diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--hbviuqx.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--hbviuqx.json
new file mode 100644
index 0000000000000000000000000000000000000000..26d09f42b01962ab7aa2a698241e622f0d055f77
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--hbviuqx.json
@@ -0,0 +1,177 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-privileged--hbviuqx",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "live_check_success": 1,
+   "deterministic": 1,
+   "inserted": 1,
+   "tip_to_axis_m": 0.00112,
+   "axial_t": 0.549,
+   "socket_z": 0.0228,
+   "peg_z": 0.021,
+   "time_s": 56.2,
+   "n_actions": 2810,
+   "video_rendered": 1,
+   "grader_error": 0,
+   "agent_sessions": 68,
+   "agent_resets": 147,
+   "agent_cost_self_reported": 1,
+   "agent_traces": 90,
+   "agent_trace_videos": 6,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-privileged--hbviuqx/replay.mp4",
+  "duration_s": 14.12,
+  "fps": 25.0,
+  "frames": 353,
+  "sha256": "2f9df38d954e23e25bbe68e5266d4315db61cc41822679bd9a01a15075f2e587"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [
+     [
+      12.2,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00 s 的两个全景视图显示红色插销和蓝色套筒分开放在桌面，插入关系初始不成立。"
+     },
+     {
+      "t": 6.6,
+      "log_step": null,
+      "text_zh": "观察：视频 6.40–6.80 s，机器人右夹爪的两指夹住红色插销并将其带离原桌面位置，随后移向蓝色套筒；左夹爪没有拿起套筒，蓝管仍在原桌面位置。"
+     },
+     {
+      "t": 12.2,
+      "log_step": null,
+      "text_zh": "推断：视频 11.68–12.40 s 的右腕视图显示右夹爪继续将红色插销送入蓝色套筒，红条插入端被蓝管遮住，外露长度进一步缩短；将视频约 12.2 s 记为达到验收深度的近似起点，时间不确定度约 0.5 s。无逐帧状态日志，图像不能独立精确测量 5 毫米轴距和 12 毫米深度阈值；末态达标由评分器原生读数确认。"
+     },
+     {
+      "t": 14.08,
+      "log_step": 146,
+      "text_zh": "观察：视频 13.12–14.08 s 逐帧抽查中，右夹爪仍握住插销，红条端部保持在蓝管内，外露长度继续缩短，未见退出；左夹爪仍未持套筒。log step 146 对最终交付的 2810 行动作重新生成并独立重放，均返回 inserted=True、tip_to_axis_m=0.00112、axial_t=0.549；inputs/run.json 的 reward 同样给出 inserted=1、success=1、deterministic=1，末态满足 DAG 的几何判据，且不要求松手。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "手指/夹爪",
+      "how": "抓握",
+      "note_zh": "观察：视频 6.40–6.80 s，机器人右侧二指夹爪夹住并搬运红色插销；视频 8.00–14.08 s 持续握持插销并向留在桌面上的蓝色套筒送入。左夹爪未抓握套筒，本次实际为单夹爪抓握插销完成插入。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.00112
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": 0.549
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": 0.549
+  }
+ ],
+ "failure": null,
+ "summary_zh": "机器人右夹爪抓住并搬运红色插销，向桌面上的蓝色套筒逐步插入，视频约 12.2 s 达标并保持至末帧，最终完成 1/1 节点；评分器确认末态轴距为 0.00112 米、轴向参数为 0.549。任务设计问题：inputs/run.json 的 task_text 要求双臂分别抓住并抬起两件物体，但视频 0.00–14.08 s 中蓝色套筒始终留在桌面、左夹爪未拿取，评分器仍判成功，故按实际几何验收标注。没有状态日志，毫米级首次达标时刻只能结合腕部画面近似判断，末态判定有评分器和 log step 146 的最终交付重放支持。",
+ "conf": 7,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "单节点 peg_inserted 标注（约12.2 s 起、末态成立、单夹爪抓握）与帧和评分器读数一致；腕部视图中红条自约11 s 起进入蓝管并逐步缩短，12.2 s 的近似起点在不确定度内，任务设计问题说明合理。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:21:47Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--hbviuqx/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--hbviuqx/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "ca5e3fa264a3965314200229040f36c80e7b9b005ec9fe24c8022339547f5f54",
+   "video_sha256": "2f9df38d954e23e25bbe68e5266d4315db61cc41822679bd9a01a15075f2e587",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--kbikpjm.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--kbikpjm.json
new file mode 100644
index 0000000000000000000000000000000000000000..594d4f9308785b8b26bea6e58577d825721cfbd3
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--kbikpjm.json
@@ -0,0 +1,171 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-privileged--kbikpjm",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "live_check_success": 1,
+   "deterministic": 1,
+   "inserted": 1,
+   "tip_to_axis_m": 0.00258,
+   "axial_t": 0.4887,
+   "socket_z": 0.1872,
+   "peg_z": 0.1852,
+   "time_s": 28.8,
+   "n_actions": 1440,
+   "video_rendered": 1,
+   "grader_error": 0,
+   "agent_sessions": 7,
+   "agent_resets": 13,
+   "agent_cost_self_reported": 1,
+   "agent_traces": 4,
+   "agent_trace_videos": 4
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-privileged--kbikpjm/replay.mp4",
+  "duration_s": 14.48,
+  "fps": 25.0,
+  "frames": 362,
+  "sha256": "49df866ab4e92644044e29c1e5c1e29ff0f0a0fae149df83dc6d938f60e908aa"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [
+     [
+      13.0,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00 s，蓝色套筒与红色插销分开放在桌面上，机器人两只夹爪尚未夹持物体，插入关系不成立。"
+     },
+     {
+      "t": 7.6,
+      "log_step": null,
+      "text_zh": "观察：视频 4.8–6.4 s，两侧二指夹爪分别在蓝色套筒和红色插销两侧合拢；视频 6.8–8.4 s，两物体随各自夹爪上升并离开原来的桌面位置，正面视图可见两物体悬空。"
+     },
+     {
+      "t": 11.6,
+      "log_step": null,
+      "text_zh": "观察：视频 9.0–11.6 s，机器人左夹爪持套筒，右夹爪持插销并横向移至套筒开口方向，随后沿轴向推进；仅对准的阶段不计完成。"
+     },
+     {
+      "t": 13.0,
+      "log_step": null,
+      "text_zh": "推断：结合视频 12.4、12.6、12.8、13.0、13.2、13.4 s 的连续腕部视图，右夹爪所持红色插销端部逐渐进入蓝色套筒内，端部被管壁遮住；约视频 13.0 s 达到所需插入深度，首次达标时刻有约 ±0.5 s 的视觉不确定性。无逐帧状态日志，不能从像素单独精确测量轴距和深度阈值；末态达标由本次评分器读数支持。"
+     },
+     {
+      "t": 14.44,
+      "log_step": 22,
+      "text_zh": "观察：视频 13.4–14.44 s，两夹爪持续持物，插销端部仍在蓝色套筒内，未见退出；inputs/run.json 的 reward 给出 success=1、inserted=1、deterministic=1，端点轴距 tip_to_axis_m=0.00258，小于 0.005 m，axial_t=0.4887，处于 [-0.875,0.75]。log step 19–22 记录最终动作文件保存并从初态重放，log step 22 的重放为 1440 个控制步，读数与评分器一致；不以早期 trial 的失败结果标记本次视频。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "双臂协作",
+      "how": "抓握",
+      "note_zh": "观察：视频 4.8–8.4 s，机器人左侧二指夹爪夹住蓝色套筒、右侧二指夹爪夹住红色插销并各自抬起；视频 9.0–14.44 s，左夹爪持续持套筒，右夹爪持插销完成对准和推进，两臂协同插入，末态均未松手。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.00258
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": 0.4887
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": 0.4887
+  }
+ ],
+ "failure": null,
+ "summary_zh": "机器人双臂分别抓握套筒和插销并抬离桌面，约视频 13.0 s 完成插入，保持至视频 14.44 s，唯一目标节点末态为真。评分器确认成功且重放确定，末态端点轴距为 2.58 毫米、轴向参数为 0.4887，均满足验收。无逐帧状态日志，毫米级首次达标时刻只能由连续画面估计，约有半秒误差，末态几何判断由评分器读数补足。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "唯一目标节点 peg_inserted 的区间 [13.0, null]、末态为真、双臂协作抓握均与画面和评分器读数一致；成功运行无断点，无异常。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:14:05Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--kbikpjm/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--kbikpjm/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "d4d53290611e757ec0717a683befa4b7dc6767b04f5f232c280d04d094d5f5d7",
+   "video_sha256": "49df866ab4e92644044e29c1e5c1e29ff0f0a0fae149df83dc6d938f60e908aa",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--kehyyib.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--kehyyib.json
new file mode 100644
index 0000000000000000000000000000000000000000..96461db47313417e3a98ea1eaef1cb21b9dc8849
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--kehyyib.json
@@ -0,0 +1,187 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-privileged--kehyyib",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "live_check_success": 1,
+   "deterministic": 1,
+   "inserted": 1,
+   "tip_to_axis_m": 0.0006,
+   "axial_t": 0.1096,
+   "socket_z": 0.1746,
+   "peg_z": 0.1778,
+   "time_s": 13.2,
+   "n_actions": 660,
+   "video_rendered": 1,
+   "grader_error": 0,
+   "agent_sessions": 41,
+   "agent_resets": 55,
+   "agent_cost_self_reported": 1,
+   "agent_traces": 15,
+   "agent_trace_videos": 6,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-privileged--kehyyib/replay.mp4",
+  "duration_s": 13.28,
+  "fps": 25.0,
+  "frames": 332,
+  "sha256": "fde313c7dd9d095ff297dc35865e2003f01137eaae7324d19cf92a0bae6994ba"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [
+     [
+      10.4,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0 s 的外部视图中，蓝色套筒与红色插销分开放在桌面，插销端点未进入套筒。"
+     },
+     {
+      "t": 4.8,
+      "log_step": null,
+      "text_zh": "观察：视频 3.2–4.0 s，机器人两只二指夹爪分别从两侧夹住蓝色套筒和红色插销；视频 4.4–5.6 s 的正面视图中，两物体随各自夹爪升起，原桌面位置空出，物体与桌面之间出现间隙。"
+     },
+     {
+      "t": 10.4,
+      "log_step": null,
+      "text_zh": "观察：视频 9.0–9.6 s，机器人持红色插销的夹爪将其端点对准另一夹爪所持蓝管的开口，端点仍在开口外；视频 9.8–11.2 s 的左腕视图中，红色端点进入蓝管，露在管外的红条长度持续缩短。"
+     },
+     {
+      "t": 10.4,
+      "log_step": null,
+      "text_zh": "推断：结合视频 9.8–11.2 s 的连续深入过程及最终评分器 inserted=1，将达到规定插入深度的起点估在视频 10.4 s，误差约 0.4 s；无逐帧状态日志，图像不能精确确定轴距小于 5 毫米且轴向参数首次不大于 0.75 的瞬间。"
+     },
+     {
+      "t": 13.24,
+      "log_step": 58,
+      "text_zh": "观察：视频 12.2、12.4、12.6、12.8、13.0、13.2 和 13.24 s，外部视图中机器人继续持住两物体，未见退出或掉落，腕部视图的插入口被另一夹爪遮挡；inputs/run.json 的 reward 给出 success=1、inserted=1、deterministic=1、tip_to_axis_m=0.0006、axial_t=0.1096。log step 57 重新生成交付文件并重放成功，log step 58 再次重放所得终态几何读数与评分器一致。"
+     },
+     {
+      "t": 13.24,
+      "log_step": null,
+      "text_zh": "推断：评分器的末态轴距 0.0006 米小于 0.005 米，轴向参数 0.1096 位于 [-0.875,0.75] 内；结合视频持续插入并保持的过程，节点末态为真。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "双臂协作",
+      "how": "抓握",
+      "note_zh": "观察：视频 3.2–5.6 s，机器人左臂二指夹爪夹持蓝色套筒外壁，右臂二指夹爪夹持红色插销，两物体均随夹爪离桌；视频 9.8–11.2 s，左爪持稳套筒，右爪持插销向开口推进完成插入，之后双爪继续持物至视频 13.24 s。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.0006
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": 0.1096
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": 0.1096
+  }
+ ],
+ "failure": null,
+ "summary_zh": "机器人双臂分别抓握并抬起套筒和插销，在空中对准后于约视频 10.4 s 达到插入状态，唯一节点末态完成（1/1），无失败断点。评分器判定成功且重放确定，末态轴距为 0.6 毫米、轴向参数为 0.1096，均达标。无逐帧状态日志，首次跨越毫米级验收阈值的时刻只能近似判断，视频末段插入口受夹爪遮挡，末态几何判定结合评分器字段。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "单节点 peg_inserted 的末态为真、约 10.4 s 到达，与视频帧（9.0 s 端点在开口外，10.4 s 红条已进入蓝管）及评分器 inserted=1、tip_to_axis 0.0006 一致；成功运行无断点，contact_actual（双臂协作+抓握）与帧相符。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:16:50Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--kehyyib/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--kehyyib/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "68011c507e6a35f79e19a33da07de3420d6cc46f3ee65401550648abdeb8e8b1",
+   "video_sha256": "fde313c7dd9d095ff297dc35865e2003f01137eaae7324d19cf92a0bae6994ba",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--qknakkt.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--qknakkt.json
new file mode 100644
index 0000000000000000000000000000000000000000..71c404eb30c79cb2a5eda6837d361808d9a0825c
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--qknakkt.json
@@ -0,0 +1,177 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-privileged--qknakkt",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "live_check_success": 1,
+   "deterministic": 1,
+   "inserted": 1,
+   "tip_to_axis_m": 0.00275,
+   "axial_t": 0.6549,
+   "socket_z": 0.0228,
+   "peg_z": 0.0207,
+   "time_s": 11.94,
+   "n_actions": 597,
+   "video_rendered": 1,
+   "grader_error": 0,
