chore: remove obsolete dead code and stale references

.gitignore

@@ -83,9 +83,7 @@ src/unilab/assets/bots/
 src/unilab/assets/robots/x2/meshes/*.STL
 src/unilab/assets/robots/g1/assets/
 src/unilab/assets/robots/g1/textures/
-src/unilab/assets/robots/go1/assets/
 src/unilab/assets/robots/go2/assets/
-src/unilab/assets/robots/a2/assets/
 src/unilab/assets/robots/allegro_hand/assets/
 src/unilab/assets/robots/sharpa_wave/meshes/
 

docs/sphinx/source/api_reference/index.md

@@ -56,7 +56,7 @@ PPO / APPO / SAC / TD3 variants moved to the uni_rl package (issue #1480).
 :::{grid-item-card} 🏋 `unilab.training`
 :link: training/index
 :link-type: doc
-Runtime helpers, monitoring, reward bookkeeping, runner orchestration.
+Runtime helpers, reward bookkeeping, runner orchestration.
 :::
 
 :::{grid-item-card} 🔗 Shared-memory runtime → `uni_rl.ipc`

docs/sphinx/source/api_reference/training/index.md

@@ -2,7 +2,7 @@
 
 Glue between `algos`, `envs` and `ipc`: experiment lifecycle and the
 top-level `run` helpers invoked by the `train` / `eval` / `demo` CLI
-entrypoints. Layer-0 helpers (seeding, monitoring, reward bookkeeping,
+entrypoints. Layer-0 helpers (seeding, reward bookkeeping,
 checkpoint resolution, sim2sim contracts) live in `unilab.utils`; resolved
 env config adaptation lives in `unilab.base.config_adapter`.
 

docs/sphinx/source/api_reference/utils/index.md

@@ -1,7 +1,7 @@
 # `unilab.utils` — Utilities
 
-Device probing, tensor helpers, training seeding, hardware monitoring,
-reward bookkeeping, checkpoint resolution, sim2sim contract checks, NaN
+Device probing, tensor helpers, training seeding, reward bookkeeping,
+checkpoint resolution, sim2sim contract checks, NaN
 guards, and pure-numpy geometry/rotation helpers shared across envs and
 scripts.
 

docs/sphinx/source/en/3-deployment/3-framework_migration/2-from_legged_gym.md

@@ -36,7 +36,7 @@ mostly mechanical.
 3. Mirror your reward terms; keep the same names so reward parity is
    diff-able.
 4. Translate command sampling — configure `UniformVelocityCommandCfg` under
-   the owner YAML's `env.commands` (see the Go1 flat owner).
+   the owner YAML's `env.commands` (see the `go2_joystick_flat` owner).
 5. Translate terrain — Legged Gym's heightfield generator has a UniLab
    counterpart at `unilab.terrains.heightfield_terrains`.
 

docs/sphinx/source/en/4-developer_guide/3-extending/4-new_terrain.md

@@ -25,8 +25,6 @@ materialization out of `step()`, `reset()`, and hot domain-randomization loops.
 - Terrain generator shape and numerical behavior:
   `tests/terrains/test_terrain_generator.py`
 - Rough locomotion height-scan and spawn behavior:
-  `tests/envs/locomotion/test_go2_rough_height_scan.py`,
-  `tests/envs/locomotion/test_go2_terrain_spawn.py`,
   `tests/envs/locomotion/test_terrain_spawn.py`
 - Backend materialization boundaries: `tests/utils/test_xml_utils.py`
 

docs/sphinx/source/zh_CN/3-deployment/3-framework_migration/2-from_legged_gym.md

@@ -32,7 +32,7 @@ Legged Gym 曾是那套 GPU 常驻的 PPO 模板，教会了整个领域如何
 2. 在 `src/unilab/tasks/locomotion/<robot>/` 下创建一个任务模块。
 3. 镜像你的 reward 项；保持名称相同，以便 reward 一致性可被 diff。
 4. 翻译命令采样 —— 在 owner YAML 的 `env.commands` 下配置
-   `UniformVelocityCommandCfg`，参考 Go1 flat owner。
+   `UniformVelocityCommandCfg`，参考 `go2_joystick_flat` owner。
 5. 翻译地形 —— Legged Gym 的高度场生成器在 UniLab 中有一个对应物，位于
    `unilab.terrains.heightfield_terrains`。
 

docs/sphinx/source/zh_CN/4-developer_guide/3-extending/4-new_terrain.md

@@ -25,8 +25,6 @@
 - 地形生成器形状与数值行为：
   `tests/terrains/test_terrain_generator.py`
 - Rough locomotion 高度扫描与 spawn 行为：
-  `tests/envs/locomotion/test_go2_rough_height_scan.py`、
-  `tests/envs/locomotion/test_go2_terrain_spawn.py`、
   `tests/envs/locomotion/test_terrain_spawn.py`
 - 后端 materialization 边界：`tests/utils/test_xml_utils.py`
 

pyproject.rocm.toml

@@ -163,7 +163,6 @@ venv = ".venv"
 include = ["src/unilab"]
 exclude = [
     "src/unilab/envs/",                     # lazy optional `mujoco` import; not installed in the pyright env
-    "src/unilab/utils/monitoring.py",       # optional pynvml/psutil deps
     "src/unilab/visualization/",            # direct mujoco C bindings + optional viser deps
 ]
 reportMissingImports = "warning"

pyproject.toml

@@ -250,7 +250,6 @@ venv = ".venv"
 include = ["src/unilab"]
 exclude = [
     "src/unilab/envs/",                     # lazy optional `mujoco` import; not installed in the pyright env
-    "src/unilab/utils/monitoring.py",       # optional pynvml/psutil deps
     "src/unilab/visualization/",            # direct mujoco C bindings + optional viser deps
 ]
 reportMissingImports = "warning"