diff --git a/docs/cli/list-info.en.md b/docs/cli/list-info.en.md
index ae97e35c..0a60668b 100644
--- a/docs/cli/list-info.en.md
+++ b/docs/cli/list-info.en.md
@@ -11,20 +11,29 @@ genelab list envs
 genelab list tasks
 ```
 
-If an extension is not installed, import it explicitly:
+If an extension is not installed as an entry point, import it explicitly:
 
 ```bash
 PYTHONPATH=examples/inverted_pendulum/src \
   genelab --import genelab_inverted_pendulum.tasks list tasks
 ```
 
-## Inspect a task
+To see only the bundled asset zoo (skip user-installed entry-point extensions):
 
 ```bash
-genelab info GeneLab-Inverted-Pendulum-v0
+genelab --no-entry-points list robots
 ```
 
-The task view includes:
+## Inspect any registered name
+
+```bash
+genelab info GeneLab-Inverted-Pendulum-v0   # task
+genelab info CartPole-Env-v0                # env
+genelab info anymal_c                       # robot
+```
+
+`info` does not require you to know which registry the name belongs to. Task views additionally
+include:
 
 | Section | Meaning |
 |---|---|
@@ -32,6 +41,9 @@ The task view includes:
 | Task config | `TaskCfg` summary: env, robot, trainable flag, agent. |
 | Overridable paths | Dotted paths accepted by `play` and `train`. |
 
+If the inspected task needs an asset that isn't cached yet, `info` fetches it inline behind the
+same progress bar used by `play` / `train` / `asset download`.
+
 ## Use copied override paths
 
 ```bash
@@ -39,9 +51,11 @@ genelab play GeneLab-Inverted-Pendulum-v0 \
   --env.rewards_cfg.pole_upright.weight 4.0
 ```
 
-When `play_env` exists, play-mode shortcut flags target `play_env`; explicit paths remain explicit.
+When `play_env` exists, play-mode shortcut flags (`--vis` / `--gpu` / `--dt` / `--steps`) target
+`play_env`; explicit dotted paths still write the exact subtree you spelled.
 
 ## See also
 
 - [Configuration Reference](../reference/configuration.md)
 - [CLI Reference](../reference/cli.md)
+- [Asset zoo](../concepts/asset_zoo.md)
diff --git a/docs/cli/list-info.zh.md b/docs/cli/list-info.zh.md
index b5bb671b..82365f24 100644
--- a/docs/cli/list-info.zh.md
+++ b/docs/cli/list-info.zh.md
@@ -10,20 +10,28 @@ genelab list envs
 genelab list tasks
 ```
 
-如果扩展尚未安装，显式导入：
+如果扩展尚未通过 entry point 安装，显式导入：
 
 ```bash
 PYTHONPATH=examples/inverted_pendulum/src \
   genelab --import genelab_inverted_pendulum.tasks list tasks
 ```
 
-## 查看任务
+要只看随包 asset zoo（不加载用户已安装的扩展），加 `--no-entry-points`：
 
 ```bash
-genelab info GeneLab-Inverted-Pendulum-v0
+genelab --no-entry-points list robots
 ```
 
-任务视图包括：
+## 查看任意已注册名字
+
+```bash
+genelab info GeneLab-Inverted-Pendulum-v0   # task
+genelab info CartPole-Env-v0                # env
+genelab info anymal_c                       # robot
+```
+
+`info` 不要求事先知道是哪一类。Task 视图额外包含：
 
 | 区域 | 含义 |
 |---|---|
@@ -31,6 +39,9 @@ genelab info GeneLab-Inverted-Pendulum-v0
 | 任务配置 | `TaskCfg` 摘要：env、robot、trainable、agent。 |
 | 可覆盖路径 | `play` 和 `train` 接受的 dotted path。 |
 
+若该 task 需要的 asset 尚未缓存，`info` 会即时下载并显示进度条
+（和 `play` / `train` / `asset download` 走的是同一条 progress UI）。
+
 ## 使用复制出的 override 路径
 
 ```bash
@@ -38,9 +49,11 @@ genelab play GeneLab-Inverted-Pendulum-v0 \
   --env.rewards_cfg.pole_upright.weight 4.0
 ```
 
-当 task 有 `play_env` 时，play 模式短标志会作用到 `play_env`；显式路径仍保持显式语义。
+当 task 提供 `play_env` 时，play 模式下的简写 flag（`--vis` / `--gpu` / `--dt` /
+`--steps`）会作用到 `play_env`；显式的 dotted 路径仍按字面写的子树生效。
 