+   "agent_sessions": 40,
+   "agent_resets": 48,
+   "agent_cost_self_reported": 1,
+   "agent_traces": 15,
+   "agent_trace_videos": 6,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 1.084,
+   "peak_processes": 14,
+   "peak_disk_gb": 0.24,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-privileged--qknakkt/replay.mp4",
+  "duration_s": 12.0,
+  "fps": 25.0,
+  "frames": 300,
+  "sha256": "5d9df3acf1fde936194ce208c0f3120a2dca56ad15ed8efa858d873bbfee9add"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [
+     [
+      10.0,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0 s 四视图显示红色插销与蓝色套筒分离，初始插入关系不成立。"
+     },
+     {
+      "t": 4.8,
+      "log_step": null,
+      "text_zh": "观察：视频 3.2–4.0 s 右侧机械臂的二指夹爪在红色插销两侧合拢；视频 4.8–5.2 s 红销随右夹爪离开原来桌面位置并抬起，随后向蓝色套筒开口移动；蓝色套筒仍在桌面上，左夹爪未接触它。"
+     },
+     {
+      "t": 10.0,
+      "log_step": null,
+      "text_zh": "推断：逐帧对比视频 7.0–11.8 s（每隔视频 0.2 s 取帧），右腕视图中红销端部从开口进入蓝管遮挡区域，露出长度逐渐缩短；结合末态通过插入验收的评分器读数，将满足深度要求的起点估为视频约 10.0 s，误差约 0.5 s。视频 11.0–11.96 s 保持插入，未见退出；无逐帧几何状态日志，首次越过毫米级阈值的时刻为视觉估计。"
+     },
+     {
+      "t": 11.96,
+      "log_step": 39,
+      "text_zh": "观察：视频 11.96 s 末帧仍见红销端部被蓝管遮住、右夹爪夹持红销，形态与参考成功视频 14.08 s 相符；参考仅用于识别可见线索。本次 run.json 的评分器 reward 为 success=1、inserted=1、deterministic=1、tip_to_axis_m=0.00275、axial_t=0.6549，满足轴距严格小于 0.005 m 且轴向参数位于 [-0.875,0.75]；log step 39 对已保存的 597 行动作独立重放得到相同末态读数。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "手指/夹爪",
+      "how": "抓握",
+      "note_zh": "观察：视频 3.2–5.2 s 右机械臂二指夹爪合拢夹住红色插销并将其抬离桌面，视频 7.0–11.96 s 继续握持并送入蓝色套筒；套筒由桌面承托，另一夹爪未抓握套筒。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.00275
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": 0.6549
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": 0.6549
+  }
+ ],
+ "failure": null,
+ "summary_zh": "机器人用右夹爪抓起红色插销，向桌面上的蓝色套筒对准并插入，约在视频 10.0 s 达成唯一节点，末态评分器确认成功（轴距 2.75 毫米，轴向参数 0.6549），无失败断点。任务设计问题：inputs/run.json 的任务文本要求双臂分别抓持并抬起两件物体，但本次视频 0.0–11.96 s 套筒留在桌面、单臂插入仍被评分器接受，验收仅检查插入几何关系。无逐帧状态日志，首次达到精确深度阈值的时刻只能由视频估计，约有 0.5 s 误差，末态几何数值采用评分器字段。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "单节点 peg_inserted 的值、约 10.0 s 的起点估计（已标注为推断，误差 0.5 s）、末态 final=true、contact_actual（右夹爪抓握）与评分器读数一致；成功 run 无失败断点，任务设计问题说明有 run.json 任务文本支撑。未发现需修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:19:01Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--qknakkt/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--qknakkt/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "50959e39294563f78256502012db84876512795887aa36c9a99add4fd853afa4",
+   "video_sha256": "5d9df3acf1fde936194ce208c0f3120a2dca56ad15ed8efa858d873bbfee9add",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--sd6tel3.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--sd6tel3.json
new file mode 100644
index 0000000000000000000000000000000000000000..99aa1036b38b3b9666c6ab1d3f178f70275385f8
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--sd6tel3.json
@@ -0,0 +1,177 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-privileged--sd6tel3",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "live_check_success": 1,
+   "deterministic": 1,
+   "inserted": 1,
+   "tip_to_axis_m": 0.00122,
+   "axial_t": -0.0009,
+   "socket_z": 0.1512,
+   "peg_z": 0.1504,
+   "time_s": 10.4,
+   "n_actions": 520,
+   "video_rendered": 1,
+   "grader_error": 0,
+   "agent_sessions": 15,
+   "agent_resets": 25,
+   "agent_cost_self_reported": 1,
+   "agent_traces": 4,
+   "agent_trace_videos": 4,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-privileged--sd6tel3/replay.mp4",
+  "duration_s": 10.48,
+  "fps": 25.0,
+  "frames": 262,
+  "sha256": "3982528519bddb769ab10758d7ee05a3fd65a1c03ad70fe4ea1971f309e2ddb4"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [
+     [
+      7.6,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频起始四视图中，红色插销与蓝色套筒分开放在桌面上，插销端尚未进入套筒。"
+     },
+     {
+      "t": 3.4,
+      "log_step": null,
+      "text_zh": "观察：视频 2.2–3.4 s，机器人两侧二指夹爪分别闭合夹住蓝色套筒与红色插销；视频 3.4–4.5 s，两件随各自夹爪离开原桌面位置并抬到桌面上方，随后相向对准。"
+     },
+     {
+      "t": 7.6,
+      "log_step": null,
+      "text_zh": "推断：视频 6.8–8.2 s 的逐段帧显示机器人持红色插销沿蓝色套筒开口持续送入；腕部视图中红条插入端逐渐被蓝管遮住、外露长度缩短，视频约 7.6 s 已有明显插入深度，故以该时刻作为达标起点，精确越过 12 毫米深度阈值的时间约有 ±0.5 s 误差。后续视频至 10.44 s 未见退出或错开；无逐帧状态日志，不能逐帧直接核验毫米级轴距。"
+     },
+     {
+      "t": 10.44,
+      "log_step": 26,
+      "text_zh": "观察：视频 9.48–10.44 s（含末帧），机器人双爪持续持有两件，红色插销一端保持进入蓝色套筒内。inputs/run.json 的评分器 reward 给出 success=1、inserted=1、deterministic=1、tip_to_axis_m=0.00122、axial_t=-0.0009，满足轴距小于 0.005 米且轴向参数在 [-0.875,0.75]；log step 26 对最终保存的 520 个动作执行独立重放，报告相同读数。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "双臂协作",
+      "how": "抓握",
+      "note_zh": "观察：视频 2.2–4.5 s，两侧二指夹爪分别合拢夹持蓝色套筒与红色插销，并共同将两件抬离桌面；视频 6.8–10.44 s，持套筒的夹爪固定套筒，另一夹爪持插销沿开口送入并保持，两爪各自夹住一件。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.00122
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": -0.0009
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": -0.0009
+  }
+ ],
+ "failure": null,
+ "summary_zh": "机器人双爪分别抓握并抬起套筒和插销，完成空中对准与插入；插入节点约从视频 7.6 s 起成立，并保持至视频 10.44 s 末帧，最终完成 1/1。评分器确认成功且末态轴距为 1.22 毫米、轴向参数为 -0.0009；无逐帧状态日志，首次达到精确几何阈值的时刻仅能由视频估计，约有半秒误差。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "单节点 peg_inserted 成功，评分器 inserted=1、末态轴距 1.22 mm；视频帧显示红条露出长度在 6.8–8.2 s 逐渐缩短，7.6 s 起点在半秒误差内可接受；接触 双臂协作+抓握 符合帧所见；未发现需要修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:16:25Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--sd6tel3/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--sd6tel3/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "83dddcd782425a832bfee7a8042057cfe9dc116d3e8bad49331e7ae63caf19f4",
+   "video_sha256": "3982528519bddb769ab10758d7ee05a3fd65a1c03ad70fe4ea1971f309e2ddb4",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--tzyml4h.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--tzyml4h.json
new file mode 100644
index 0000000000000000000000000000000000000000..ca0eb0782901bb5ad7f610e89e48d48ba2207b34
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-privileged--tzyml4h.json
@@ -0,0 +1,182 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-privileged--tzyml4h",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "live_check_success": 1,
+   "deterministic": 1,
+   "inserted": 1,
+   "tip_to_axis_m": 0.00097,
+   "axial_t": 0.2118,
+   "socket_z": 0.1548,
+   "peg_z": 0.1588,
+   "time_s": 13.3,
+   "n_actions": 665,
+   "video_rendered": 1,
+   "grader_error": 0,
+   "agent_sessions": 4,
+   "agent_resets": 6,
+   "agent_cost_self_reported": 1,
+   "agent_traces": 1,
+   "agent_trace_videos": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.492,
+   "peak_processes": 8,
+   "peak_disk_gb": 0.239,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-privileged--tzyml4h/replay.mp4",
+  "duration_s": 13.36,
+  "fps": 25.0,
+  "frames": 334,
+  "sha256": "6f57f1a6c194e1f08a250842d4f463ce68ceb163b9036726f833eb86730f8296"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [
+     [
+      10.0,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00 s 的外部视图显示蓝色套筒与红色插销分开放在桌面，尚未建立插入关系。"
+     },
+     {
+      "t": 4.8,
+      "log_step": null,
+      "text_zh": "观察：视频 3.0–3.8 s 两个二指夹爪分别在套筒和插销两侧闭合；视频 4.2–4.8 s 两件物体随各自夹爪上升，离开原桌面位置，随后由双臂持物移向彼此。"
+     },
+     {
+      "t": 10.0,
+      "log_step": null,
+      "text_zh": "推断：视频 9.56 s 红色插销端仍在蓝色套筒开口外，视频 9.68–10.16 s 左腕视图显示红端沿开口进入内腔、被蓝管遮住且外露长度逐渐缩短；据此将满足插入深度的首次时刻近似标在视频 10.0 s（约有 0.3 s 不确定性）。初次进入开口不等同于达到至少 12 毫米深度，画面不能精确测量毫米级阈值。"
+     },
+     {
+      "t": 13.32,
+      "log_step": null,
+      "text_zh": "观察：视频 12.00、12.36、12.60、12.84、13.08 和 13.32 s 的左腕视图仍可见红色插销伸入蓝管，外部视图显示两件物体保持离桌、两爪继续握持，未见退出或掉落；评分器 reward 报告 success=1、inserted=1、deterministic=1、tip_to_axis_m=0.00097、axial_t=0.2118，确认末态同时满足轴距与深度判据。"
+     },
+     {
+      "t": null,
+      "log_step": 8,
+      "text_zh": "观察：log step 7 的脚本输出记录成功重放后保存 665 行动作；log step 8 在新进程读取已保存文件，重放结果仍为 success=True、inserted=True、tip_to_axis_m=0.00097、axial_t=0.2118，与本次评分器末态字段一致。此处用于核对交付对象，未将 agent 工作时间换算为视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "双臂协作",
+      "how": "抓握",
+      "note_zh": "观察：视频 3.0–4.8 s，机器人左臂二指夹爪夹住蓝色套筒，右臂二指夹爪夹住红色插销，分别抬离桌面；视频 9.56–13.32 s 左臂持套筒、右臂持插销送入并保持，两爪均以对向夹持方式抓握各自物体。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.00097
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": 0.2118
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": 0.2118
+  }
+ ],
+ "failure": null,
+ "summary_zh": "机器人双臂分别抓握并抬起套筒与插销，视频约 10.0 s 完成插入，保持至末帧视频 13.32 s，最终完成 1/1 个节点。评分器确认末态轴距为 0.00097 米、轴向参数为 0.2118，交付文件独立重放成功（log step 8），无失败断点。无逐帧状态日志，腕部视图后段有夹爪遮挡，毫米级验收依赖评分器原生读数，首次达成时间为视觉近似值。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "peg_inserted 在视频约10 s 达成并保持到末帧，与评分器读数（inserted=1，tip_to_axis 0.00097，axial_t 0.2118）及帧一致；接触双臂协作+抓握正确，无失败断点。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:14:21Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--tzyml4h/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-privileged--tzyml4h/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "fd6cfde9e47e3bdaa06a61c1c31b2d182f0390584284c7d090c859d8fbe0a270",
+   "video_sha256": "6f57f1a6c194e1f08a250842d4f463ce68ceb163b9036726f833eb86730f8296",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--8g8mjwh.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--8g8mjwh.json
new file mode 100644
index 0000000000000000000000000000000000000000..da9a3bb2f52abe53f6dc32996d3cdbf0592c4453
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--8g8mjwh.json
@@ -0,0 +1,238 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-standard--8g8mjwh",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "finalized": 1,
+   "control_steps": 5910,
+   "calls": 127,
+   "elapsed_s": 986.8,
+   "video_rendered": 1,
+   "infra_error": 0,
+   "grader_error": 0,
+   "no_verdict": 0,
+   "sim_error": 0,
+   "resets": 0,
+   "reset_budget": 0,
+   "rows": 880,
+   "done": 1,
+   "video_done": 1,
+   "episodes": 1,
+   "control_steps_total": 5910,
+   "live_success": 0,
+   "inserted": 0,
+   "tip_to_axis_m": 0.00029,
+   "axial_t": 2.4482,
+   "socket_z": 0.0228,
+   "peg_z": 0.0231,
+   "time_s": 118.2,
+   "video": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-standard--8g8mjwh/replay.mp4",
+  "duration_s": 26.32,
+  "fps": 25.0,
+  "frames": 658,
+  "sha256": "2576ab6a9fdc21f469b8fa8431920b648bfe9ed1a02384554b7a6338885c9126"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0 s，蓝色套筒与红色插销分开放在桌面上，尚无插入关系。"
+     },
+     {
+      "t": 4.6,
+      "log_step": null,
+      "text_zh": "观察：视频 2.5 s 双臂分别夹持并抬起蓝色套筒和红色插销；视频 4.4–5.2 s 两件物体靠拢，红条靠套筒的一端被蓝管遮住，随后腕部视图被机器人遮挡。该外观只能证明一端进入，不能辨认命名站点 peg_end2。"
+     },
