
+    pole_pos: torch.Tensor,
+    pole_vel: torch.Tensor,
+    pendulum_pos: torch.Tensor,
+    pendulum_vel: torch.Tensor,
+    reset_terminated: torch.Tensor,
+):
+    rew_alive = rew_scale_alive * (1.0 - reset_terminated.float())
+    rew_termination = rew_scale_terminated * reset_terminated.float()
+    rew_pole_pos = rew_scale_pole_pos * torch.sum(torch.square(pole_pos).unsqueeze(dim=1), dim=-1)
+    rew_pendulum_pos = rew_scale_pendulum_pos * torch.sum(
+        torch.square(pole_pos + pendulum_pos).unsqueeze(dim=1), dim=-1
+    )
+    rew_cart_vel = rew_scale_cart_vel * torch.sum(torch.abs(cart_vel).unsqueeze(dim=1), dim=-1)
+    rew_pole_vel = rew_scale_pole_vel * torch.sum(torch.abs(pole_vel).unsqueeze(dim=1), dim=-1)
+    rew_pendulum_vel = rew_scale_pendulum_vel * torch.sum(torch.abs(pendulum_vel).unsqueeze(dim=1), dim=-1)
+
+    total_reward = {
+        "cart": rew_alive + rew_termination + rew_pole_pos + rew_cart_vel + rew_pole_vel,
+        "pendulum": rew_alive + rew_termination + rew_pendulum_pos + rew_pendulum_vel,
+    }
+    return total_reward

source/mindbot/mindbot/tasks/direct/mindbot_marl/mindbot_marl_env_cfg.py

@@ -0,0 +1,55 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+from isaaclab_assets.robots.cart_double_pendulum import CART_DOUBLE_PENDULUM_CFG
+
+from isaaclab.assets import ArticulationCfg
+from isaaclab.envs import DirectMARLEnvCfg
+from isaaclab.scene import InteractiveSceneCfg
+from isaaclab.sim import SimulationCfg
+from isaaclab.utils import configclass
+
+
+@configclass
+class MindbotMarlEnvCfg(DirectMARLEnvCfg):
+    # env
+    decimation = 2
+    episode_length_s = 5.0
+    # multi-agent specification and spaces definition
+    possible_agents = ["cart", "pendulum"]
+    action_spaces = {"cart": 1, "pendulum": 1}
+    observation_spaces = {"cart": 4, "pendulum": 3}
+    state_space = -1
+
+    # simulation
+    sim: SimulationCfg = SimulationCfg(dt=1 / 120, render_interval=decimation)
+
+    # robot(s)
+    robot_cfg: ArticulationCfg = CART_DOUBLE_PENDULUM_CFG.replace(prim_path="/World/envs/env_.*/Robot")
+
+    # scene
+    scene: InteractiveSceneCfg = InteractiveSceneCfg(num_envs=4096, env_spacing=4.0, replicate_physics=True)
+
+    # custom parameters/scales
+    # - controllable joint
+    cart_dof_name = "slider_to_cart"
+    pole_dof_name = "cart_to_pole"
+    pendulum_dof_name = "pole_to_pendulum"
+    # - action scale
+    cart_action_scale = 100.0  # [N]
+    pendulum_action_scale = 50.0  # [Nm]
+    # - reward scales
+    rew_scale_alive = 1.0
+    rew_scale_terminated = -2.0
+    rew_scale_cart_pos = 0
+    rew_scale_cart_vel = -0.01
+    rew_scale_pole_pos = -1.0
+    rew_scale_pole_vel = -0.01
+    rew_scale_pendulum_pos = -1.0
+    rew_scale_pendulum_vel = -0.01
+    # - reset states/conditions
+    initial_pendulum_angle_range = [-0.25, 0.25]  # pendulum angle sample range on reset [rad]
+    initial_pole_angle_range = [-0.25, 0.25]  # pole angle sample range on reset [rad]
+    max_cart_pos = 3.0  # reset if cart exceeds this position [m]

source/mindbot/mindbot/tasks/manager_based/__init__.py

@@ -0,0 +1,6 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+import gymnasium as gym  # noqa: F401