 ## 另见
 
 - [配置参考](../reference/configuration.md)
 - [CLI 参考](../reference/cli.md)
+- [Asset zoo](../concepts/asset_zoo.md)
diff --git a/docs/cli/overview.en.md b/docs/cli/overview.en.md
index afe339be..31d2fc6c 100644
--- a/docs/cli/overview.en.md
+++ b/docs/cli/overview.en.md
@@ -12,10 +12,21 @@ genelab [global options] <command> [arguments]
 
 | Area | Commands |
 |---|---|
-| Registry discovery | `list robots`, `list envs`, `list tasks`, `info NAME` |
-| Runtime | `play TASK`, `train TASK` |
+| Registry | `list robots\|envs\|tasks`, `info NAME`, `asset list\|info\|download\|purge` |
+| Runtime | `play TASK`, `train TASK`, `eval TASK CHECKPOINT`, `export TASK CHECKPOINT`, `benchmark --suite ...` |
+| Project | `project new NAME` |
 | Utilities | `cache`, `prof open` |
-| Project scaffolding | `project new NAME` |
+
+Running `genelab` with no subcommand prints a landing page with quickstart commands and a count of
+registered robots / envs / tasks.
+
+## Global options
+
+| Option | Meaning |
+|---|---|
+| `--version` | Print the version and exit. |
+| `--import MODULE` | Import an extension module before dispatch; repeatable. |
+| `--no-entry-points` | Skip extensions from the `genelab.extensions` entry-point group. |
 
 ## Extension loading order
 
@@ -30,7 +41,7 @@ Use entry points for daily work and `--import` for local experiments.
 ## Overrides
 
 Runtime commands accept unknown `--a.b.c VALUE` options after the task id. The CLI forwards them to
-`apply_overrides`.
+`apply_overrides`:
 
 ```bash
 genelab play TASK_ID --env.simulation.dt 0.005
@@ -48,3 +59,4 @@ When stdin is a TTY, the CLI can prompt for a missing task id, unknown registry
 - [CLI Reference](../reference/cli.md)
 - [Discovery: list and info](list-info.md)
 - [Play and Train](play-train.md)
+- [Eval and export](../concepts/eval-and-export.md)
diff --git a/docs/cli/overview.zh.md b/docs/cli/overview.zh.md
index 6916f131..d6ad9aa3 100644
--- a/docs/cli/overview.zh.md
+++ b/docs/cli/overview.zh.md
@@ -1,6 +1,7 @@
 # CLI 总览
 
-`genelab` CLI 是注册表和任务配置上的薄调度层。它不拥有任务逻辑；它负责发现扩展、解析已注册对象、应用 override，然后调用 task 或 runner。
+`genelab` CLI 是注册表（registry）与任务配置之上的薄分发层。它不持有 task 逻辑——
+发现扩展、解析已注册对象、应用 override，然后调用 task 或 runner。
 
 ## 命令模型
 
@@ -8,40 +9,53 @@
 genelab [全局选项] <命令> [参数]
 ```
 
-| 区域 | 命令 |
+| 分组 | 命令 |
 |---|---|
-| 注册表发现 | `list robots`、`list envs`、`list tasks`、`info NAME` |
-| 运行时 | `play TASK`、`train TASK` |
-| 工具 | `cache`、`prof open` |
-| 项目骨架 | `project new NAME` |
+| Registry | `list robots\|envs\|tasks`、`info NAME`、`asset list\|info\|download\|purge` |
+| Runtime | `play TASK`、`train TASK`、`eval TASK CHECKPOINT`、`export TASK CHECKPOINT`、`benchmark --suite ...` |
+| Project | `project new NAME` |
+| Utilities | `cache`、`prof open` |
+
+不带子命令运行 `genelab` 会打印 landing 页：包含 quickstart 命令以及已注册的
+robots / envs / tasks 数量。
+
+## 全局选项
+
+| 选项 | 含义 |
+|---|---|
+| `--version` | 打印版本并退出。 |
+| `--import MODULE` | 分发前导入扩展模块；可重复。 |
+| `--no-entry-points` | 跳过 `genelab.extensions` entry point 组的扩展。 |
 
 ## 扩展加载顺序
 
-任何需要注册表数据的命令都会按顺序加载：
+每个需要 registry 数据的命令按以下顺序加载扩展：
 
-1. 通过 `load_bundled_asset_zoo()` 加载随包 asset zoo robot。
-2. 已安装的 `genelab.extensions` entry point，除非设置 `--no-entry-points`。
-3. 重复传入的显式 `--import MODULE`。
+1. 通过 `load_bundled_asset_zoo()` 装入随包的 asset zoo 机器人。
+2. 已安装的 `genelab.extensions` entry points（除非传 `--no-entry-points`）。
+3. 多次显式的 `--import MODULE`。
 
-日常工作用 entry point，本地实验用 `--import`。
+日常工作用 entry points，临时实验用 `--import`。
 
 ## Overrides
 
-运行时命令在 task id 后接受未知的 `--a.b.c VALUE` 选项。CLI 会把它们转发给
-`apply_overrides`。
+Runtime 命令在 task id 后接受未知的 `--a.b.c VALUE` 选项，由 CLI 转发到
+`apply_overrides`：
 
 ```bash
 genelab play TASK_ID --env.simulation.dt 0.005
 ```
 