+     {
+      "t": 8.4,
+      "log_step": null,
+      "text_zh": "观察：视频 8.2–8.4 s 红色插销从右夹爪下方落到桌面；视频 10.5–15.0 s 机器人重新抓取并抬起插销，视频 15.4 s 蓝色套筒也落到桌面。"
+     },
+     {
+      "t": 21.5,
+      "log_step": null,
+      "text_zh": "观察：视频 17.5–19.25 s 右夹爪持红色插销向桌上的蓝色套筒推进，视频 19.5–21.0 s 右夹爪松开后以闭合指端从外端推红条，视频 21.5 s 再夹住红条；左夹爪此时悬空。"
+     },
+     {
+      "t": 26.28,
+      "log_step": 67,
+      "text_zh": "观察：视频 24.5–26.28 s 右夹爪仍夹持红色插销，蓝管遮住其靠套筒的一段，外端仍明显伸出；成功参考视频 12.0–14.08 s 也有相似遮挡外观，因此不能凭露出长度给本次受检端点计分。评分器末态 inserted=0、tip_to_axis_m=0.00029 m、axial_t=2.4482，轴距合格但受检端点远在开口之外，超过轴向上界 0.75。"
+     },
+     {
+      "t": 26.28,
+      "log_step": 55,
+      "text_zh": "推断：log step 55 的视觉脚本以 xend-0.096 作为插入端，log step 67 承认假定朝套筒的负 x 端就是 peg_end2；该估计的轴向参数约 0.37 与评分器的 2.4482 不同，支持实际进入的是另一端。视频全过程未见机器人将插销掉头或改从另一端插入，故不认定存在满足指定端点判据的区间；末态为假由原生几何读数直接决定。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "双臂协作",
+      "how": "抓握",
+      "note_zh": "视频约 2.5–8.2 s，左二指夹爪夹持蓝色套筒，右二指夹爪夹持红色插销，抬起并相向靠拢尝试插入；随后插销掉落。"
+     },
+     {
+      "who": "手指/夹爪",
+      "how": "抓握",
+      "note_zh": "视频约 14.0 s 右夹爪重新夹住红色插销；蓝色套筒于视频约 15.4 s 掉落后由桌面承托，右夹爪单独搬运、推进和调整插销，末态仍夹持红条。"
+     },
+     {
+      "who": "手指/夹爪",
+      "how": "推/扫",
+      "note_zh": "视频约 19.5–21.0 s，右夹爪先松开红色插销，再用闭合的指端顶住红条露出的外端向桌上套筒方向推；此阶段不是夹住红条搬运。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.00029
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": 2.4482
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": 2.4482
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "peg_inserted",
+   "t": 4.6,
+   "log_step": null,
+   "where": "操作中",
+   "evidence_zh": "推断：视频约 4.6 s 双臂首次将红条一端送进蓝管，后来掉落、重新抓取并改用桌面插入，但直到视频 26.28 s 都未见掉头换端。末态评分器 axial_t=2.4482>0.75，指定端点仍在套筒开口外；首次插入方向的端点偏差没有得到纠正。"
+  },
+  "unmet": {
+   "peg_inserted": false
+  },
+  "phenomena": [
+   {
+    "cat": "运动",
+    "item": "位姿不准",
+    "t": 24.5,
+    "log_step": null,
+    "evidence_zh": "推断：视频 24.5–26.28 s 红色插销一段进入蓝管，另一端仍露在套筒外，机器人保持该姿态结束；评分器记录受检端点距轴 0.00029 m、轴向参数 2.4482，说明该端点虽对轴却没有进入套筒规定深度。结合 log step 55、67 的端点假设，实际进入的一端与受检端点不一致。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "认知",
+    "item": "感知",
+    "log_step": 55,
+    "source_zh": "日志 / 评分器字段",
+    "evidence_zh": "推断：log step 55 的脚本把可见红条最大 x 坐标减去 0.096 当作 tip，log step 62 得到该点约 x=-0.0742；log step 67 明确说未确认哪一端是 peg_end2，只假定负 x 端是受检端。评分器末态却给出 axial_t=2.4482，而 agent 在 log step 67 估计约 0.37；视频 24.5–26.28 s 末态仍有红条外端露出，支持受检端点识别错误，导致控制器围绕错误端点调整。"
+   },
+   {
+    "cat": "认知",
+    "item": "误以为完成",
+    "log_step": 67,
+    "source_zh": "agent 自述 / 日志",
+    "evidence_zh": "推断：agent 在 log step 67 自述插销已插入、视觉估计符合评分条件，并明确说已停止发送命令；虽承认没有成功信号及端点身份风险，仍以该估计接受末态。log step 65 的服务 time_left_s 仍为 878.8，任务预算为 1800 秒，说明停止并非预算耗尽；评分器 inserted=0 与该完成判断矛盾。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志或沿程原生 inserted/轴向读数；视频 5.4–10.0 s 腕部视图受遮挡，且图像不能直接辨认 peg_end2 站点或精确核验毫米阈值。因此全过程未达成及插错端的判断结合了可见运动方向、脚本端点假设和末态评分器，瞬时是否擦过阈值不能完全排除；没有将 agent 自估的插入深度当作节点达成证据。"
+ },
+ "summary_zh": "机器人先双臂持件尝试空中插入，掉落后重新抓取插销并改在桌面推进、调整，最终插入节点为 0/1。红条虽有一端进入蓝管，评分器受检端点轴向参数仍为 2.4482，未进入允许范围；日志支持识别错端并据错误完成估计停止。缺少逐帧几何状态，端点身份及全过程未达成包含推断，末态未达成有原生读数直接支持。",
+ "conf": 7,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "单节点 peg_inserted 末态为假由评分器 inserted=0、axial_t=2.4482 直接决定，视频帧（红条始终从外端伸出、无掉头）与之相符；接触标注、断点、位姿不准、感知（log 55/62/67 端点假设，估计 0.37 vs 评分器 2.45，且红条约 96 mm 长，几何上吻合插错端）与误以为完成均有证据支持，时钟与引用均无问题。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:14:00Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--8g8mjwh/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--8g8mjwh/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "842e31ddf92072288cce4a46683a5ad87ca7d853eef45f1e302a6293f2219db2",
+   "video_sha256": "2576ab6a9fdc21f469b8fa8431920b648bfe9ed1a02384554b7a6338885c9126",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--8qb9v4r.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--8qb9v4r.json
new file mode 100644
index 0000000000000000000000000000000000000000..197410702a59bd1c74b0133ddbeeace07cbb6a40
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--8qb9v4r.json
@@ -0,0 +1,186 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-standard--8qb9v4r",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "finalized": 1,
+   "control_steps": 550,
+   "calls": 3,
+   "elapsed_s": 1416.9,
+   "video_rendered": 1,
+   "infra_error": 0,
+   "grader_error": 0,
+   "no_verdict": 0,
+   "sim_error": 0,
+   "resets": 0,
+   "reset_budget": 0,
+   "rows": 550,
+   "done": 1,
+   "video_done": 1,
+   "episodes": 1,
+   "control_steps_total": 550,
+   "live_success": 1,
+   "inserted": 1,
+   "tip_to_axis_m": 0.00057,
+   "axial_t": 0.0745,
+   "socket_z": 0.1675,
+   "peg_z": 0.1665,
+   "time_s": 11.0,
+   "video": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-standard--8qb9v4r/replay.mp4",
+  "duration_s": 11.08,
+  "fps": 25.0,
+  "frames": 277,
+  "sha256": "749b4c2cb11fe0294dff94fbc4b4b05cb705a50364b7919e26b637a49458389d"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [
+     [
+      9.2,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频开头红色插销与蓝色套筒分开放在桌面上，插入关系尚未成立。"
+     },
+     {
+      "t": 4.8,
+      "log_step": 44,
+      "text_zh": "观察：视频 3.2–4.0 s 两个二指夹爪分别闭合夹住蓝色套筒和红色插销；视频 4.4–5.6 s 两件物体随各自夹爪离开桌面，原放置处变空，随后机器人将二者移到桌面上方相对的位置。"
+     },
+     {
+      "t": 9.2,
+      "log_step": 47,
+      "text_zh": "推断：对照视频 7.8–8.2 s 的开口外对准与视频 8.4–9.4 s 的连续推进，腕部视图中红色插销逐渐进入蓝色套筒，露出长度缩短，至视频 9.4 s 插入端被蓝管遮住；将达到规定插入深度的起点估为视频 9.2 s，误差约 0.3 s。无逐帧状态日志，不能由像素精确确定毫米级阈值越界时刻。"
+     },
+     {
+      "t": 11.04,
+      "log_step": null,
+      "text_zh": "观察：视频 10.0、10.2、10.4、10.6、10.8、11.0 和 11.04 s，机器人双爪仍各持一件，红色插销保持在蓝色套筒内，未见退回或掉落；评分器 reward 报告 success=1、inserted=1、tip_to_axis_m=0.00057、axial_t=0.0745，末态满足距轴线小于 0.005 m 且轴向参数位于 [-0.875,0.75] 的验收。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "双臂协作",
+      "how": "抓握",
+      "note_zh": "观察：视频 3.2–5.6 s 左臂二指夹爪夹住蓝色套筒、右臂二指夹爪夹住红色插销并将两者抬起；视频 8.4–10.0 s 左爪持稳套筒、右爪持销推进完成插入，双爪保持抓握至视频 11.04 s。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.00057
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": 0.0745
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": 0.0745
+  }
+ ],
+ "failure": null,
+ "summary_zh": "机器人双爪分别抓握并抬起套筒与插销，对准后推进，约在视频 9.2 s 达成插入，保持至视频 11.04 s，最终完成 1/1 个节点。评分器确认插入成功，末态距轴线 0.57 毫米、轴向参数 0.0745，均满足验收，无失败断点。没有状态日志，首次达标时刻由视频估计，误差约 0.3 s；末态毫米级几何关系由评分器读数佐证。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标注与视频一致：双爪各夹套筒与插销抬起、对准后约视频 9.2 s 插入，至 11.04 s 保持，评分器读数佐证 final=true；contact_actual 双臂协作+抓握正确，无断点，校验通过。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:20:20Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--8qb9v4r/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--8qb9v4r/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "f832ce33190a2ec78504eba58835cc29f067bcf972da69999e5f0abe35658191",
+   "video_sha256": "749b4c2cb11fe0294dff94fbc4b4b05cb705a50364b7919e26b637a49458389d",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--aabtuh3.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--aabtuh3.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a6b62e0ca66c3ad1e417dd6f4f821497958e642
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--aabtuh3.json
@@ -0,0 +1,229 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-standard--aabtuh3",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "graded_by_live": 1,
+   "replays": 0,
+   "finalized": 1,
+   "missing_log": 0,
+   "infra_error": 0,
+   "sim_error": 0,
+   "grader_error": 0,
+   "control_steps": 3820,
+   "calls": 76,
+   "elapsed_s": 1827.8,
+   "video_rendered": 1,
+   "no_verdict": 0,
+   "instance_mismatch": 0,
+   "gave_up": 0,
+   "sim_crashed_on_call": 0,
+   "resets": 0,
+   "reset_budget": 0,
+   "rows": 62,
+   "done": 1,
+   "video_done": 1,
+   "episodes": 1,
+   "control_steps_total": 3820,
+   "n_actions": 3820,
+   "video": 1,
+   "video_pending": 0,
+   "inserted": 0,
+   "tip_to_axis_m": 0.03436,
+   "axial_t": -0.58,
+   "socket_z": 0.0237,
+   "peg_z": 0.0112,
+   "time_s": 76.4,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-standard--aabtuh3/replay.mp4",
+  "duration_s": 25.52,
+  "fps": 25.0,
+  "frames": 638,
+  "sha256": "e92e064bf5d3ff1d0800efab15d6326345af90282c9a47d66de6e3a60f3df4e3"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0 s 四视图中红色插销与蓝色套筒分离。视频 0.0–25.48 s 的已查帧中，机器人先抓取套筒，再反复夹取红销，后将两件移近；未见红销端部沿开口进入内腔的可见特征。"
+     },
+     {
+      "t": 12.48,
+      "log_step": 118,
+      "text_zh": "观察：视频 12.16–12.32 s 红销随右夹爪离桌，视频 12.40–12.80 s 右夹爪继续抬升而红销脱离夹爪落回桌面，蓝色套筒仍在另一夹爪中；log step 118 的动作保持右夹爪闭合并要求继续抬升红销。"
+     },
+     {
+      "t": 21.6,
+      "log_step": null,
+      "text_zh": "观察：视频 20.4–22.0 s 腕部视图显示红销与蓝管并排，红销端部在蓝管外侧，未从管口进入。成功参考录像视频 8.0–12.0 s 的相同腕部视图可见红销一端被蓝管遮住、露出段缩短，仅用于比较达成特征。"
+     },
+     {
+      "t": 25.48,
+      "log_step": null,
+      "text_zh": "观察：视频 25.48 s 末帧腕部视图显示红销仍沿蓝管外侧，开口处未形成插入关系；本次评分器 reward 报告 inserted=0、tip_to_axis_m=0.03436、axial_t=-0.58。末态轴向参数在允许范围，但距轴线 34.36 毫米超过严格小于 5 毫米的阈值，故节点末态为假。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "双臂协作",
+      "how": "抓握",
+      "note_zh": "观察：视频 6.0–12.32 s 左侧机械臂二指夹爪夹持蓝色套筒，右侧机械臂夹爪反复闭合夹取红销；视频 12.16–12.32 s 红销短暂随右夹爪离桌，随后滑脱。两臂各自夹持零件参与同一插入步骤。"
+     },
+     {
+      "who": "手指/夹爪",
+      "how": "推/扫",
+      "note_zh": "观察：视频 17.0–22.0 s 右夹爪接近桌面推动红销，使其向蓝色套筒靠近；红销在桌面上移位，并未保持为稳定的悬空抓持。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.03436
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": -0.58
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": -0.58
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "peg_inserted",
+   "t": 12.48,
+   "log_step": 118,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 12.40–12.80 s，机器人右夹爪抬升时红销脱离并落回桌面，插入操作中的持物失败；随后机器人重新接近、推移并调整两件，视频 20.4–22.0 s 及视频 25.48 s 仍显示红销位于蓝管外侧，直到结束也没有完成插入。"
+  },
+  "unmet": {
+   "peg_inserted": false
+  },
+  "phenomena": [
+   {
+    "cat": "接触",
+    "item": "没保持住",
+    "t": 12.48,
+    "log_step": 118,
+    "evidence_zh": "观察：视频 12.16–12.32 s 红销被右夹爪夹起，视频 12.40–12.80 s 夹爪继续上移而红销滑脱落回桌面；log step 118 的右夹爪目标仍为闭合值 0.002，没有计划松手。"
+   },
+   {
+    "cat": "运动",
+    "item": "位姿不准",
+    "t": 21.6,
+    "log_step": null,
+    "evidence_zh": "观察：视频 20.4–22.0 s，机器人把红销移至套筒旁，但红销与蓝管并排、端部偏在管口外侧；视频 25.48 s 仍是外侧错开关系。评分器末态 tip_to_axis_m=0.03436，支持最终横向偏差超限。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：log step 112、118 连续要求闭爪抬升红销，视频 12.16–12.80 s 却显示夹持失效、红销滑落；log step 168、171、173 转为朝套筒推动红销，视频 20.4–22.0 s 及视频 25.48 s 仍显示两件并排错开。机器人未能实现稳定持物与相对位姿控制，因此归于控制；现有证据不能再区分夹持几何、接触力或轨迹跟踪问题。",
+    "source_zh": "视频 / 日志"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志或接触力数据，无法精确判定夹持滑脱的力学原因，也不能从像素测量全过程的毫米级轴距；部分后段腕部画面被机械臂遮挡。末态由视频 25.48 s 的清晰外侧关系及评分器几何读数共同确认，未将日志中的抓取动作名称当作已抓稳的证据。"