source/mindbot/mindbot/tasks/manager_based/mindbot/__init__.py

@@ -0,0 +1,29 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+import gymnasium as gym
+
+from . import agents
+
+##
+# Register Gym environments.
+##
+
+
+gym.register(
+    id="Template-Mindbot-v0",
+    entry_point="isaaclab.envs:ManagerBasedRLEnv",
+    disable_env_checker=True,
+    kwargs={
+        "env_cfg_entry_point": f"{__name__}.mindbot_env_cfg:MindbotEnvCfg",
+        "rl_games_cfg_entry_point": f"{agents.__name__}:rl_games_ppo_cfg.yaml",
+        "rsl_rl_cfg_entry_point": f"{agents.__name__}.rsl_rl_ppo_cfg:PPORunnerCfg",
+        "skrl_amp_cfg_entry_point": f"{agents.__name__}:skrl_amp_cfg.yaml",
+        "skrl_ippo_cfg_entry_point": f"{agents.__name__}:skrl_ippo_cfg.yaml",
+        "skrl_mappo_cfg_entry_point": f"{agents.__name__}:skrl_mappo_cfg.yaml",
+        "skrl_cfg_entry_point": f"{agents.__name__}:skrl_ppo_cfg.yaml",
+        "sb3_cfg_entry_point": f"{agents.__name__}:sb3_ppo_cfg.yaml",
+    },
+)

source/mindbot/mindbot/tasks/manager_based/mindbot/agents/__init__.py

@@ -0,0 +1,4 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause

source/mindbot/mindbot/tasks/manager_based/mindbot/agents/rl_games_ppo_cfg.yaml

@@ -0,0 +1,78 @@
+params:
+  seed: 42
+
+  # environment wrapper clipping
+  env:
+    # added to the wrapper
+    clip_observations: 5.0
+    # can make custom wrapper?
+    clip_actions: 1.0
+
+  algo:
+    name: a2c_continuous
+
+  model:
+    name: continuous_a2c_logstd
+
+  # doesn't have this fine grained control but made it close
+  network:
+    name: actor_critic
+    separate: False
+    space:
+      continuous:
+        mu_activation: None
+        sigma_activation: None
+
+        mu_init:
+          name: default
+        sigma_init:
+          name: const_initializer
+          val: 0
+        fixed_sigma: True
+    mlp:
+      units: [32, 32]
+      activation: elu
+      d2rl: False
+
+      initializer:
+        name: default
+      regularizer:
+        name: None
+
+  load_checkpoint: False # flag which sets whether to load the checkpoint
+  load_path: '' # path to the checkpoint to load
+
+  config:
+    name: cartpole_direct
+    env_name: rlgpu
+    device: 'cuda:0'
+    device_name: 'cuda:0'
+    multi_gpu: False
+    ppo: True
+    mixed_precision: False
+    normalize_input: True
+    normalize_value: True
+    num_actors: -1  # configured from the script (based on num_envs)
+    reward_shaper:
+      scale_value: 0.1
+    normalize_advantage: True
+    gamma: 0.99
+    tau : 0.95
+    learning_rate: 5e-4
+    lr_schedule: adaptive
+    kl_threshold: 0.008
+    score_to_win: 20000
+    max_epochs: 150
+    save_best_after: 50
+    save_frequency: 25
+    grad_norm: 1.0
+    entropy_coef: 0.0
+    truncate_grads: True
+    e_clip: 0.2
+    horizon_length: 32
+    minibatch_size: 16384
+    mini_epochs: 8
+    critic_coef: 4
+    clip_value: True
+    seq_length: 4
+    bounds_loss_coef: 0.0001

source/mindbot/mindbot/tasks/manager_based/mindbot/agents/rsl_rl_ppo_cfg.py

@@ -0,0 +1,38 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+from isaaclab.utils import configclass
+
+from isaaclab_rl.rsl_rl import RslRlOnPolicyRunnerCfg, RslRlPpoActorCriticCfg, RslRlPpoAlgorithmCfg
+
+
+@configclass
+class PPORunnerCfg(RslRlOnPolicyRunnerCfg):
+    num_steps_per_env = 16
+    max_iterations = 150
+    save_interval = 50
+    experiment_name = "cartpole_direct"
+    policy = RslRlPpoActorCriticCfg(
+        init_noise_std=1.0,
+        actor_obs_normalization=False,
+        critic_obs_normalization=False,
+        actor_hidden_dims=[32, 32],
+        critic_hidden_dims=[32, 32],
+        activation="elu",
+    )
+    algorithm = RslRlPpoAlgorithmCfg(
+        value_loss_coef=1.0,
+        use_clipped_value_loss=True,
+        clip_param=0.2,
+        entropy_coef=0.005,
+        num_learning_epochs=5,
+        num_mini_batches=4,
+        learning_rate=1.0e-3,
+        schedule="adaptive",
+        gamma=0.99,
+        lam=0.95,
+        desired_kl=0.01,
+        max_grad_norm=1.0,
+    )