-用 `genelab info TASK_ID` 查看有效路径。
+可用路径见 `genelab info TASK_ID`。
 
 ## 交互模式
 
-stdin 是 TTY 时，CLI 可以为缺失 task id、未知注册表名字、非法 `--agent`、未知 override 路径弹出选择器。CI 和脚本中会直接抛出相同错误。
+当 stdin 是 TTY 时，CLI 可以为缺失的 task id、未知的注册名、非法 `--agent` 或
+未知 override 路径弹出交互选择器；CI 与脚本中则直接抛出同样的错误。
 
 ## 另见
 
 - [CLI 参考](../reference/cli.md)
 - [发现：list 与 info](list-info.md)
 - [play 与 train](play-train.md)
+- [Eval 与 export](../concepts/eval-and-export.md)
diff --git a/docs/cli/play-train.en.md b/docs/cli/play-train.en.md
index 51d8a24a..5c49bd41 100644
--- a/docs/cli/play-train.en.md
+++ b/docs/cli/play-train.en.md
@@ -1,7 +1,8 @@
-# Play and Train
+# Runtime: play and train
 
 `play` runs a registered task. `train` runs a registered task through a supported runner when the
-task provides an agent config.
+task provides an agent config. This page also covers the post-training runtime subcommands —
+`eval`, `export`, and `benchmark` — which take the same checkpoint produced by `train`.
 
 ## Play
 
@@ -23,8 +24,8 @@ The policy options (`--agent`, `--checkpoint`, `--num-envs`, `--prof*`) apply on
 tasks — those whose play env config is a `ManagerBasedRlEnvCfg`. Non-RL **scene-playback
 demos** (e.g. `GeneLab-Rubiks-Play-v0`, `GeneLab-Wuji-Hand-Playback-v0`), whose config
 subclasses the base `ManagerBasedEnvCfg`, run their own built-in playback; passing those
-options prints a warning and they are ignored. `--steps` / `--vis` / `--gpu` and dotted
-config overrides still apply to both.
+options prints a warning and they are ignored. `--steps` / `--vis` / `--headless` / `--gpu` /
+`--dt` and dotted config overrides still apply to both.
 
 Checkpoint replay:
 
@@ -44,6 +45,19 @@ genelab play TASK_ID \
       --checkpoint <ckpt> --headless
     ```
 
+## Shortcut flags
+
+Both `play` and `train` rewrite the following shortcuts into `env.simulation.*` overrides:
+
+| Shortcut | Override |
+|---|---|
+| `-v`, `--vis` | `env.simulation.vis=true` |
+| `--headless` | `env.simulation.vis=false` (mutually exclusive with `--vis`) |
+| `--gpu` | `env.simulation.gpu=true` |
+| `--steps N` | play: `env.simulation.steps=N`; train: alias for `--max_iterations N` |
+| `--dt SECONDS` | `env.simulation.dt=SECONDS` |
+| `--a.b.c VALUE` | Any dotted cfg path |
+
 ## Train
 
 ```bash
@@ -57,7 +71,38 @@ genelab train TASK_ID --gpus 4 --num_envs 4096
 ```
 
 `--num_envs` is total across ranks and must divide evenly by `--gpus`. Use
-`--num_envs_per_gpu` for per-rank semantics.
+`--num_envs_per_gpu` for per-rank semantics (mutually exclusive with `--num_envs`).
+Multi-GPU is RSL-RL only; the first entry automatically relaunches under `torchrun`.
+
+### In-training eval
+
+Pass `--eval_every K` to run a deterministic rollout every K iterations. On improvement, the
+runner writes `best_model.<ext>` into `--log_dir`:
+
+```bash
+genelab train TASK_ID --eval_every 50 --eval_episodes 20
+```
+
+| Option | Meaning (default) |
+|---|---|
+| `--eval_every K` | Evaluate every K iterations. |
+| `--eval_episodes N` | Episodes per evaluation (10). |
+| `--eval_num_envs N` | Parallel envs during eval (matches training). |
+| `--eval_seed N` | RNG seed for the eval rollout (0). |
+
+### Multi-seed train
+
+`--seeds 1,2,3` fans out the current `train` invocation into one independent subprocess per seed:
+
+```bash
+genelab train TASK_ID --seeds 1,2,3,4 --parallel 2 --num_envs 4096
+```
+
+- `--parallel N` caps concurrency (default 1 — sequential).
+- Each child is invoked with `--seed S` and `--log_dir <parent>/seed_<S>`.
+- Without an explicit `--log_dir`, the parent is
+  `logs/multi-seed/<task_id>/<YYYY-MM-DD_HH-MM-SS>/`.
+- If any seed fails, the command exits non-zero.
 