+ },
+ "summary_zh": "机器人夹持蓝色套筒并反复夹取红销，但视频 12.40–12.80 s 红销在抬升中滑脱，随后推移和对准也未形成插入，唯一节点全程未见达成、末态为 0/1。视频 25.48 s 红销仍位于蓝管外侧，评分器端点距轴线 34.36 毫米，失败表现为没保持住及位姿不准，根因归于控制。缺少逐帧状态及接触力数据，无法进一步确定控制子类或精确测量中途几何误差。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 6,
+    "summary_zh": "唯一节点 peg_inserted 末态为假，与视频 25.4 s 腕部视图（红销在蓝管外侧）及评分器 inserted=0、tip_to_axis_m=0.03436 一致；断点（视频约 12.5 s 红销滑脱）、现象和根因「控制」均有证据支持，未发现需要改动的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:21:36Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--aabtuh3/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--aabtuh3/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "a49aae85c8ab26df9844c7889fecdb2126e5052410a19c0108c81bb4f0f28334",
+   "video_sha256": "e92e064bf5d3ff1d0800efab15d6326345af90282c9a47d66de6e3a60f3df4e3",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--cur5s7c.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--cur5s7c.json
new file mode 100644
index 0000000000000000000000000000000000000000..d0bd142bed6a113fe0696b483c10d033848e9542
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--cur5s7c.json
@@ -0,0 +1,203 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-standard--cur5s7c",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "finalized": 1,
+   "control_steps": 5006,
+   "calls": 108,
+   "elapsed_s": 3084.2,
+   "video_rendered": 1,
+   "infra_error": 0,
+   "grader_error": 0,
+   "no_verdict": 0,
+   "sim_error": 0,
+   "rows": 5006,
+   "done": 1,
+   "video_done": 1,
+   "live_success": 0,
+   "inserted": 0,
+   "tip_to_axis_m": 0.01083,
+   "axial_t": 1.0123,
+   "socket_z": 0.0228,
+   "peg_z": 0.0137,
+   "time_s": 100.12,
+   "video": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-standard--cur5s7c/replay.mp4",
+  "duration_s": 25.08,
+  "fps": 25.0,
+  "frames": 627,
+  "sha256": "d2bfe376dc66f7d5551dbca06be575f5c8e045419f7d6269c4b030672a0fa9d6"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0 s 四视图中，红色插销与蓝色套筒分离，均在桌面上；初态不满足插入关系。"
+     },
+     {
+      "t": 21.5,
+      "log_step": null,
+      "text_zh": "观察：视频 0–20.75 s 的全程粗看及视频 8–25.04 s 的加密帧中，机器人先多次向红销合爪、抬升和重新抓取，随后于视频约 20.75–21.5 s 把红销移至蓝管口；此前红销与蓝管分开，之后红销仍大段外露，后续修正未出现成功参考视频 8–14.08 s 中红条深入蓝管、露出长度明显缩短的特征。"
+     },
+     {
+      "t": 25.04,
+      "log_step": 219,
+      "text_zh": "观察：视频 24.25、24.5、24.75、25.04 s 中，机器人右夹爪仍夹着红销停在蓝管口附近，红条未深入至成功参考的可见位置；inputs/run.json 的评分器末态字段 inserted=0、tip_to_axis_m=0.01083、axial_t=1.0123，分别不满足轴距严格小于 0.005 m 及轴向参数不大于 0.75 的要求。log step 219 的机器人服务末态状态为 control_steps=5006、sim_time_s=100.12；这里视频时间取自所看帧，未用 agent 工作时间换算。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "手指/夹爪",
+      "how": "抓握",
+      "note_zh": "观察：视频约 20.75–21.5 s，右臂二指夹爪夹持红色插销，将其从原位置移至蓝色套筒入口附近，随后至视频 25.04 s 保持夹持并尝试推送、修正；蓝色套筒留在桌面，另一夹爪未抓住套筒。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.01083
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": 1.0123
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": 1.0123
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "peg_inserted",
+   "t": 21.5,
+   "log_step": null,
+   "where": "操作中",
+   "evidence_zh": "观察：视频约 21.5 s，机器人右夹爪持红销靠向蓝管入口，之后反复推进和微调至视频 25.04 s，仍未形成成功参考中红条深入管内的关系；评分器末态 inserted=0、轴距 0.01083 m、轴向参数 1.0123，插入操作始终未完成。"
+  },
+  "unmet": {
+   "peg_inserted": false
+  },
+  "phenomena": [
+   {
+    "cat": "运动",
+    "item": "位姿不准",
+    "t": 21.5,
+    "log_step": null,
+    "evidence_zh": "观察：视频 21.2–25.04 s，机器人已将红销带到蓝管口，但红条在右腕视图中仍几乎整段外露，推送和侧向微调后也未出现成功参考视频 8–14.08 s 中端部被蓝管明显遮住的深入关系。评分器给出的末态轴距 10.83 毫米、轴向参数 1.0123 确认最终既偏离轴线又深度不足。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "轨迹或逆解",
+    "t": null,
+    "log_step": 191,
+    "source_zh": "日志",
+    "evidence_zh": "推断：agent 的插入控制未找到并执行满足目标的臂位姿轨迹。log step 158 的运动学程序已打印多次 IK miss；log step 174 指定腕相机末端高度 0.163 m，执行后的估计高度为 0.1504 m。log step 191 读取右臂关节并计算限位裕量，腕角约 -1.869 rad、裕量约 -0.001 rad，且其旋转搜索中最佳位置残差仍为 0.02044 m；log step 202 指令沿负 x 推进并向上抬，执行后腕相机估计位置从 [-0.0018,0.0588,0.1478] m 变为 [-0.0069,0.0735,0.1448] m，推进很小且高度反降。结合视频 21.5–25.04 s 插销停留在入口附近，这些诊断支持逆解及受限轨迹执行问题。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志或全过程原生几何读数，无法用像素精确量出中途毫米级轴距和深度；节点区间依据四视图可见关系及成功参考判断，末态由评分器几何字段明确判定。根因中的腕位姿和逆解残差来自 agent 自建运动学程序的诊断，缺少独立关节轨迹和接触力记录，无法进一步分离限位、跟踪偏差与接触影响。"
+ },
+ "summary_zh": "机器人多次重新抓取后，以右侧夹爪将红销送到蓝色套筒口，但未完成插入，节点末态为 0/1。视频约 21.5–25.04 s 的推进与微调始终未修正插入位姿，评分器末态轴距 10.83 毫米、轴向参数 1.0123；日志中的腕关节限位与逆解残差支持轨迹或逆解问题。缺少逐帧状态和接触力记录，根因细分仍有上述诊断依赖。",
+ "conf": 7,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 6,
+    "summary_zh": "节点 peg_inserted 末态 false、无区间、抓握接触、断点 peg_inserted(21.5 s 附近，操作中) 与现象“位姿不准”都与视频帧和评分器字段(inserted=0, tip_to_axis 0.01083, axial_t 1.0123)一致；根因“轨迹或逆解”已标为推断，log step 158/174/191/202 的内容与引用相符，未发现需要修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:21:35Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--cur5s7c/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--cur5s7c/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "8b4ad35ebacd8fe49f2a9c033fcaef538f11038c6f02e5f7bb68a5ec03403b39",
+   "video_sha256": "d2bfe376dc66f7d5551dbca06be575f5c8e045419f7d6269c4b030672a0fa9d6",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--ekz8zex.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--ekz8zex.json
new file mode 100644
index 0000000000000000000000000000000000000000..8c3ee9eeef6cf7210e3661ff1d172dd9f5a48b42
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--ekz8zex.json
@@ -0,0 +1,223 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-standard--ekz8zex",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "graded_by_live": 1,
+   "replays": 0,
+   "finalized": 1,
+   "missing_log": 0,
+   "infra_error": 0,
+   "sim_error": 0,
+   "grader_error": 0,
+   "control_steps": 7231,
+   "calls": 166,
+   "elapsed_s": 1686.8,
+   "video_rendered": 1,
+   "no_verdict": 0,
+   "instance_mismatch": 0,
+   "gave_up": 0,
+   "sim_crashed_on_call": 0,
+   "resets": 0,
+   "reset_budget": 0,
+   "rows": 7231,
+   "done": 1,
+   "video_done": 1,
+   "episodes": 1,
+   "control_steps_total": 7231,
+   "n_actions": 7231,
+   "video": 1,
+   "video_pending": 0,
+   "inserted": 0,
+   "tip_to_axis_m": 0.03181,
+   "axial_t": 0.3314,
+   "socket_z": 0.0363,
+   "peg_z": 0.0108,
+   "time_s": 144.62,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.307,
+   "peak_processes": 8,
+   "peak_disk_gb": 0.521,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-standard--ekz8zex/replay.mp4",
+  "duration_s": 24.16,
+  "fps": 25.0,
+  "frames": 604,
+  "sha256": "bf897ddcb4acce78e14cdc7f71ac37a4cc8044b77dabfb9960bda4b484e3601d"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0 s 四视图中红色插销与蓝色套筒分开放在桌面，初态没有插入关系。"
+     },
+     {
+      "t": 7.8,
+      "log_step": null,
+      "text_zh": "观察：视频 7.4–8.8 s 机器人两爪各夹一件并靠近，腕部视图中的红条与蓝管存在侧向及高度错位，红条露在管外；随后视频 9.2–18.4 s 的多次调整仍可见红条沿蓝管外侧伸出，没有成功参考视频末段那种从开口进入、露出长度缩短的关系。"
+     },
+     {
+      "t": 20.8,
+      "log_step": null,
+      "text_zh": "观察：视频 18.6–21.2 s 机器人再次调整两臂，红条留在桌面附近，蓝管转向并移到其旁边；视频 21.4–24.12 s 两臂保持最后姿态，右腕视图的夹爪前没有夹住红条，俯视及正视仍显示红条偏在蓝管外。"
+     },
+     {
+      "t": 24.12,
+      "log_step": null,
+      "text_zh": "观察：末帧视频 24.12 s 未见插销端从套筒开口进入；inputs/run.json 的在线末态评分字段 inserted=0、tip_to_axis_m=0.03181（米），大于严格小于 0.005 米的要求，axial_t=0.3314 则在允许范围内。该节点末态为假，不能从轴向参数单项合格认定插入。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "双臂协作",
+      "how": "抓握",
+      "note_zh": "观察：视频 7.4–8.8 s 左侧二指夹爪夹持蓝色套筒，右侧二指夹爪夹持红色插销，两臂共同移动并尝试对准送入；此前先后取物，后续又反复重抓、调整。视频 24.12 s 右爪前已没有夹住红条。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.03181
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": 0.3314
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": 0.3314
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "peg_inserted",
+   "t": 7.8,
+   "log_step": null,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 7.4–8.8 s 机器人正在用两爪将红色插销送向蓝色套筒，但红条与开口错位，仍在管外；之后多轮调整直至视频 24.12 s 都未完成该插入步骤。末态评分 inserted=0、端点轴距 31.81 毫米，确认这一步最终没有修复。"
+  },
+  "unmet": {
+   "peg_inserted": false
+  },
+  "phenomena": [
+   {
+    "cat": "运动",
+    "item": "位姿不准",
+    "t": 7.8,
+    "log_step": null,
+    "evidence_zh": "观察：视频 7.8 s 两腕视图显示红色插销与蓝色套筒没有同轴，红条贴近管外而非从开口进入；视频 12.4–18.4 s 红条仍偏在蓝管一侧，视频 24.12 s 最终红条与套筒也未形成插入关系。末态评分 tip_to_axis_m=0.03181 米，轴距超出 5 毫米要求。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "视频 / 日志",
+    "evidence_zh": "推断：机器人在视频 7.4–8.8 s、9.8–18.4 s 反复执行靠近和送入，但没有把两件物体的相对位姿控制到插入所需精度。log step 145、304、323、328 的脚本明确以对准并插入为目标、执行关节插值轨迹，视频中实际物体仍在管外，说明问题发生在插入的具体操控执行；缺少逐帧物体姿态及接触记录，控制子类不能确定。"
+   },
+   {
+    "cat": "认知",
+    "item": "误以为完成",
+    "source_zh": "日志 / 评分器字段 / 视频",
+    "evidence_zh": "推断：agent 的错误完成判断促使它保留失败姿态并停止纠偏。log step 334 将控制器改为仅保持当前关节角，称其为 winning pose；log step 337 实际执行保持，log step 339 最终声称已核验轴距小于 1.5 毫米且插入成功，随后结束，而 log step 338 的资源记录仍有 124.9 秒预算。视频 23.04–24.12 s 保持姿态未形成插入，在线最终评分却为 inserted=0、轴距 31.81 毫米，否定其完成判断。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志或接触力记录；多视图足以判断可见的错位和末态失败，但无法精确测量全过程的毫米级轴距、插深，亦无法确定具体控制偏差来自持物相对位姿变化、轨迹求解还是跟踪。节点区间按已检查画面判定，末态使用原生评分字段；agent 自算的点云指标及最终成功声明不当作评分器读数，也未把 agent 时间换算成视频时间。"
+ },
+ "summary_zh": "机器人曾用双臂分别抓握蓝色套筒和红色插销并多次尝试对准，唯一的插入节点未见达成，末态为 0/1。断点是插入操作中的位姿错位：最终轴距 31.81 毫米超限，agent 却误判成功并保持该姿态后结束。无逐帧状态和接触记录，具体控制子类及全过程毫米级几何无法精确核验。",
+ "conf": 7,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 6,
+    "summary_zh": "节点 peg_inserted 末态为假、末态评分 inserted=0 与视频一致；断点、位姿不准现象及两个根因有视频和日志依据，时钟未换算，未发现需要修改的内容。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:10:57Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--ekz8zex/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--ekz8zex/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "ada6ef14f1237ffb3adbdbe200d53830e9b361fc6a00e44e725da2c1baca2a43",
+   "video_sha256": "bf897ddcb4acce78e14cdc7f71ac37a4cc8044b77dabfb9960bda4b484e3601d",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--m4vtq8s.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--m4vtq8s.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c7bd0a1b2f26b57208cfd2e7130e61f1296a7cf
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--m4vtq8s.json
@@ -0,0 +1,192 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-standard--m4vtq8s",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "finalized": 1,
+   "control_steps": 12605,
+   "calls": 891,
+   "elapsed_s": 1695.2,
+   "video_rendered": 1,
+   "grader_error": 0,
+   "sim_error": 0,
+   "rows": 2461,
+   "done": 1,
+   "live_success": 0,
+   "inserted": 0,
+   "tip_to_axis_m": 0.08908,
+   "axial_t": 1.6346,
+   "socket_z": 0.0228,
+   "peg_z": 0.0108,
+   "time_s": 252.1,
+   "video": 1
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-standard--m4vtq8s/replay.mp4",