source/mindbot/mindbot/tasks/manager_based/mindbot/agents/sb3_ppo_cfg.yaml

@@ -0,0 +1,20 @@
+# Reference: https://github.com/DLR-RM/rl-baselines3-zoo/blob/master/hyperparams/ppo.yml#L32
+seed: 42
+
+n_timesteps: !!float 1e6
+policy: 'MlpPolicy'
+n_steps: 16
+batch_size: 4096
+gae_lambda: 0.95
+gamma: 0.99
+n_epochs: 20
+ent_coef: 0.01
+learning_rate: !!float 3e-4
+clip_range: !!float 0.2
+policy_kwargs:
+  activation_fn: nn.ELU
+  net_arch: [32, 32]
+  squash_output: False
+vf_coef: 1.0
+max_grad_norm: 1.0
+device: "cuda:0"

source/mindbot/mindbot/tasks/manager_based/mindbot/agents/skrl_amp_cfg.yaml

@@ -0,0 +1,111 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: True
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: -2.9
+    fixed_log_std: True
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [1024, 512]
+        activations: relu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [1024, 512]
+        activations: relu
+    output: ONE
+  discriminator:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [1024, 512]
+        activations: relu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+# AMP memory (reference motion dataset)
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+motion_dataset:
+  class: RandomMemory
+  memory_size: 200000
+
+# AMP memory (preventing discriminator overfitting)
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+reply_buffer:
+  class: RandomMemory
+  memory_size: 1000000
+
+
+# AMP agent configuration (field names are from AMP_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/agents/amp.html
+agent:
+  class: AMP
+  rollouts: 16
+  learning_epochs: 6
+  mini_batches: 2
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 5.0e-05
+  learning_rate_scheduler: null
+  learning_rate_scheduler_kwargs: null
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  amp_state_preprocessor: RunningStandardScaler
+  amp_state_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 0.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.5
+  discriminator_loss_scale: 5.0
+  amp_batch_size: 512
+  task_reward_weight: 0.0
+  style_reward_weight: 1.0
+  discriminator_batch_size: 4096
+  discriminator_reward_scale: 2.0
+  discriminator_logit_regularization_scale: 0.05
+  discriminator_gradient_penalty_scale: 5.0
+  discriminator_weight_decay_scale: 1.0e-04
+  # rewards_shaper_scale: 1.0
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "humanoid_amp_run"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 80000
+  environment_info: log

source/mindbot/mindbot/tasks/manager_based/mindbot/agents/skrl_ippo_cfg.yaml

@@ -0,0 +1,80 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: False
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: 0.0
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+
+# IPPO agent configuration (field names are from IPPO_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/multi_agents/ippo.html
+agent:
+  class: IPPO
+  rollouts: 16
+  learning_epochs: 8
+  mini_batches: 1
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 3.0e-04
+  learning_rate_scheduler: KLAdaptiveLR
+  learning_rate_scheduler_kwargs:
+    kl_threshold: 0.008
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 1.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.0
+  kl_threshold: 0.0
+  rewards_shaper_scale: 1.0
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "cart_double_pendulum_direct"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 4800
+  environment_info: log

source/mindbot/mindbot/tasks/manager_based/mindbot/agents/skrl_mappo_cfg.yaml

@@ -0,0 +1,82 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: True
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: 0.0
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+
+# MAPPO agent configuration (field names are from MAPPO_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/multi_agents/mappo.html
+agent:
+  class: MAPPO
+  rollouts: 16
+  learning_epochs: 8
+  mini_batches: 1
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 3.0e-04
+  learning_rate_scheduler: KLAdaptiveLR
+  learning_rate_scheduler_kwargs:
+    kl_threshold: 0.008
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  shared_state_preprocessor: RunningStandardScaler
+  shared_state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 1.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.0
+  kl_threshold: 0.0
+  rewards_shaper_scale: 1.0
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "cart_double_pendulum_direct"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 4800
+  environment_info: log