 ## RL backends
 
@@ -95,6 +140,47 @@ GENELAB_SB3_DEMO_PATH=/tmp/franka_pp_demos.npz \
   --gpu --num-envs 32 --max-iterations 2000000
 ```
 
+## Post-training subcommands
+
+`eval`, `export`, and `benchmark` all take a registered task plus a checkpoint and reuse the
+task's play env config — see [Eval and export](../concepts/eval-and-export.md) for the
+conceptual picture and script-level details.
+
+### Eval
+
+Deterministic rollout that writes `eval.json` (`return_mean`, `length_mean`, and
+`success_rate` if the task publishes `extras['is_success']`):
+
+```bash
+genelab eval TASK_ID logs/.../model_300.pt \
+  --num-envs 64 --episodes 100 --out eval.json
+```
+
+`--deterministic` / `--stochastic` toggles the policy mode; `--max-steps` caps the rollout.
+
+### Export
+
+Export the policy as TorchScript or ONNX (per-term scale/clip baked into the model):
+
+```bash
+genelab export TASK_ID logs/.../model_300.pt --format onnx --out policy.onnx
+```
+
+A sibling `<OUTPUT>.metadata.json` records the observation schema.
+
+### Benchmark
+
+Batch eval driven by a JSON suite, aggregated into one report:
+
+```bash
+genelab benchmark --suite suite.json --out report.json
+genelab benchmark --suite suite.json --reference baseline.json --tolerance 0.1
+```
+
+`suite.json` is `[{"task": ..., "checkpoint": ..., "episodes": ..., "seed": ..., "num_envs": ...}, ...]`.
+With `--reference`, the command compares `return_mean` against the baseline and exits non-zero
+when any task drops more than `--tolerance` — usable directly as a CI regression gate.
+
 ## Config overrides
 
 Any unknown option after the task id is treated as a dotted config override:
@@ -110,3 +196,4 @@ genelab play TASK_ID \
 - [Run RL Experiments](../best-practices/rl-experiments.md)
 - [CLI Reference](../reference/cli.md)
 - [RL runner](../concepts/rl-runner.md)
+- [Eval and export](../concepts/eval-and-export.md)
diff --git a/docs/cli/play-train.zh.md b/docs/cli/play-train.zh.md
index 1c925e4a..9a629979 100644
--- a/docs/cli/play-train.zh.md
+++ b/docs/cli/play-train.zh.md
@@ -1,6 +1,8 @@
-# play 与 train
+# 运行时：play 与 train
 
 `play` 运行已注册任务。`train` 在 task 提供 agent 配置时，通过支持的 runner 训练任务。
+这一页还覆盖训练后用同一个 checkpoint 做 `eval` / `export` / `benchmark` 的几条
+runtime 子命令。
 
 ## Play
 
@@ -16,12 +18,13 @@ genelab play TASK_ID --agent random --steps 128
 |---|---|
 | `zero` | 零动作。未提供 checkpoint 时默认。 |
 | `random` | `[-1, 1]` 均匀随机动作。 |
-| `trained` | 加载 checkpoint 并使用 runner inference policy。 |
+| `trained` | 加载 checkpoint 并使用 runner 的 inference policy。 |
 
-策略相关选项（`--agent`、`--checkpoint`、`--num-envs`、`--prof*`）仅适用于 RL 任务，即 play
-环境配置为 `ManagerBasedRlEnvCfg` 的任务。对于配置继承自基类 `ManagerBasedEnvCfg` 的非 RL
-**场景回放示例**（如 `GeneLab-Rubiks-Play-v0`、`GeneLab-Wuji-Hand-Playback-v0`），将运行其
-自带的回放逻辑；传入这些选项会打印告警并被忽略。`--steps` / `--vis` / `--gpu` 以及点号配置
+策略相关选项（`--agent`、`--checkpoint`、`--num-envs`、`--prof*`）仅适用于 RL 任务，
+即 play 环境配置为 `ManagerBasedRlEnvCfg` 的任务。对于配置继承自基类
+`ManagerBasedEnvCfg` 的非 RL **场景回放示例**（如 `GeneLab-Rubiks-Play-v0`、
+`GeneLab-Wuji-Hand-Playback-v0`），将运行其自带的回放逻辑；传入这些选项会打印
+警告并被忽略。`--steps` / `--vis` / `--headless` / `--gpu` / `--dt` 以及点号配置
 覆盖对两类任务都生效。
 
 checkpoint 回放：
@@ -42,6 +45,19 @@ genelab play TASK_ID \
       --checkpoint <ckpt> --headless
     ```
 
+## 简写 flag
+
+`play` 和 `train` 都会把以下简写改写为 `env.simulation.*` override：
+
+| 简写 | 等价 override |
+|---|---|
+| `-v`、`--vis` | `env.simulation.vis=true` |
+| `--headless` | `env.simulation.vis=false`（与 `--vis` 互斥） |
+| `--gpu` | `env.simulation.gpu=true` |
+| `--steps N` | play: `env.simulation.steps=N`；train: 等价于 `--max_iterations N` |
+| `--dt SECONDS` | `env.simulation.dt=SECONDS` |
+| `--a.b.c VALUE` | 任意 dotted cfg path |
+
 ## Train
 