+  "duration_s": 24.08,
+  "fps": 25.0,
+  "frames": 602,
+  "sha256": "087f118f5adb54973076430e142c39ebd194dc0472227e65f8d9aae8f8fa529c"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0 s 的两幅外部视图显示红色插销和蓝色套筒分开放在桌面，初始没有插入。"
+     },
+     {
+      "t": 4.2,
+      "log_step": null,
+      "text_zh": "观察：视频 3.8–5.2 s，右侧机械臂的空夹爪从物体附近抬起并向外摆动，红色插销仍在桌面；视频 6.0–15.6 s 的反复末端运动也未带起插销，未出现红端进入蓝色套筒的关系。"
+     },
+     {
+      "t": 22.4,
+      "log_step": null,
+      "text_zh": "观察：视频 18.0–22.0 s，右夹爪在红色插销附近反复调整，随后视频 22.2–23.4 s 向桌边偏离；插销留在桌面，未随夹爪移动。"
+     },
+     {
+      "t": 24.04,
+      "log_step": null,
+      "text_zh": "观察：视频 23.2、23.4、23.6、23.8、24.04 s，外部视图一直显示两件物体分离，双臂最终悬停；没有参考成功视频 13.8 s 所见的红条一端深入蓝管并被遮住的特征。评分器末态字段 inserted=0、tip_to_axis_m=0.08908、axial_t=1.6346，与可见未插入一致。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "手指/夹爪",
+      "how": "抓握",
+      "note_zh": "观察：右臂二指夹爪在插销附近调整并尝试夹取红色插销（视频 18.0–22.4 s），但未见夹住或带离桌面；此处记录失败的抓取尝试，不表示已建立抓握。左夹爪未见抓起蓝色套筒，未形成双臂各持一件的协作。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "value": 0.08908,
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better"
+  },
+  {
+   "name": "axial_t",
+   "value": 1.6346,
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better"
+  },
+  {
+   "name": "axial_t",
+   "value": 1.6346,
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "peg_inserted",
+   "t": 4.2,
+   "log_step": null,
+   "where": "操作中",
+   "evidence_zh": "观察：视频约 4.2 s，右臂空夹爪抬离桌上物体区域，红色插销未随之拿起；之后虽多次恢复臂姿并重试，直到视频 24.04 s 仍未建立持销并插入的关系。该时刻是抓取阶段偏离的粗略导航锚点，不是由 agent 时钟换算；抓取属于 DAG 的 peg_inserted 操作阶段。"
+  },
+  "phenomena": [
+   {
+    "cat": "接触",
+    "item": "没建立接触",
+    "t": 4.2,
+    "log_step": null,
+    "evidence_zh": "观察：视频 3.8–5.2 s 的右夹爪抬离，以及视频 18.0–23.4 s 的再次靠近和偏离，均未使红色插销离开桌面或随夹爪运动；视频 24.04 s 红销与蓝套筒仍分离。推断：插入所需的持销抓握未建立，而非持住后滑落；不另将同一空夹事件重复标为位姿不准。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "source_zh": "日志",
+    "evidence_zh": "观察：单局在线会话的 log step 79 中，agent 脚本明确把右夹爪目标设在红销上方，自己的伺服误差输出从 0.2033 m 增至 0.3624 m，并在后续轮次仍不收敛；log step 113 再次接近时误差从 0.0596 m 增至 0.3023 m，脚本闭爪前后的红销估计位置不变。视频 4.2–5.2 s 和 22.2–23.4 s 也可见空夹爪向外偏离，物体仍留在桌面。推断：接近抓取目标的视觉反馈运动未收敛，导致所需抓握没有建立；这些日志数值用于解释控制失败，不作为精确物体状态或视频时间映射。"
+   }
+  ],
+  "gap_zh": "无状态日志，不能逐帧确认夹爪是否曾短暂擦碰物体，也不能从像素测出毫米级插入几何。视频已检查全程每约 0.2 s 的帧及视频 24.04 s 末帧；本次没有可见插入特征，末态另有评分器几何字段支持。日志仿真末态为仿真 252.1 s，而视频仅 24.08 s，未建立逐控制步到视频的映射，故不将日志动作与视频时刻一一对应。agent 在 log step 106、117 自述末端特征误跟到桌框、雅可比不可靠，但缺少可独立核验的逐帧估计真值，无法进一步拆分感知误差与雅可比辨识误差。",
+  "unmet": {
+   "peg_inserted": false
+  }
+ },
+ "summary_zh": "本次插入节点全程未达成，末态为 0/1；机器人反复调整夹爪，但未见红色插销被抓住并送入蓝色套筒，视频 24.04 s 两件物体仍分离。断点位于插入步骤的抓取操作阶段，现象是所需抓握未建立，日志支持接近目标的视觉伺服反复发散。无状态日志且缺少日志到视频的精确映射，不能进一步核验短暂擦碰或区分末端定位与雅可比估计误差。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 6,
+    "summary_zh": "peg_inserted 全程未达成（评分器 inserted=0、tip_to_axis 0.089 m）与视频一致；断点在插入步骤的抓取阶段，现象“没建立接触”与根因“反馈控制”有日志（id 79、113、117）和视频支持，时钟没有把 agent 时间换算成视频时间。未发现需要改内容的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:13:43Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--m4vtq8s/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--m4vtq8s/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "19e16ac6ace858addd9756e6ebc0a3ed826df55f03445a52c686c6cfd5db8f82",
+   "video_sha256": "087f118f5adb54973076430e142c39ebd194dc0472227e65f8d9aae8f8fa529c",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--yfzetnw.json b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--yfzetnw.json
new file mode 100644
index 0000000000000000000000000000000000000000..f631b79d11b10325726946772f8d147afe534a22
--- /dev/null
+++ b/phase2/mujoco-playground/aloha_single_peg_insertion/trajectories/mujoco_playground_aloha_single_peg_insertion-standard--yfzetnw.json
@@ -0,0 +1,234 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_aloha_single_peg_insertion-standard--yfzetnw",
+ "taskKey": "mujoco-playground/aloha_single_peg_insertion",
+ "dag_sha256": "9709e85780759f3204a58aae58bbc9f03fd320c1c8d74829d8d296fb394797b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "graded_by_live": 1,
+   "replays": 0,
+   "finalized": 1,
+   "missing_log": 0,
+   "infra_error": 0,
+   "sim_error": 0,
+   "grader_error": 0,
+   "control_steps": 1512,
+   "calls": 6,
+   "elapsed_s": 1819.1,
+   "video_rendered": 1,
+   "no_verdict": 0,
+   "instance_mismatch": 0,
+   "gave_up": 0,
+   "sim_crashed_on_call": 0,
+   "resets": 0,
+   "reset_budget": 0,
+   "rows": 1512,
+   "done": 1,
+   "video_done": 1,
+   "episodes": 1,
+   "control_steps_total": 1512,
+   "n_actions": 1512,
+   "video": 1,
+   "video_pending": 0,
+   "inserted": 0,
+   "tip_to_axis_m": 0.14157,
+   "axial_t": 8.9131,
+   "socket_z": 0.0228,
+   "peg_z": 0.0108,
+   "time_s": 30.24,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_aloha_single_peg_insertion-standard--yfzetnw/replay.mp4",
+  "duration_s": 30.32,
+  "fps": 25.0,
+  "frames": 758,
+  "sha256": "0e6773892d2aa868610ad664c1c8e41e827959baa5c43bfb12e4a041bdabfa65"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "peg_inserted": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0 s 红色插销与蓝色套筒分别位于桌面中央附近，插销未进入套筒。"
+     },
+     {
+      "t": 2.8,
+      "log_step": 30,
+      "text_zh": "观察：视频 0.2–3.2 s 左夹爪附近的蓝套筒被带动、翻动并移向桌面左后方；视频 12.8–13.2 s 右夹爪扫过红销，红销在桌面滑动转向。两者始终分开；视频全程的两臂运动未出现持红销向蓝管开口对准并送入的过程。log step 30 的实际提交代码为双臂随机关节扫描。"
+     },
+     {
+      "t": 30.28,
+      "log_step": null,
+      "text_zh": "观察：视频 29.2–30.28 s 两物体仍分离留在桌面，最后一帧蓝套筒在左后方、红销在较前方；未出现成功参考录像视频约 8–14 s 所示红条端部进入蓝管且露出长度缩短的特征。评分器末态 inserted=0、tip_to_axis_m=0.14157、axial_t=8.9131，轴距和深度均不满足本节点。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "手指/夹爪",
+      "how": "推/扫",
+      "note_zh": "观察：标定扫描中，视频约 2.4–3.2 s 左侧二指夹爪碰拨蓝色套筒，使其移向桌面左后方；视频约 12.8–13.2 s 右侧二指夹爪扫动红色插销，使其在桌面滑动并转向。这些是先后发生的单夹爪接触，未见双臂各持一件配合插入。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "tip_to_axis_m",
+   "unit": "m",
+   "threshold": 0.005,
+   "direction": "lower_better",
+   "value": 0.14157
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": -0.875,
+   "direction": "higher_better",
+   "value": 8.9131
+  },
+  {
+   "name": "axial_t",
+   "unit": "1",
+   "threshold": 0.75,
+   "direction": "lower_better",
+   "value": 8.9131
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "peg_inserted",
+   "t": null,
+   "log_step": 30,
+   "where": "首步前",
+   "evidence_zh": "观察：评分视频 0.0–30.28 s 两臂做广泛姿态扫描并偶然拨动物体，没有开始面向套筒开口的插销抓取、对准和送入操作。log step 30 的代码生成 54 组随机双臂关节姿态，分四批提交 476+476+476+84=1512 个控制步；评分器 control_steps=1512、resets=0，未记录后续操作提交。推断：该评分尝试止于控制标定，插入步骤的定向操作尚未启动，不能把扫描中的碰拨当作插入尝试；没有可定位的单一插入失败时刻。"
+  },
+  "unmet": {
+   "peg_inserted": false
+  },
+  "phenomena": [
+   {
+    "cat": "没尝试",
+    "item": "没尝试",
+    "t": null,
+    "log_step": 30,
+    "evidence_zh": "观察：视频 0.0–30.28 s 两臂反复摆动，蓝套筒与红销没有接合；log step 30 唯一实际运动脚本是随机关节扫描，后续 log step 31–40 仅做运动学拟合和代码修正。推断：机器人没有开始有目标的抓取对准插入操作，物体在扫描中被拨动不构成该操作的启动。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "认知",
+    "item": "推理/规划",
+    "source_zh": "日志",
+    "evidence_zh": "观察：inputs/run.json 的 log step 9–24 反复处理相机坐标、点云和标定约定，log step 25–27 转向腕相机方案；log step 28 中 agent 自述“Starting with the sweep first”，决定先拟合各臂运动学再执行任务。log step 30 的实际代码生成 54 组随机双臂关节姿态，四批提交 476+476+476+84=1512 个控制步；评分器 reward.control_steps=1512、episodes=1、resets=0，log step 31–40 只继续拟合及排错，没有后续运动提交。视频 0.0–30.28 s 两臂广泛扫动、先后拨动物体，未见持红销对准蓝套筒并送入。推断：agent 将相机标定和完整运动学拟合置于任务操作之前，反复排错后仍将评分回合内全部实际运动用于随机标定扫描，未转入目标操作；这是跨阶段策略与执行顺序的规划问题。此归因依据具体标定路线和动作提交内容，而非以时间耗尽或未交付本身作根因；没有证据证明某次面向插入的轨迹、逆解或跟踪控制失败。"
+   }
+  ],
+  "gap_zh": "观察：inputs/ 中有完整评分视频和运行日志，无逐帧状态日志；视频能确认两物体全程未插合及夹爪拨动，但视频约 0.2–1.4 s 左夹爪与蓝套筒的短时接触部分遮挡，不能确证是否曾形成对向夹持。log step 40 未给出最终完整运动学拟合结果，且没有提交面向抓取与插入的控制，无法评估该模型最终能否支持任务控制。"
+ },
+ "summary_zh": "评分轨迹仅包含双臂随机标定扫描，夹爪先后拨动蓝套筒和红销，插入节点从未达成，末态为 0/1，轴距 0.14157 米、轴向参数 8.9131 均超出验收范围。插入的定向操作尚未启动；日志中的相机标定、运动学拟合路线与实际扫描提交支持“认知／推理/规划”根因：反复标定排错后，评分回合内的全部运动仍用于扫描。无状态日志，初段蓝套筒是否曾短时被夹住无法确定；未提交任务控制，不能判断其控制能力。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 2,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "fix",
+    "conf": 7,
+    "summary_zh": "WHERE 标签（peg_inserted 未达成、contact_actual、首步前/没尝试）与视频一致，评分器 inserted=0 也支持。问题在根因：把「控制」归因于 agent 未把运动学建模落实为控制器，实质是预算耗尽/未交付的复述，证据不支持控制子类；日志显示的是策略选择（唯一的评分回合被用来做标定扫描）。",
+    "n_issues": 1,
+    "n_major": 1
+   },
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 2,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "peg_inserted 未达成、contact_actual、首步前/没尝试、认知·推理/规划根因均与视频、log step 28/30/31 及评分器 inserted=0 一致；上一轮的控制根因问题已修正。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:28:34Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--yfzetnw/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--yfzetnw/04_review_r1"
+   },
+   {
+    "stage": "reconcile",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--yfzetnw/05_reconcile_r1"
+   },
+   {
+    "stage": "traj_review",
+    "round": 2,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_aloha_single_peg_insertion-standard--yfzetnw/04_review_r2"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "8dbb36c0a866cbe0be157135ce25d2f6792dbd9ab9bfc2b0b133525cfa6db89d",
+   "video_sha256": "0e6773892d2aa868610ad664c1c8e41e827959baa5c43bfb12e4a041bdabfa65",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--b8k85uu.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--b8k85uu.json
new file mode 100644
index 0000000000000000000000000000000000000000..53c193cda5872dc0c1d99780253a3ae771f4a40e
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--b8k85uu.json
@@ -0,0 +1,198 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-privileged--b8k85uu",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.3178,
+   "worst_yaw_err": 0.7281,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 33,
+   "agent_resets": 85,
+   "agent_sim_steps": 171215,
+   "agent_replays": 9,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.474,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.241,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-privileged--b8k85uu/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "1e0c15e8a5d5f301beb2c9dcfeb52c73aa9a0a70dd6a3e37b14e7cfb36d136a7"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.000–10.143 s中，Barkour四足机器人在棋盘格地面上交替迈腿并移动，前半段机身航向有来回偏摆，视频5.143–10.143 s继续移动并逐渐转向；视频9.143–10.143 s逐帧检查仍见机身由腿支起，没有倒地。"
+     },
+     {
+      "t": 3.286,
+      "log_step": null,
+      "text_zh": "观察：视频2.500–4.500 s逐帧可见机器人一边迈步一边改变航向，随后又转回；成功参考视频同一前半段的机身朝向较稳定。参考仅用于识别行进和转向外观，不替代本次评分。"
+     },
+     {
+      "t": 10.143,
+      "log_step": 62,
+      "text_zh": "观察：log step 62对最终交付的trajectory.npz执行评分器回放，输出steps=1000、fell_at=null、tracking_ok=false；两段平移误差分别为0.2927和0.3178 m/s，偏航误差分别为0.5540和0.7281 rad/s。run.json评分字段也给出n_actions=1000、fell=0、deterministic=1及相同最差误差。"