source/mindbot/mindbot/tasks/manager_based/mindbot/agents/skrl_ppo_cfg.yaml

@@ -0,0 +1,80 @@
+seed: 42
+
+
+# Models are instantiated using skrl's model instantiator utility
+# https://skrl.readthedocs.io/en/latest/api/utils/model_instantiators.html
+models:
+  separate: False
+  policy:  # see gaussian_model parameters
+    class: GaussianMixin
+    clip_actions: False
+    clip_log_std: True
+    min_log_std: -20.0
+    max_log_std: 2.0
+    initial_log_std: 0.0
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ACTIONS
+  value:  # see deterministic_model parameters
+    class: DeterministicMixin
+    clip_actions: False
+    network:
+      - name: net
+        input: OBSERVATIONS
+        layers: [32, 32]
+        activations: elu
+    output: ONE
+
+
+# Rollout memory
+# https://skrl.readthedocs.io/en/latest/api/memories/random.html
+memory:
+  class: RandomMemory
+  memory_size: -1  # automatically determined (same as agent:rollouts)
+
+
+# PPO agent configuration (field names are from PPO_DEFAULT_CONFIG)
+# https://skrl.readthedocs.io/en/latest/api/agents/ppo.html
+agent:
+  class: PPO
+  rollouts: 32
+  learning_epochs: 8
+  mini_batches: 8
+  discount_factor: 0.99
+  lambda: 0.95
+  learning_rate: 5.0e-04
+  learning_rate_scheduler: KLAdaptiveLR
+  learning_rate_scheduler_kwargs:
+    kl_threshold: 0.008
+  state_preprocessor: RunningStandardScaler
+  state_preprocessor_kwargs: null
+  value_preprocessor: RunningStandardScaler
+  value_preprocessor_kwargs: null
+  random_timesteps: 0
+  learning_starts: 0
+  grad_norm_clip: 1.0
+  ratio_clip: 0.2
+  value_clip: 0.2
+  clip_predicted_values: True
+  entropy_loss_scale: 0.0
+  value_loss_scale: 2.0
+  kl_threshold: 0.0
+  rewards_shaper_scale: 0.1
+  time_limit_bootstrap: False
+  # logging and checkpoint
+  experiment:
+    directory: "cartpole_direct"
+    experiment_name: ""
+    write_interval: auto
+    checkpoint_interval: auto
+
+
+# Sequential trainer
+# https://skrl.readthedocs.io/en/latest/api/trainers/sequential.html
+trainer:
+  class: SequentialTrainer
+  timesteps: 4800
+  environment_info: log

source/mindbot/mindbot/tasks/manager_based/mindbot/mdp/__init__.py

@@ -0,0 +1,10 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+"""This sub-module contains the functions that are specific to the environment."""
+
+from isaaclab.envs.mdp import *  # noqa: F401, F403
+
+from .rewards import *  # noqa: F401, F403

source/mindbot/mindbot/tasks/manager_based/mindbot/mdp/rewards.py

@@ -0,0 +1,26 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+from __future__ import annotations
+
+import torch
+from typing import TYPE_CHECKING
+
+from isaaclab.assets import Articulation
+from isaaclab.managers import SceneEntityCfg
+from isaaclab.utils.math import wrap_to_pi
+
+if TYPE_CHECKING:
+    from isaaclab.envs import ManagerBasedRLEnv
+
+
+def joint_pos_target_l2(env: ManagerBasedRLEnv, target: float, asset_cfg: SceneEntityCfg) -> torch.Tensor:
+    """Penalize joint position deviation from a target value."""
+    # extract the used quantities (to enable type-hinting)
+    asset: Articulation = env.scene[asset_cfg.name]
+    # wrap the joint positions to (-pi, pi)
+    joint_pos = wrap_to_pi(asset.data.joint_pos[:, asset_cfg.joint_ids])
+    # compute the reward
+    return torch.sum(torch.square(joint_pos - target), dim=1)