 ```bash
@@ -55,7 +71,38 @@ genelab train TASK_ID --gpus 4 --num_envs 4096
 ```
 
 `--num_envs` 表示所有 rank 的总数，必须能被 `--gpus` 整除。每 rank 语义用
-`--num_envs_per_gpu`。
+`--num_envs_per_gpu`（与 `--num_envs` 互斥）。多 GPU 仅 RSL-RL 后端支持，
+首次进入会自动通过 `torchrun` 重启。
+
+### 训练中评估
+
+加 `--eval_every K` 即可每 K 次迭代跑一次确定性 rollout；如有改进，会把
+`best_model.<ext>` 写入 `--log_dir`：
+
+```bash
+genelab train TASK_ID --eval_every 50 --eval_episodes 20
+```
+
+| 选项 | 含义（默认） |
+|---|---|
+| `--eval_every K` | 每 K 次迭代评估一次。 |
+| `--eval_episodes N` | 每次评估的回合数（10）。 |
+| `--eval_num_envs N` | 评估用的并行 env 数（与训练相同）。 |
+| `--eval_seed N` | 评估的 RNG seed（0）。 |
+
+### 多 seed train
+
+`--seeds 1,2,3` 把当前 `train` 调用拆成多个独立子进程，每个 seed 一个：
+
+```bash
+genelab train TASK_ID --seeds 1,2,3,4 --parallel 2 --num_envs 4096
+```
+
+- `--parallel N` 限制并发数，默认 1（顺序跑）。
+- 每个子进程会被注入 `--seed S` 和 `--log_dir <parent>/seed_<S>`。
+- 不显式指定 `--log_dir` 时，父目录是
+  `logs/multi-seed/<task_id>/<YYYY-MM-DD_HH-MM-SS>/`。
+- 任一 seed 失败，命令以非零状态退出。
 
 ## RL 后端
 
@@ -90,6 +137,47 @@ GENELAB_SB3_DEMO_PATH=/tmp/franka_pp_demos.npz \
   --gpu --num-envs 32 --max-iterations 2000000
 ```
 
+## 训练后的工具命令
+
+`eval`、`export`、`benchmark` 都以已注册 task + checkpoint 作为入参，复用 task 的
+play env 配置——理论上的概念与脚本细节见
+[Eval 与 export](../concepts/eval-and-export.md)。
+
+### Eval
+
+确定性 rollout，写出 `eval.json`（包含 `return_mean`、`length_mean`，若 task 暴露
+`extras['is_success']` 则附带 `success_rate`）：
+
+```bash
+genelab eval TASK_ID logs/.../model_300.pt \
+  --num-envs 64 --episodes 100 --out eval.json
+```
+
+`--deterministic` / `--stochastic` 切换策略；`--max-steps` 设安全步上限。
+
+### Export
+
+把 policy 导出为 TorchScript 或 ONNX（per-term scale/clip 烘焙进模型）：
+
+```bash
+genelab export TASK_ID logs/.../model_300.pt --format onnx --out policy.onnx
+```
+
+同时写出 `<OUTPUT>.metadata.json`，记录 obs schema。
+
+### Benchmark
+
+按 JSON 套件批量 eval 并聚合参考数：
+
+```bash
+genelab benchmark --suite suite.json --out report.json
+genelab benchmark --suite suite.json --reference baseline.json --tolerance 0.1
+```
+
+`suite.json` 是 `[{"task": ..., "checkpoint": ..., "episodes": ..., "seed": ..., "num_envs": ...}, ...]`。
+传 `--reference` 时与 baseline 比 `return_mean`，下降超过 `--tolerance` 视为回归，
+命令以非零状态退出——可直接作为 CI 回归门。
+
 ## 配置 override
 
 task id 后的任何未知选项都会被当作 dotted config override：
@@ -105,3 +193,4 @@ genelab play TASK_ID \
 - [运行 RL 实验](../best-practices/rl-experiments.md)
 - [CLI 参考](../reference/cli.md)
 - [RL runner](../concepts/rl-runner.md)
+- [Eval 与 export](../concepts/eval-and-export.md)
diff --git a/docs/cli/profiling.en.md b/docs/cli/profiling.en.md
index 378040f5..eddc8c21 100644
--- a/docs/cli/profiling.en.md
+++ b/docs/cli/profiling.en.md
@@ -15,21 +15,26 @@ genelab train TASK_ID \
 Open traces:
 
 ```bash
-genelab prof open logs/torch_profile
+genelab prof open                       # defaults to logs/torch_profile
+genelab prof open <log_dir>             # explicit directory
+genelab prof open <log_dir> --port 6007 # custom TensorBoard port
+genelab prof open <log_dir> --host 0.0.0.0
 ```
 