+     },
+     {
+      "t": 10.143,
+      "log_step": 62,
+      "text_zh": "推断：完整日程虽已运行且机器人未倒地，但两段速度跟踪误差均超过平移0.2 m/s、偏航0.3 rad/s的验收阈值，因此“按指令行进”从未满足完整判据，final=false；不是仅凭success=0作此判断。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.143–10.143 s，Barkour四条腿的足端交替落在棋盘格地面，承重并推动躯干行进和转向；视频末秒仍由足腿支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.3178,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.7281,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 3.286,
+   "log_step": 62,
+   "where": "操作中",
+   "evidence_zh": "观察：视频2.500–4.500 s机器人正在连续迈步，机身航向明显偏摆后回转；该处取视频3.286 s为偏差的粗略导航锚点。log step 62最终交付回放表明第一段平移/偏航均值误差0.2927 m/s、0.5540 rad/s已经超标，第二段也未达标；机器人最终未倒地但整个按指令行进步骤始终未修复完成。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 3.286,
+    "log_step": 62,
+    "evidence_zh": "观察：视频2.500–4.500 s中机器人迈步时航向来回变化，视频5.143–10.143 s虽继续移动和转向，但与成功参考后半段相比转向进展较小；视频末秒仍保持足腿支撑。log step 62及评分器记录确认两段的平移与偏航速度均未跟上指令。这是未倒地的行进偏差。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 55,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 55的生成代码将当前指令与速度、偏航反馈叠加，生成周期足端扫动轨迹并以IK转换成关节目标后录制；log step 56保存的存活配置与log step 62最终交付回放的两段误差完全一致，仍有显著速度跟踪残差。结合视频2.500–4.500 s的航向偏摆，支持该步态控制未能实现所需平移和转向，而非未启动或途中摔倒；现有证据不能在接触设计、反馈控制及参数之间可靠细分。"
+   }
+  ],
+  "gap_zh": "没有逐控制步状态日志；视频能够确认迈步、转向及未倒地，精确速度均值采用与最终评分字段吻合的log step 62交付文件回放结果。不能定位均值误差首次超阈值的瞬间，视频3.286 s仅作可见偏摆的导航锚点，也不能进一步确认足端滑移、接触力分配或某个增益参数的独立因果作用。会话中其他配置的摔倒不属于本次交付回合。"
+ },
+ "summary_zh": "机器人运行完整1000步，视频末秒仍由足腿支撑，但两段平移和偏航速度误差均超标，唯一节点“按指令行进”最终为0/1。断点在该步骤操作中，现象为未倒地的步态跟踪偏差；日志支持步态控制未实现所需速度，不能可靠细分具体控制机制。无逐步状态日志，精确误差来自最终交付文件回放及评分字段，偏差开始时刻只能给粗略视频锚点。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "唯一节点“按指令行进”final=false、断点（操作中）、步态现象与控制根因均有视频和 log step 62 的评分器回放（steps=1000、fell_at=null、两段误差 0.2927/0.3178、0.554/0.7281，与 run.json 一致）支持；视频全程未倒地，与参考视频相比无矛盾。时钟、contact_actual（腿/躯干·支撑）和 native_metrics 均正确。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:03:32Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--b8k85uu/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--b8k85uu/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "c015a416e86a69cff78ae57f33b0063775aac6e232bde6c95dc1273762e374e6",
+   "video_sha256": "1e0c15e8a5d5f301beb2c9dcfeb52c73aa9a0a70dd6a3e37b14e7cfb36d136a7",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--flcfwwu.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--flcfwwu.json
new file mode 100644
index 0000000000000000000000000000000000000000..f4c0f3a463b4c9af286d200d6ca04b43e19cefd9
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--flcfwwu.json
@@ -0,0 +1,197 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-privileged--flcfwwu",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.3568,
+   "worst_yaw_err": 0.8218,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 85,
+   "agent_resets": 3778,
+   "agent_sim_steps": 1279530,
+   "agent_replays": 2,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.884,
+   "peak_processes": 18,
+   "peak_disk_gb": 0.981,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-privileged--flcfwwu/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "b9b42ccde9af38e61b0f2caa4a99065454b3d68a97f0ee6ca9092b9be94684d5"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.0 s起，Barkour四足机器人站在棋盘格地面上；视频0.5–5.0 s足端交替落地、机身平移且由腿支起。起始站立及前半段移动不满足节点要求的完整指令日程。"
+     },
+     {
+      "t": 5.5,
+      "log_step": null,
+      "text_zh": "观察：视频5.0、5.5、6.0及7.0 s，机器人继续迈步，机身仍大致保持原先侧向朝向；视频8.0–10.14 s才逐渐转成近端面朝相机，未见倒地。成功参考在其视频5.5–10.0 s呈现持续且大幅的转弯；参考仅用于辨认视角中的转向表现。"
+     },
+     {
+      "t": 10.14,
+      "log_step": 149,
+      "text_zh": "观察：log step 147保存最后选中的轨迹；log step 149由评分工具在新进程回放该文件，报告1000步、fell_at=null，第一段lin_err=0.0905 m/s、yaw_err=0.0945 rad/s，第二段lin_err=0.3568 m/s、yaw_err=0.8218 rad/s。inputs/run.json的reward给出相同的两项最差误差、n_actions=1000、fell=0和deterministic=1；视频10.14 s机器人仍由足端支撑。"
+     },
+     {
+      "t": 10.14,
+      "log_step": 149,
+      "text_zh": "推断：最终交付文件回放结果与评分器字段相符，且视频支持完整行进但转向不足的表现；第二段平面速度误差0.3568 m/s超过0.2 m/s、偏航角速度误差0.8218 rad/s超过0.3 rad/s，因此完整日程节点始终未达到，终值为false。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.5–10.14 s，Barkour四条腿的足端交替接触棋盘格地面，支撑躯干并在迈步中移动；视频末秒机身仍离地，未见躯干着地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.3568,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.8218,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 5.5,
+   "log_step": 149,
+   "where": "操作中",
+   "evidence_zh": "推断：log step 149对最终交付文件的回放表明第一段两项误差均达标，第二段两项均超阈值；视频5.5–7.0 s机器人换段后继续以接近原先朝向迈步，随后到视频10.14 s也未呈现所需的持续大幅转弯。断在按指令行进步骤，视频5.5 s是换段后表现的粗略导航锚点，不是由agent工作时间换算。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 5.5,
+    "log_step": 149,
+    "evidence_zh": "观察：视频5.5–7.0 s机器人迈步但朝向变化很小，视频8.0–10.14 s才缓慢转成近端面朝相机；其转弯幅度明显小于成功参考的对应阶段，全程及末秒仍有足端支撑。log step 149的最终文件回放确认第二段平移和偏航跟踪均未达标，属于未倒地的行进偏差。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：log step 143中的生成脚本把当前指令s.commands[t]、速度和角速度送入Trot.act并记录关节目标；log step 147进一步选择并保存了轨迹，log step 149回放该最终文件仍得到第二段lin_err=0.3568 m/s、yaw_err=0.8218 rad/s，与评分器一致。结合视频5.5–10.14 s持续迈步却未形成所需转弯，支持该步态控制方案未能实现第二段速度指令；没有足够证据细分为逆解、反馈或某个参数错误。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有逐控制步状态日志，无法确定第二段误差的逐时变化或归因到具体落足、逆解、反馈参数；视频只能支持运动及转向表现，精确误差采用评分器与最终文件回放记录。断点视频5.5 s为粗略锚点，不能视为首次瞬时超差时刻。"
+ },
+ "summary_zh": "机器人运行满1000个控制步，视频末秒仍由足端支撑，评分器也记录fell=0，但唯一的“按指令行进”节点未完成（0/1）。第一段跟踪达标，第二段转弯表现不足，平面速度及偏航误差分别为0.3568 m/s和0.8218 rad/s，断点在换段后的行进操作中，现象为步态，根因保留为控制。缺少逐步状态日志，不能进一步锁定控制器的具体失效机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成（终值false）、断点在换段后的行进、步态现象和控制根因均有log step 149评分器回放（第二段lin_err 0.3568、yaw_err 0.8218）与视频支持；时钟均为视频时间，无换算问题。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:02:32Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--flcfwwu/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--flcfwwu/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "967b33e9312f82c2c0c1757eb112383455a6b3efd57b5799ae7cb335f92af238",
+   "video_sha256": "b9b42ccde9af38e61b0f2caa4a99065454b3d68a97f0ee6ca9092b9be94684d5",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--gcdbhfc.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--gcdbhfc.json
new file mode 100644
index 0000000000000000000000000000000000000000..7908665a79b367e6efeaddd218ab0d402beded93
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--gcdbhfc.json
@@ -0,0 +1,202 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-privileged--gcdbhfc",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.3056,
+   "worst_yaw_err": 0.4985,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 227,
+   "agent_resets": 424,
+   "agent_sim_steps": 1936815,
+   "agent_replays": 12,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.734,
+   "peak_processes": 10,
+   "peak_disk_gb": 0.833,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-privileged--gcdbhfc/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "119a8ab825c29c0764649961832eafbba0df1603162fbe84226f3c3771f8499e"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.429,
+      "log_step": null,
+      "text_zh": "观察：视频0.429–3.214 s，Barkour机器人反复抬落腿，足端接触棋盘格地面并支撑躯干，地面相对机器人移动，行进尝试已开始。"
+     },
+     {
+      "t": 4.714,
+      "log_step": null,
+      "text_zh": "观察：视频3.429–5.357 s，机器人仍由腿支起，但黄色前端从偏向画面右侧转为较正对相机，前半段已出现明显朝向变化。"
+     },
+     {
+      "t": 9.5,
+      "log_step": null,
+      "text_zh": "观察：视频6.000–10.143 s，机器人机身保持离地，后半段朝向变化有限，视频9.071–10.143 s逐帧检查未见翻倒；成功参考视频的后半段呈持续明显转向，仅用于外观对照。"
+     },
+     {
+      "t": null,
+      "log_step": 186,
+      "text_zh": "观察：log step 186对最终交付文件trajectory.npz运行回放，报告1000/1000控制步、fell_at=null、tracking_ok=false；第0段平移/偏航误差为0.3056 m/s、0.3785 rad/s，第1段为0.3028 m/s、0.4985 rad/s。inputs/run.json评分器记录deterministic=1、fell=0，最差两项误差与该回放一致，均超过DAG阈值0.2 m/s和0.3 rad/s。"
+     },
+     {
+      "t": null,
+      "log_step": 186,
+      "text_zh": "推断：该节点要求完整日程及每段误差全部合格；交付回放和评分器均证实两段跟踪未达标，因此节点从未完成、终值为false，不能因机器人迈步且未倒地而计为达成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.429–8.357 s，Barkour四条腿的足端反复抬落、接触地面，承重并推动机身移动；视频9.071–10.143 s机身仍由足端支撑，未见躯干落地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.3056
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.4985
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 4.714,
+   "log_step": 186,
+   "where": "操作中",
+   "evidence_zh": "观察：视频4.286–5.357 s，机器人在执行前半段行进时已有明显朝向漂移；视频6.000–10.143 s后半段也未呈持续充分转向。log step 186最终交付回放的两段平移及偏航误差均超标，评分器最差误差为0.3056 m/s和0.4985 rad/s；视频4.714 s仅作可见偏航偏差的导航锚点，不代表误差均值首次超阈值时刻。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 4.714,
+    "log_step": 186,
+    "evidence_zh": "观察：视频3.429–5.357 s，Barkour在迈步中明显改变朝向，而任务前半段要求近乎不转向的平移；视频6.000–10.143 s后半段转向又很有限，机器人始终未倒地。log step 186交付回放确认两段速度跟踪均不合格：这是行进方向和转向未跟上指令的步态问题。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：log step 105记录agent明确区分前半段前进/正侧移与后半段负侧移/正偏航的目标；log step 181记录生成器采用带相位偏移的逆运动学小跑，并给平移、偏航设置分段增益。经log step 183修正生成设置后，log step 186对最终文件的回放仍在两段都出现平移和偏航误差超标，且最差值与评分器一致；结合视频4.286–5.357 s的朝向漂移及后半段转向不足，归为关节轨迹未实现所需机身速度的控制能力不足，证据不足以细分控制子类型。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有本次逐帧状态日志或完整速度时间序列；视频可见行进、朝向变化和支撑，但不能量出各帧机体系速度、足地接触力或精确累计误差。两段均值来自log step 186的最终交付回放，最差值由评分器字段核对；无法定位误差均值首次越界时刻，也无法进一步区分轨迹设计、反馈或具体参数的贡献。"
+ },
+ "summary_zh": "机器人完成了1000个控制步，视频可见迈步及足端支撑，评分器记录未翻倒，但“按指令行进”节点未达成，最终为0/1。前半段出现明显朝向漂移，后半段转向不足，最差平移误差0.3056 m/s、偏航误差0.4985 rad/s均超阈值，断点为行进操作中的步态跟踪失败，根因归为控制。缺少逐帧状态与速度序列，不能确定误差首次越界时刻或更具体的控制子类型。",
+ "conf": 7,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成（final=false）、断点和根因与评分器字段（worst_lin_err 0.3056、worst_yaw_err 0.4985、fell=0）及 log step 186 回放一致；视频帧可见持续迈步、未倒地，未发现需修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:03:06Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--gcdbhfc/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--gcdbhfc/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "2d4aeb73ff7784b0a372b19b40efe365535bf9c9cc359ae763ecf8ace2bb2379",
+   "video_sha256": "119a8ab825c29c0764649961832eafbba0df1603162fbe84226f3c3771f8499e",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--ieyzivc.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--ieyzivc.json
new file mode 100644
index 0000000000000000000000000000000000000000..1b6549d15402b9523abe07821df93a31457ad70f
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--ieyzivc.json
@@ -0,0 +1,173 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-privileged--ieyzivc",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.131,
+   "worst_yaw_err": 0.0539,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 4,
+   "agent_resets": 6,
+   "agent_sim_steps": 15000,
+   "agent_replays": 2,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.678,
+   "peak_processes": 8,