source/mindbot/mindbot/tasks/manager_based/mindbot/mindbot_env_cfg.py

@@ -0,0 +1,180 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+import math
+
+import isaaclab.sim as sim_utils
+from isaaclab.assets import ArticulationCfg, AssetBaseCfg
+from isaaclab.envs import ManagerBasedRLEnvCfg
+from isaaclab.managers import EventTermCfg as EventTerm
+from isaaclab.managers import ObservationGroupCfg as ObsGroup
+from isaaclab.managers import ObservationTermCfg as ObsTerm
+from isaaclab.managers import RewardTermCfg as RewTerm
+from isaaclab.managers import SceneEntityCfg
+from isaaclab.managers import TerminationTermCfg as DoneTerm
+from isaaclab.scene import InteractiveSceneCfg
+from isaaclab.utils import configclass
+
+from . import mdp
+
+##
+# Pre-defined configs
+##
+
+from isaaclab_assets.robots.cartpole import CARTPOLE_CFG  # isort:skip
+
+
+##
+# Scene definition
+##
+
+
+@configclass
+class MindbotSceneCfg(InteractiveSceneCfg):
+    """Configuration for a cart-pole scene."""
+
+    # ground plane
+    ground = AssetBaseCfg(
+        prim_path="/World/ground",
+        spawn=sim_utils.GroundPlaneCfg(size=(100.0, 100.0)),
+    )
+
+    # robot
+    robot: ArticulationCfg = CARTPOLE_CFG.replace(prim_path="{ENV_REGEX_NS}/Robot")
+
+    # lights
+    dome_light = AssetBaseCfg(
+        prim_path="/World/DomeLight",
+        spawn=sim_utils.DomeLightCfg(color=(0.9, 0.9, 0.9), intensity=500.0),
+    )
+
+
+##
+# MDP settings
+##
+
+
+@configclass
+class ActionsCfg:
+    """Action specifications for the MDP."""
+
+    joint_effort = mdp.JointEffortActionCfg(asset_name="robot", joint_names=["slider_to_cart"], scale=100.0)
+
+
+@configclass
+class ObservationsCfg:
+    """Observation specifications for the MDP."""
+
+    @configclass
+    class PolicyCfg(ObsGroup):
+        """Observations for policy group."""
+
+        # observation terms (order preserved)
+        joint_pos_rel = ObsTerm(func=mdp.joint_pos_rel)
+        joint_vel_rel = ObsTerm(func=mdp.joint_vel_rel)
+
+        def __post_init__(self) -> None:
+            self.enable_corruption = False
+            self.concatenate_terms = True
+
+    # observation groups
+    policy: PolicyCfg = PolicyCfg()
+
+
+@configclass
+class EventCfg:
+    """Configuration for events."""
+
+    # reset
+    reset_cart_position = EventTerm(
+        func=mdp.reset_joints_by_offset,
+        mode="reset",
+        params={
+            "asset_cfg": SceneEntityCfg("robot", joint_names=["slider_to_cart"]),
+            "position_range": (-1.0, 1.0),
+            "velocity_range": (-0.5, 0.5),
+        },
+    )
+
+    reset_pole_position = EventTerm(
+        func=mdp.reset_joints_by_offset,
+        mode="reset",
+        params={
+            "asset_cfg": SceneEntityCfg("robot", joint_names=["cart_to_pole"]),
+            "position_range": (-0.25 * math.pi, 0.25 * math.pi),
+            "velocity_range": (-0.25 * math.pi, 0.25 * math.pi),
+        },
+    )
+
+
+@configclass
+class RewardsCfg:
+    """Reward terms for the MDP."""
+
+    # (1) Constant running reward
+    alive = RewTerm(func=mdp.is_alive, weight=1.0)
+    # (2) Failure penalty
+    terminating = RewTerm(func=mdp.is_terminated, weight=-2.0)
+    # (3) Primary task: keep pole upright
+    pole_pos = RewTerm(
+        func=mdp.joint_pos_target_l2,
+        weight=-1.0,
+        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["cart_to_pole"]), "target": 0.0},
+    )
+    # (4) Shaping tasks: lower cart velocity
+    cart_vel = RewTerm(
+        func=mdp.joint_vel_l1,
+        weight=-0.01,
+        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["slider_to_cart"])},
+    )
+    # (5) Shaping tasks: lower pole angular velocity
+    pole_vel = RewTerm(
+        func=mdp.joint_vel_l1,
+        weight=-0.005,
+        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["cart_to_pole"])},
+    )
+
+
+@configclass
+class TerminationsCfg:
+    """Termination terms for the MDP."""
+
+    # (1) Time out
+    time_out = DoneTerm(func=mdp.time_out, time_out=True)
+    # (2) Cart out of bounds
+    cart_out_of_bounds = DoneTerm(
+        func=mdp.joint_pos_out_of_manual_limit,
+        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["slider_to_cart"]), "bounds": (-3.0, 3.0)},
+    )
+
+
+##
+# Environment configuration
+##
+
+
+@configclass
+class MindbotEnvCfg(ManagerBasedRLEnvCfg):
+    # Scene settings
+    scene: MindbotSceneCfg = MindbotSceneCfg(num_envs=4096, env_spacing=4.0)
+    # Basic settings
+    observations: ObservationsCfg = ObservationsCfg()
+    actions: ActionsCfg = ActionsCfg()
+    events: EventCfg = EventCfg()
+    # MDP settings
+    rewards: RewardsCfg = RewardsCfg()
+    terminations: TerminationsCfg = TerminationsCfg()
+
+    # Post initialization
+    def __post_init__(self) -> None:
+        """Post initialization."""
+        # general settings
+        self.decimation = 2
+        self.episode_length_s = 5
+        # viewer settings
+        self.viewer.eye = (8.0, 0.0, 5.0)
+        # simulation settings
+        self.sim.dt = 1 / 120
+        self.sim.render_interval = self.decimation