 ## Flags
 
-| Flag | Meaning |
-|---|---|
-| `--prof` | Enable profiling. |
-| `--prof-out PATH` | Trace output directory. |
-| `--prof-wait N` | Initial wait steps. |
-| `--prof-warmup N` | Warmup steps before recording. |
-| `--prof-active N` | Recorded steps per cycle. |
-| `--prof-repeat N` | Number of cycles. |
-| `--prof-record-shapes` | Record tensor shapes. |
-| `--prof-with-stack` | Capture Python stacks. Higher overhead. |
+| Flag | Env-var equivalent | Meaning (default) |
+|---|---|---|
+| `--prof` | `GENELAB_PROFILE` | Enable profiling. |
+| `--prof-out PATH` | `GENELAB_PROFILE_OUT` | TensorBoard trace directory. |
+| `--prof-wait N` | `GENELAB_PROFILE_WAIT` | Initial wait steps (10). |
+| `--prof-warmup N` | `GENELAB_PROFILE_WARMUP` | Warmup steps before recording (5). |
+| `--prof-active N` | `GENELAB_PROFILE_ACTIVE` | Recorded steps per cycle (10). |
+| `--prof-repeat N` | `GENELAB_PROFILE_REPEAT` | Number of cycles (2). |
+| `--prof-record-shapes` | `GENELAB_PROFILE_RECORD_SHAPES` | Record tensor shapes. |
+| `--prof-with-stack` | `GENELAB_PROFILE_WITH_STACK` | Capture Python stacks. Higher overhead. |
+
+CLI flags always take precedence over the env vars.
 
 ## Practical defaults
 
@@ -41,6 +46,11 @@ quickly on large vectorized runs.
 Only the main process writes profiler traces. Keep `--prof-active` small when profiling distributed
 training.
 
+## `prof open` prerequisites
+
+- `tensorboard` must be on `PATH` (`uv pip install tensorboard`, or via this checkout's `rl` extra).
+- The directory you pass must exist; an empty directory is fine.
+
 ## See also
 
 - [Run RL Experiments](../best-practices/rl-experiments.md)
diff --git a/docs/cli/profiling.zh.md b/docs/cli/profiling.zh.md
index d22005cf..a295d10e 100644
--- a/docs/cli/profiling.zh.md
+++ b/docs/cli/profiling.zh.md
@@ -15,30 +15,41 @@ genelab train TASK_ID \
 打开 trace：
 
 ```bash
-genelab prof open logs/torch_profile
+genelab prof open                       # 默认 logs/torch_profile
+genelab prof open <log_dir>             # 指定目录
+genelab prof open <log_dir> --port 6007 # 指定 TensorBoard 端口
+genelab prof open <log_dir> --host 0.0.0.0
 ```
 
 ## 参数
 
-| 标志 | 含义 |
-|---|---|
-| `--prof` | 启用 profiling。 |
-| `--prof-out PATH` | trace 输出目录。 |
-| `--prof-wait N` | 初始等待步数。 |
-| `--prof-warmup N` | 记录前 warmup 步数。 |
-| `--prof-active N` | 每个周期记录的步数。 |
-| `--prof-repeat N` | 周期数。 |
-| `--prof-record-shapes` | 记录 tensor shape。 |
-| `--prof-with-stack` | 捕获 Python stack，开销更高。 |
+| 标志 | 等价环境变量 | 含义（默认） |
+|---|---|---|
+| `--prof` | `GENELAB_PROFILE` | 启用 profiling。 |
+| `--prof-out PATH` | `GENELAB_PROFILE_OUT` | TensorBoard trace 目录。 |
+| `--prof-wait N` | `GENELAB_PROFILE_WAIT` | 初始等待步数（10）。 |
+| `--prof-warmup N` | `GENELAB_PROFILE_WARMUP` | 记录前 warmup 步数（5）。 |
+| `--prof-active N` | `GENELAB_PROFILE_ACTIVE` | 每个周期记录的步数（10）。 |
+| `--prof-repeat N` | `GENELAB_PROFILE_REPEAT` | 周期数（2）。 |
+| `--prof-record-shapes` | `GENELAB_PROFILE_RECORD_SHAPES` | 记录 tensor shape。 |
+| `--prof-with-stack` | `GENELAB_PROFILE_WITH_STACK` | 捕获 Python stack，开销更高。 |
+
+命令行 flag 始终优先于环境变量。
 
 ## 实用默认值
 
-先跑短 profile。GeneLab 的 RL loop 每个 env step 推进一次 profiler step，大规模 vectorized run 的 trace 会增长很快。
+先跑短 profile。GeneLab 的 RL loop 每个 env step 推进一次 profiler step，大规模
+vectorized run 的 trace 会增长很快。
 