+   "peak_disk_gb": 0.828,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-privileged--ieyzivc/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "8b16175b9799740b1fde0b75c77daa3b25f6aafbebde919affac999ae002350d"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [
+     [
+      10.143,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.571,
+      "log_step": null,
+      "text_zh": "观察：视频 0.000 s 机器人处于四足站姿；视频 0.214–0.571 s 连续帧显示足端轮换抬落，机身由腿支起，地面棋盘格相对机器人移动；视频 0.857–4.857 s 的抽帧中机器人继续迈步平移。"
+     },
+     {
+      "t": 6.0,
+      "log_step": null,
+      "text_zh": "观察：视频 4.571–6.214 s 的逐帧检查显示机器人持续迈步，随后机身朝向发生变化；视频 6.286–9.143 s 的抽帧中转向进一步扩大，平移方向相对机身也改变，未见倒地。"
+     },
+     {
+      "t": 10.143,
+      "log_step": null,
+      "text_zh": "观察：视频 9.143–10.143 s 的逐帧检查中机器人继续交替落足，躯干始终离开地面，末帧仍保持腿部支撑。inputs/run.json 的 reward 记录 success=1、replay_success=1、deterministic=1、n_actions=horizon=1000、fell=0，worst_lin_err=0.131 m/s、worst_yaw_err=0.0539 rad/s，分别小于任务阈值 0.2 m/s 与 0.3 rad/s。"
+     },
+     {
+      "t": 10.143,
+      "log_step": 11,
+      "text_zh": "推断：该节点要求完成整个日程，不能从开局站立或前半段行走计入。log step 10 的脚本保存关节目标，log step 11 的完整日程检查报告 1000 步且两段误差达标，log step 12 随后调用所存文件的回放；这些与本次评分器字段及视频终态一致，因此节点在回合末达成，导航时间取已检查的末帧视频 10.143 s（约数）。这里未把 agent 工作时间或仿真 10.0 s 直接换算为视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.214–0.571 s、4.571–6.214 s 及 9.143–10.143 s 的连续帧可见 Barkour 足端交替着地与抬起，落地足承重支撑躯干，机器人借此平移并转向；躯干未落地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.131
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.0539
+  }
+ ],
+ "failure": null,
+ "summary_zh": "机器人完成“按指令行进”节点，最终进度为 1/1：视频前半段持续迈步平移，后半段明显转向，逐帧检查视频 9.143–10.143 s 至末帧仍保持腿部支撑。评分器确认全部 1000 个控制步完成且未触发终止条件，最大分段平均平面速度误差为 0.131 m/s、偏航角速度误差为 0.0539 rad/s，均达标。无状态日志，精确误差与终止历史依据本次评分器字段；完成时间取视频末帧约 10.143 s，未假定仿真时钟与视频时钟相同。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "单节点已达成，final=1、无failure与评分器字段（success=1、fell=0、误差0.131/0.0539达标）及视频一致；时间取视频末帧，未换算agent时间，接触标注合理。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:02:36Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--ieyzivc/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--ieyzivc/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "533eb3fb3f691ffaf4177273567eb7768d98ee47607ee5f03063c2c6f71dcb36",
+   "video_sha256": "8b16175b9799740b1fde0b75c77daa3b25f6aafbebde919affac999ae002350d",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--oqmtteh.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--oqmtteh.json
new file mode 100644
index 0000000000000000000000000000000000000000..448359d9af35c8e13541af98e78f05f8c2921cb6
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--oqmtteh.json
@@ -0,0 +1,199 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-privileged--oqmtteh",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.4832,
+   "worst_yaw_err": 0.7525,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 15,
+   "agent_resets": 46,
+   "agent_sim_steps": 96610,
+   "agent_replays": 5,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.455,
+   "peak_processes": 12,
+   "peak_disk_gb": 0.241,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-privileged--oqmtteh/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "77fc0a48811eb008e6b28bbaccdf1c23e91beb6284c3eee0dbd179c18c8cb105"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 1.0,
+      "log_step": null,
+      "text_zh": "观察：逐帧检查视频 0.0–10.143 s，机器人在棋盘格地面上不断屈伸、交替迈腿，足端落地支撑躯干；视频 0.286–1.714 s 可见机身随步伐小幅摆向，全程未见侧翻或躯干落地。"
+     },
+     {
+      "t": 5.5,
+      "log_step": null,
+      "text_zh": "观察：视频 4.714–6.286 s 及视频 6.857–10.143 s 中，机器人继续迈步，黄色头部仍大致朝画面右侧，只有小幅往复摆向，未出现后半程持续明显转向。作为外观参照，成功示范视频 5.5–9.5 s 的机器人从朝右逐渐转至朝左、再朝向相机；该示范仅用于辨识转向外观。"
+     },
+     {
+      "t": 10.143,
+      "log_step": 16,
+      "text_zh": "观察：log step 16 记录机器人轨迹执行和回放均为 1000 步、fell_at=null，随后保存 (1000,12) 轨迹；两段平移误差为 0.4113、0.4832 m/s，偏航误差为 0.7525、0.6784 rad/s。最终评分器同样报告 n_actions=1000、fell=0、worst_lin_err=0.4832、worst_yaw_err=0.7525、deterministic=1；视频最后一秒至视频 10.143 s 仍在支撑迈步。"
+     },
+     {
+      "t": 10.143,
+      "log_step": 16,
+      "text_zh": "推断：已保存回放的最差误差与最终评分完全一致，且视频中缺乏持续转向的表现相符，故评分片段对应这条基线轨迹；后续 log step 22、31、37 的短程跌倒试验不是评分片段。完整日程结束仍不满足平移 0.2 m/s 和偏航 0.3 rad/s 两项阈值，因此“按指令行进”从未完成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.286–1.714 s、4.714–6.286 s 和 9.214–10.143 s，Barkour前后腿持续交替屈伸，足端接触棋盘格地面并支撑、推动躯干；未见躯干翻落地面。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.4832,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.7525,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 1.0,
+   "log_step": 16,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 0.286–1.714 s 机器人已在迈步，机身反复小幅摆向；视频 5.071–10.143 s 继续行走却未形成后半段所需的持续转向。log step 16 保存的完整回放两段速度误差均超阈值，最终评分器确认 worst_lin_err=0.4832 m/s、worst_yaw_err=0.7525 rad/s 且 fell=0。推断：失败发生在持续执行“按指令行进”期间，视频 1.0 s 仅作早段异常行进的粗略导航锚点，并非均值判据的瞬时越线时间。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 1.0,
+    "log_step": 16,
+    "evidence_zh": "观察：视频 0.286–1.714 s 及 4.714–6.286 s 中机器人不断迈步，机身随步伐往复摆向；视频 6.857–10.143 s 头部仍大致朝右，未实现持续转向，直到末帧仍由腿支起。log step 16 的已保存回放与最终评分的误差数值一致，且最终 fell=0。推断：这是未能实现所需平移、转向跟踪的行进步态，未见翻倒。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": null,
+    "log_step": 17,
+    "source_zh": "日志",
+    "evidence_zh": "观察：log step 10 的脚本以步态相位、足端目标和逆运动学生成关节轨迹，并用速度、偏航反馈修正目标；log step 16 运行后保存基线。log step 17 在修改该控制器前重跑诊断，第一段平均平移速度约为 [0.123,-0.004] m/s，对应指令 [0.459,0.221] m/s；第二段平均速度约为 [0.166,0.073] m/s，对应指令 [0.090,-0.389] m/s，第二段平均偏航速度约 -0.0039 rad/s，对应指令 +0.643 rad/s。该诊断是会话证据；评分片段本身的跟踪失败另由 log step 16 的已保存回放、最终评分器误差字段以及视频 5.071–10.143 s 缺乏持续转向共同支持。推断：脚本虽设置了正确的速度日程，生成的动作未能产生所需的机体速度与转向，根因可归于控制；现有证据不足以唯一细分接触设计、反馈律或参数问题。"
+   }
+  ],
+  "gap_zh": "没有逐帧仿真状态日志，不能独立重算速度误差均值、足地接触力或足端净空；精确验收依赖最终评分器及 log step 16 的已保存回放。log step 18 以后关于下沉、摆腿拖地等解释来自会话诊断和 agent 自述，未在本次评分片段中独立验证，故不据此指定更细控制根因。连续误差的失败没有唯一瞬时事件，断点时刻仅为视频导航锚点。"
+ },
+ "summary_zh": "机器人跑满 1000 个控制步并持续由腿支撑，但“按指令行进”未完成，最终进度为 0/1；后半程没有形成明显持续转向，最差平移误差 0.4832 m/s、偏航误差 0.7525 rad/s 均超过阈值。断点位于该步骤操作中，现象为步态未达到要求，日志支持控制层面的速度跟踪不足；本次交付是 log step 16 保存的基线，后续会话试验中的跌倒不属于评分视频。缺少状态日志，无法进一步确认足地接触及更细控制机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成（0/1）、断点、步态现象与控制根因均有视频、log step 16/17 与评分器字段支持；未发现需要修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:03:04Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--oqmtteh/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--oqmtteh/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "6e3b9ab4d15a2cc1071211bf8b04ede193aa7559c8a62e22b0fb80ab40cf28e2",
+   "video_sha256": "77fc0a48811eb008e6b28bbaccdf1c23e91beb6284c3eee0dbd179c18c8cb105",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--r6czfhq.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--r6czfhq.json
new file mode 100644
index 0000000000000000000000000000000000000000..17da34ed4d855d2cec74de34d52c8fc23ffbda7e
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--r6czfhq.json
@@ -0,0 +1,173 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-privileged--r6czfhq",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.1631,
+   "worst_yaw_err": 0.28,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 584,
+   "agent_resets": 1418,
+   "agent_sim_steps": 1468815,
+   "agent_replays": 18,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 6.747,
+   "peak_processes": 13,
+   "peak_disk_gb": 0.256,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-privileged--r6czfhq/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "12fd5cd903a233ed92b026c73b509786d6586394dce689d5b2b29d5a827418c2"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [
+     [
+      10.14,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0–5.0 s 的连续抽帧中，Barkour 四足机器人在棋盘格地面上交替迈腿，地面相对机器人移动，躯干由腿支起；起始站立尚未完成整段日程。"
+     },
+     {
+      "t": 5.5,
+      "log_step": null,
+      "text_zh": "观察：视频 4.50–6.50 s 逐帧核对显示机器人继续迈步并改变机身朝向；视频 7.0–10.14 s 可见持续转向后的行进，视频 9.14–10.14 s 逐帧未见倒地，足端落地支撑躯干。"
+     },
+     {
+      "t": null,
+      "log_step": 187,
+      "text_zh": "观察：log step 181 的生成脚本在执行后保存关节目标轨迹，log step 182 生成完成；log step 183 对该交付文件的两次官方回放均成功且 state_sha 相同，log step 187 最终回放报告 steps=1000、horizon=1000、success=true、fell_at=null、tracking_ok=true；两段 lin_err 分别为 0.1631、0.1617 m/s，yaw_err 分别为 0.2662、0.2800 rad/s，与本次 reward 的最差误差及成功字段一致。"
+     },
+     {
+      "t": 10.14,
+      "log_step": null,
+      "text_zh": "推断：结合本次评分器 n_actions=horizon=1000、fell=0、replay_success=1，以及视频全程行进、后半段转向和视频 10.14 s 最后帧仍受腿支撑的表现，完整日程验收成立，节点 final=true；达成时间以视频末帧约 10.14 s 为锚点（约半秒精度），不把会话中的早期试验计入本次进度。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–10.14 s，Barkour 四条腿的足端交替落在棋盘格地面上，支撑躯干完成平移及转向；视频 4.50–6.50 s 和 9.14–10.14 s 逐帧可见迈腿与落足。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.1631
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.28
+  }
+ ],
+ "failure": null,
+ "summary_zh": "机器人完成“按指令行进”节点（最终 1/1）：视频前半段持续平移，后半段明显转向，直到视频 10.14 s 末帧仍由足端支撑，评分器确认执行满 1000 个控制步且成功。最差平面速度误差 0.1631 m/s、偏航角速度误差 0.28 rad/s 均达标；本次没有状态日志，精确跟踪误差及终止判据采用评分器字段和交付文件的最终官方回放记录，节点达成时间仅以视频末尾近似定位。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "成功运行，唯一节点 final=true、contact 腿/躯干·支撑、无失败层，均与视频帧及 log step 182/183/187 的回放结果（steps=1000、success、lin_err 0.1631、yaw_err 0.28）一致，未发现需要修改之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:05:40Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--r6czfhq/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--r6czfhq/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "7a55154ba6a4a74417493da0f84d9d453be1dd3fcd637d3f7daf5d5e4242013f",
+   "video_sha256": "12fd5cd903a233ed92b026c73b509786d6586394dce689d5b2b29d5a827418c2",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--zvvvnnm.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--zvvvnnm.json
new file mode 100644
index 0000000000000000000000000000000000000000..13f4086159010cbb231baf22878b71b945c938de
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-privileged--zvvvnnm.json
@@ -0,0 +1,173 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-privileged--zvvvnnm",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.0852,
+   "worst_yaw_err": 0.1101,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 52,
+   "agent_resets": 938,
+   "agent_sim_steps": 305790,
+   "agent_replays": 5,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-privileged--zvvvnnm/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "8205ffc08a4db65fe8d9e40f4b79d4cc898129e614818011df4930ebd3a0d479"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [
+     [
+      10.143,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.5,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0–5.0 s 的连续抽帧中，Barkour四足机器人在棋盘格地面上交替迈腿，地面纹理相对跟随相机移动，机身保持离地支起。"
+     },
+     {
+      "t": 5.5,