source/mindbot/mindbot/ui_extension_example.py

@@ -0,0 +1,46 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+import omni.ext
+
+
+# Functions and vars are available to other extension as usual in python: `example.python_ext.some_public_function(x)`
+def some_public_function(x: int):
+    print("[mindbot] some_public_function was called with x: ", x)
+    return x**x
+
+
+# Any class derived from `omni.ext.IExt` in top level module (defined in `python.modules` of `extension.toml`) will be
+# instantiated when extension gets enabled and `on_startup(ext_id)` will be called. Later when extension gets disabled
+# on_shutdown() is called.
+class ExampleExtension(omni.ext.IExt):
+    # ext_id is current extension id. It can be used with extension manager to query additional information, like where
+    # this extension is located on filesystem.
+    def on_startup(self, ext_id):
+        print("[mindbot] startup")
+
+        self._count = 0
+
+        self._window = omni.ui.Window("My Window", width=300, height=300)
+        with self._window.frame:
+            with omni.ui.VStack():
+                label = omni.ui.Label("")
+
+                def on_click():
+                    self._count += 1
+                    label.text = f"count: {self._count}"
+
+                def on_reset():
+                    self._count = 0
+                    label.text = "empty"
+
+                on_reset()
+
+                with omni.ui.HStack():
+                    omni.ui.Button("Add", clicked_fn=on_click)
+                    omni.ui.Button("Reset", clicked_fn=on_reset)
+
+    def on_shutdown(self):
+        print("[mindbot] shutdown")

source/mindbot/pyproject.toml

@@ -0,0 +1,3 @@
+[build-system]
+requires = ["setuptools", "wheel", "toml"]
+build-backend = "setuptools.build_meta"

source/mindbot/setup.py

@@ -0,0 +1,47 @@
+# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+"""Installation script for the 'mindbot' python package."""
+
+import os
+import toml
+
+from setuptools import setup
+
+# Obtain the extension data from the extension.toml file
+EXTENSION_PATH = os.path.dirname(os.path.realpath(__file__))
+# Read the extension.toml file
+EXTENSION_TOML_DATA = toml.load(os.path.join(EXTENSION_PATH, "config", "extension.toml"))
+
+# Minimum dependencies required prior to installation
+INSTALL_REQUIRES = [
+    # NOTE: Add dependencies
+    "psutil",
+]
+
+# Installation operation
+setup(
+    name="mindbot",
+    packages=["mindbot"],
+    author=EXTENSION_TOML_DATA["package"]["author"],
+    maintainer=EXTENSION_TOML_DATA["package"]["maintainer"],
+    url=EXTENSION_TOML_DATA["package"]["repository"],
+    version=EXTENSION_TOML_DATA["package"]["version"],
+    description=EXTENSION_TOML_DATA["package"]["description"],
+    keywords=EXTENSION_TOML_DATA["package"]["keywords"],
+    install_requires=INSTALL_REQUIRES,
+    license="Apache-2.0",
+    include_package_data=True,
+    python_requires=">=3.10",
+    classifiers=[
+        "Natural Language :: English",
+        "Programming Language :: Python :: 3.10",
+        "Programming Language :: Python :: 3.11",
+        "Isaac Sim :: 4.5.0",
+        "Isaac Sim :: 5.0.0",
+        "Isaac Sim :: 5.1.0",
+    ],
+    zip_safe=False,
+)