 ## 分布式运行
 
 只有 main process 写 profiler trace。profile 分布式训练时保持 `--prof-active` 较小。
 
+## `prof open` 前置条件
+
+- `tensorboard` 必须在 PATH 上（`uv pip install tensorboard`，或本仓库的 `rl` extra）。
+- 指定的目录必须存在；空目录是合法的。
+
 ## 另见
 
 - [运行 RL 实验](../best-practices/rl-experiments.md)
diff --git a/docs/cli/project-new.en.md b/docs/cli/project-new.en.md
index bea81543..0dc81cc2 100644
--- a/docs/cli/project-new.en.md
+++ b/docs/cli/project-new.en.md
@@ -1,7 +1,8 @@
-# Project New
+# Project: project new
 
 `genelab project new` creates a standalone extension package. Use it for downstream projects rather
-than editing `src/genelab/`.
+than editing `src/genelab/`. `project` is a Typer subcommand group: `genelab project` (with no
+subcommand) lists the available actions.
 
 ## Usage
 
@@ -13,10 +14,10 @@ Options:
 
 | Option | Description |
 |---|---|
-| `--path PATH`, `-p PATH` | Parent directory for the generated project. |
+| `--path PATH`, `-p PATH` | Parent directory for the generated project (default `.`). |
 | `--package NAME` | Python package name. Defaults to normalized `NAME`. |
 | `--task-id ID` | Initial task id. Defaults to `<PackageName>-Example-v0`. |
-| `--force` | Overwrite scaffold files if the target exists. |
+| `--force` | Overwrite scaffold files when the target directory exists and is non-empty. Targets that exist but are not directories still error out. |
 
 ## Generated structure
 
@@ -32,6 +33,9 @@ my_robot_project/
     └── tasks.py
 ```
 
+The generated `pyproject.toml` references the local GeneLab source by relative path
+(`tool.uv.sources`), so `uv pip install -e .` works without publishing anything.
+
 ## After scaffolding
 
 ```bash
@@ -40,6 +44,9 @@ genelab list tasks
 genelab play MyRobotProject-Example-v0 --steps 3
 ```
 
+If the target directory is non-empty and `--force` is not set, the command exits non-zero without
+overwriting any existing file.
+
 ## See also
 
 - [Build an Extension Project](../best-practices/extension-projects.md)
diff --git a/docs/cli/project-new.zh.md b/docs/cli/project-new.zh.md
index 1353b2a7..d16d56c6 100644
--- a/docs/cli/project-new.zh.md
+++ b/docs/cli/project-new.zh.md
@@ -1,6 +1,7 @@
-# 新建项目
+# 项目：project new
 
 `genelab project new` 创建独立扩展包。下游项目应使用它，而不是直接修改 `src/genelab/`。
+`project` 是 Typer 子命令组：`genelab project`（不带子命令）会列出可用动作。
 
 ## 用法
 
@@ -12,10 +13,10 @@ genelab project new my_robot_project
 
 | 选项 | 说明 |
 |---|---|
-| `--path PATH`、`-p PATH` | 生成项目的父目录。 |
+| `--path PATH`、`-p PATH` | 生成项目的父目录（默认 `.`）。 |
 | `--package NAME` | Python 包名，默认由 `NAME` 规范化得到。 |
 | `--task-id ID` | 初始 task id，默认 `<PackageName>-Example-v0`。 |
-| `--force` | 目标已存在时覆盖骨架文件。 |
+| `--force` | 目标已存在且非空时覆盖骨架文件；目标存在但非目录时仍会报错。 |
 
 ## 生成结构
 
@@ -31,6 +32,9 @@ my_robot_project/
     └── tasks.py
 ```
 
+生成的 `pyproject.toml` 通过相对路径 (`tool.uv.sources`) 引用本地 GeneLab 源码，
+以便不必发布也能 `uv pip install -e .`。
+
 ## 生成后
 
 ```bash
@@ -39,6 +43,8 @@ genelab list tasks
 genelab play MyRobotProject-Example-v0 --steps 3
 ```
 
+如果目标目录非空且未传 `--force`，命令会以非零状态退出而不会覆盖现有文件。
+
 ## 另见
 
 - [构建扩展项目](../best-practices/extension-projects.md)
diff --git a/mkdocs.yml b/mkdocs.yml
index efefd895..423e0fe7 100644
--- a/mkdocs.yml
+++ b/mkdocs.yml
@@ -201,11 +201,11 @@ nav:
           - PointCloudTactileSensor: reference/sensors/tactile_pointcloud.md
       - API Reference: api/reference.md
   - CLI:
-      - Overview: cli/overview.md
+      - CLI Overview: cli/overview.md
       - 'Discovery: list and info': cli/list-info.md
-      - Play and Train: cli/play-train.md
+      - 'Runtime: play and train': cli/play-train.md
       - Profiling: cli/profiling.md
-      - Project New: cli/project-new.md
+      - 'Project: project new': cli/project-new.md
   - Concepts:
       - Registry: concepts/registry.md
       - Configs: concepts/configs.md
diff --git a/pyproject.toml b/pyproject.toml
index c8a5a600..f521a29a 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -185,12 +185,11 @@ addopts = ["-ra"]
 pythonpath = ["examples/genelab_examples/src", "examples/unitree/src", "examples/inverted_pendulum/src", "examples/genelab_showcase/src", "examples/franka_pick_and_place/src"]
 testpaths = ["tests"]
 
-# Layering contracts (ROADMAP §9 PR R0.3 / ADR-0009).
+# Layering contracts.
 #
-# Currently runs in CI as a non-required step (`continue-on-error: true`).
-# Lint-only mode is the explicit choice for R0–R6: every R-phase that lands
-# trims the baseline violation list, and R7 (ADR-0009 §"R7 — flip to blocking")
-# converts this job to a required check once the list is empty.
+# Currently runs in CI as a non-required step (`continue-on-error: true`)
+# until the baseline violation list is empty, at which point this job
+# converts to a required check.
 #
 # ``exclude_type_checking_imports`` skips ``if TYPE_CHECKING: from ... import``
 # blocks. The manager-based MDP design intentionally has domain modules
@@ -201,7 +200,7 @@ testpaths = ["tests"]
 root_packages = ["genelab"]
 exclude_type_checking_imports = true
 
-# Domain modules must not reach UP into cli / rl. (After R7.3a + R7.3b this is clean.)
+# Domain modules must not reach UP into cli / rl.
 [[tool.importlinter.contracts]]
 name = "Domain modules are below cli / rl"
 type = "forbidden"
@@ -225,7 +224,7 @@ forbidden_modules = [
 
 # Term logic must not trigger asset downloads. ``asset_zoo`` is the deliberate
 # exception: it is the asset catalog, so fetching assets via ``utils.download``
-# (a downward domain -> utils import) is its purpose (ADR-0009 / R7.3c).
+# (a downward domain -> utils import) is its purpose.
 [[tool.importlinter.contracts]]
 name = "Domain (except asset_zoo) does not import utils.download"
 type = "forbidden"
@@ -245,7 +244,7 @@ forbidden_modules = [
     "genelab.utils.download",
 ]
 
-# ADR-0001 / R1: this contract reports violations until the cycle is broken.
+# This contract reports violations until the cycle is broken.
 [[tool.importlinter.contracts]]
 name = "rl.backends does not import rl.runner"
 type = "forbidden"
@@ -258,7 +257,7 @@ forbidden_modules = ["genelab.rl.runner"]
 # importlinter treats as mutually *independent* — that wrongly forbids the legitimate
 # intra-domain imports a manager-based env relies on (envs -> scene, mdp -> managers,
 # entity -> actuator, …). The contracts below capture "no band imports a band above
-# it" while leaving intra-band imports free (ADR-0009 / ROADMAP §9 R7.3d).
+# it" while leaving intra-band imports free.
 
 # rl sits below cli.
 [[tool.importlinter.contracts]]
@@ -303,7 +302,7 @@ modules = [
     "genelab.rl.backends.rsl_rl",
 ]
 
-# ADR-0014: the contracts band (ports) is dependency-free within genelab — it must
+# The contracts band (ports) is dependency-free within genelab — it must
 # not reach up into the domain / rl / cli bands. (Importing torch / stdlib is fine.)
 [[tool.importlinter.contracts]]
 name = "contracts band imports nothing in domain / rl / cli"
@@ -325,10 +324,10 @@ forbidden_modules = [
     "genelab.envs",
 ]
 
-# ADR-0014: domain terms type-hint against ``contracts.EnvContext`` rather than the
+# Domain terms type-hint against ``contracts.EnvContext`` rather than the
 # concrete env. Runtime imports of ``genelab.envs`` from the domain are forbidden;
 # the remaining ``TYPE_CHECKING`` env hints are excluded globally
-# (``exclude_type_checking_imports``) and are migrated to ``contracts`` in PR 3.2–3.4.
+# (``exclude_type_checking_imports``) and will be migrated to ``contracts``.
 [[tool.importlinter.contracts]]
 name = "Domain does not import concrete envs at runtime"
 type = "forbidden"
diff --git a/uv.lock b/uv.lock
index 7ea5ef4d..ecf65ba8 100644
--- a/uv.lock
+++ b/uv.lock
@@ -863,7 +863,7 @@ wheels = [
 
 [[package]]
 name = "genelab"
-version = "0.2.0"
+version = "0.3.0"
 source = { editable = "." }
 dependencies = [
     { name = "genesis-world" },