+      "log_step": null,
+      "text_zh": "观察：视频 4.786–5.857 s 的逐帧画面显示机器人继续迈步并开始改变机身朝向；视频 6.0–10.143 s 的画面中转向持续，视频 9.0–10.143 s 逐帧检查未见倒地，末帧机身仍由腿支撑。"
+     },
+     {
+      "t": null,
+      "log_step": 55,
+      "text_zh": "观察：log step 55 中，agent脚本重新生成并保存最终交付轨迹，随后对该文件进行两次新进程回放；工具均报告steps=1000、horizon=1000、success=true、fell_at=null、tracking_ok=true，最终状态哈希均为311c62c55717。两段平面速度误差均值分别为0.0527、0.0852米/秒，偏航角速度误差均值分别为0.0837、0.1101弧度/秒；这些数值与本次reward的最大误差及成功结果一致。"
+     },
+     {
+      "t": 10.143,
+      "log_step": null,
+      "text_zh": "推断：结合本次评分器reward的success=1、replay_success=1、deterministic=1、n_actions=horizon=1000、fell=0，以及视频末帧仍在支撑行进，完整日程及误差验收已通过；节点只在整段日程结束时计入，达到时间以已检查的末帧视频 10.143 s 为近似锚点，不把开局站立或前半段行进计为完成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–10.143 s，Barkour机器人的足端随四条腿交替落向棋盘格地面，支撑躯干完成平移与转向；视频 4.786–5.857 s 的密集画面可见抬足与落足交替。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.0852
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.1101
+  }
+ ],
+ "failure": null,
+ "summary_zh": "本次完成唯一节点“按指令行进”（最终1/1）：视频前半段持续迈步平移，后半段明显转向，视频 10.143 s 末帧仍由腿支撑。评分器确认完整执行1000步、fell=0且确定性回放成功，最大分段平面速度误差0.0852米/秒、偏航角速度误差0.1101弧度/秒均低于阈值，无失败断点。没有状态日志，精确误差与全程终止条件依据本次评分器结果，完成时间以视频末帧作近似锚点。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "成功运行，唯一节点按指令行进 final=true 有评分器结果（success=1、fell=0、1000步、误差低于阈值）支持，无失败断点；log step 55 的数值与 reward 一致，时钟与接触标注无误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:02:00Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--zvvvnnm/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-privileged--zvvvnnm/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "3e40a1ec1f2d38d15765ecaac1c494831f83e3e6735e58911da9b8d6c020e3b9",
+   "video_sha256": "8205ffc08a4db65fe8d9e40f4b79d4cc898129e614818011df4930ebd3a0d479",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--8u3fim4.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--8u3fim4.json
new file mode 100644
index 0000000000000000000000000000000000000000..6512351bb79f963ddcd275a20bc52da0f730d167
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--8u3fim4.json
@@ -0,0 +1,190 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard--8u3fim4",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "live_success": 1,
+   "replay_success": 1,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 1000,
+   "sim_steps": 1000,
+   "physics_steps": 5000,
+   "steps": 783,
+   "elapsed_s": 615.8,
+   "video_rendered": 1,
+   "fell": 0,
+   "worst_lin_err": 0.1045,
+   "worst_yaw_err": 0.1371,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 2,
+   "reset_budget": 10,
+   "episodes": 3,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.527,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.829,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard--8u3fim4/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "d6f123ac9f29e917ffb91ac4faf3a661f7dc1def5ceb66ef1e03061bb9455178"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [
+     [
+      10.14,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.00 s机器人从棋盘格地面上的站立姿态开始；视频0.21–5.07 s跟随视角中四条腿交替迈步，地面格纹相对机器人连续移动，躯干由足端支撑。开局站立尚未完成完整指令日程。"
+     },
+     {
+      "t": 5.5,
+      "log_step": null,
+      "text_zh": "观察：视频5.14–6.79 s机器人继续迈步，地面相对移动方向发生改变，机身朝向逐渐转动；视频7.00–10.14 s持续转向行进。逐帧检查视频9.21–10.14 s，机器人仍由腿支撑，未见倒地。"
+     },
+     {
+      "t": null,
+      "log_step": 44,
+      "text_zh": "观察：log step 28的脚本重置后开始第三回合；log step 36服务状态为episode=3、control_steps=400、done=False，log step 37脚本接续该现场回合，未再重置；log step 44服务最终状态为episode=3、control_steps=1000、done=True、reason=horizon。前两回合调试中的失败不是本次评分视频的事件。"
+     },
+     {
+      "t": null,
+      "log_step": null,
+      "text_zh": "观察：inputs/run.json的reward报告success=live_success=replay_success=1、live_matches_replay=1、deterministic=1、n_actions=sim_steps=1000、fell=0；worst_lin_err=0.1045 m/s，小于0.2 m/s；worst_yaw_err=0.1371 rad/s，小于0.3 rad/s。精确误差采用评分器记录，而非agent在log step 45自行计算的抽样均值。"
+     },
+     {
+      "t": 10.14,
+      "log_step": null,
+      "text_zh": "推断：完整视频的持续行进、后半段换向和转向与本次评分器的1000控制步成功记录相符，故按指令行进节点最终成立；该节点要求完整日程完成，以检查过的末帧视频10.14 s作为完成时间近似锚点，不将agent工作时间换算为视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.21–10.14 s，Barkour四条腿的足端依次落在棋盘格地面并交替抬起，在平移、侧移及转向时支撑躯干；未见躯干倒地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.1045,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.1371,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": null,
+ "summary_zh": "本次评分为第三回合，机器人完成完整指令日程，唯一节点“按指令行进”在视频末段约10.14 s达到，最终1/1；前两回合调试失败不计入本次轨迹。视频显示机器人持续由足端支撑，前半段平移、后半段换向并转向，评分器确认1000控制步、fell=0，平移与偏航误差分别为0.1045 m/s和0.1371 rad/s，均达标。无状态日志，精确误差与历史终止条件依赖本次评分器记录，完成时刻取检查过的末帧作近似锚点。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "单节点“按指令行进”final true，与评分器 success=1、fell=0、1000 控制步、误差 0.1045/0.1371 一致；视频中机器人全程站立迈步，换向转向可见。接触与时钟标注无误，无失败标签。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:20:47Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--8u3fim4/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--8u3fim4/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "87e37e120553e61622c94f43349485354422b7da73e93c5c0c69d275ec47b77d",
+   "video_sha256": "d6f123ac9f29e917ffb91ac4faf3a661f7dc1def5ceb66ef1e03061bb9455178",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--a2jyusj.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--a2jyusj.json
new file mode 100644
index 0000000000000000000000000000000000000000..68b1344483d95e5564682fe549290534054891b1
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--a2jyusj.json
@@ -0,0 +1,206 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard--a2jyusj",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 524,
+   "sim_steps": 524,
+   "physics_steps": 2620,
+   "steps": 3422,
+   "elapsed_s": 870.0,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.489,
+   "worst_yaw_err": 0.4328,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.21,
+   "peak_processes": 6,
+   "peak_disk_gb": 0.286,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard--a2jyusj/replay.mp4",
+  "duration_s": 5.28,
+  "fps": 25.0,
+  "frames": 132,
+  "sha256": "a948a57b31621b02291d3d9003249eb318a3d04f0e60406c70fab43b61405a41"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 152,
+      "text_zh": "观察：视频0.00–5.24 s中，机器人四足交替落地迈步并移动；log step 152显示脚本重置到episode=11、resets_left=0，log step 164、166均报告该回合control_steps=524、done=True、reason=fell。此前回合的测试不作为本节点进度。"
+     },
+     {
+      "t": 1.2,
+      "log_step": 163,
+      "text_zh": "观察：视频约1.12–2.88 s中机器人由侧向逐渐转为面对跟随相机，视频3.20–4.96 s继续偏转；未保持前半段近乎固定的朝向。log step 163所载本回合控制步120、180的脚本传感汇总分别为vx=-0.20、-0.21 m/s，目标vx约0.46 m/s，控制步420的wz=-0.80 rad/s，目标约0.02 rad/s。"
+     },
+     {
+      "t": 5.24,
+      "log_step": 164,
+      "text_zh": "推断：log step 9给出100 Hz控制率，log step 164的本回合524步对应仿真5.24 s；已与视频5.24 s末帧和5.28 s片长核对（末帧显示时长约0.04 s）。机器人尚未完成1000步/10秒完整日程就被服务终止；评分器worst_lin_err=0.489 m/s、worst_yaw_err=0.4328 rad/s又分别超过0.2和0.3阈值，因此节点全程未达成，不能把开局站立或部分迈步计为完成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00–5.24 s中，Barkour的四条腿以足端交替接触棋盘格地面，承重并推动机身移动；可见近侧足端反复落地、抬起。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.489,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.4328,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 1.2,
+   "log_step": 163,
+   "where": "操作中",
+   "evidence_zh": "观察：视频约1.2 s起，机器人仍在迈步，却在第一段指令尚未切换时逐渐大幅改变朝向，至视频2.24–2.88 s面对跟随相机，之后继续偏转；log step 163记录本回合前向速度多次与目标反号。该行进步骤没有完成；log step 164显示服务最终于524控制步以fell终止。断点时间取持续跟踪偏离的粗略起点。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 1.2,
+    "log_step": 163,
+    "evidence_zh": "观察：视频1.12–4.96 s中机器人不断迈步，但在前半段持续偏转，从侧向转为正对相机再转向另一侧，没有沿近乎不变的朝向平移。log step 163的目标偏航角速度约0.02 rad/s，而控制步300、420的脚本传感汇总分别为-0.67、-0.80 rad/s；这是行进方向和转向跟踪偏离。视频4.08–5.24 s逐帧查看未见明确侧翻，故不凭fell字段另加平衡现象。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": 1.2,
+    "log_step": 150,
+    "source_zh": "日志 / 视频 / 评分器字段",
+    "evidence_zh": "推断：log step 150中agent编写对角小跑关节位置控制脚本，设置0.24秒步态周期、12步提交块，并读取IMU线速度、角速度和重力反馈；log step 151–152启动该脚本并重置为最终第11回合。视频1.12–4.96 s的持续转向偏离与log step 163中多次反向的前向速度、负偏航速度相符，评分器两项速度误差也均超限，支持该控制器未实现所要求的速度跟踪。没有足够证据把问题进一步定位到逆解、接触布局或某个特定参数。"
+   }
+  ],
+  "gap_zh": "观察：inputs中没有状态日志；run.json中log step 150的控制脚本中段已被截去，且最终服务仅报告reason=fell，评分器仅给出fell=1，未区分躯干翻倒、实际关节越界或躯干过低。视频末秒已逐帧检查，末帧仍可见机身在腿上、没有清楚侧翻，无法据此确定触发终止的具体条件，也无法从画面复算速度误差均值；根因只保留控制层级。log step 164的0.4834、0.3921是agent脚本自行汇总，原生指标采用评分器的0.489、0.4328。",
+  "unmet": {
+   "joystick_schedule_followed": false
+  }
+ },
+ "summary_zh": "评分回合为最后一次重置后的第11回合：机器人确实以足端支撑并持续迈步，但视频约1.2 s起明显偏转，未完成“按指令行进”，节点最终0/1。控制器未能实现速度跟踪，两项评分误差均超限，服务在524控制步（对应视频约5.24 s）以fell结束回合。缺少状态日志和具体终止条件，且视频未清楚显示侧翻，因此现象标为步态偏离，根因保留为控制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点joystick_schedule_followed为false、final 0与评分器（success=0, fell=1, 两项误差超限）一致；断点、步态现象和控制根因均有视频与log step 163/164支持，时钟采用控制步524=5.24 s并已核对，未见需改动之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:10:03Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--a2jyusj/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--a2jyusj/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "546a980e9b1e9ca9962866375d73cd8af5beabf269d6bcf5de0a4b30e8558c8b",
+   "video_sha256": "a948a57b31621b02291d3d9003249eb318a3d04f0e60406c70fab43b61405a41",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--jreayho.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--jreayho.json
new file mode 100644
index 0000000000000000000000000000000000000000..23fe23202f35db1969787a62014386c584ed1430
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--jreayho.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard--jreayho",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 117,
+   "sim_steps": 117,
+   "physics_steps": 585,
+   "steps": 2042,
+   "elapsed_s": 997.5,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.2772,
+   "worst_yaw_err": 0.4842,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard--jreayho/replay.mp4",
+  "duration_s": 2.36,
+  "fps": 25.0,
+  "frames": 59,
+  "sha256": "af697b08fed010cc4c72a5f16b8e5468ec8edb28cd8b5f66fcbbc0d1b28c1bd2"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.6,
+      "log_step": null,
+      "text_zh": "观察：视频0.00–1.20 s中，Barkour四足机器人在棋盘格地面上反复迈腿并移动，足端接触地面，已开始行进尝试；开局站立不满足完整日程验收。"
+     },
+     {
+      "t": 2.32,
+      "log_step": 102,
+      "text_zh": "观察：视频1.20–2.32 s逐帧可见机器人腿部折低、躯干下沉，最后仍保持近水平朝向，未出现参考成功视频后半段的持续转向过程。log step 102记录最终回合control_steps=117、sim_time_s=1.17、done=true、reason=fell；log step 104与106确认该回合为episode 11且无剩余重置，故此前测试回合不计入节点。"
+     },
+     {
+      "t": null,
+      "log_step": null,
+      "text_zh": "观察：inputs/run.json的评分器字段n_actions=117、fell=1、worst_lin_err=0.2772 m/s、worst_yaw_err=0.4842 rad/s，live_matches_replay=1。完整验收要求1000个控制步以及两项误差分别不大于0.2 m/s、0.3 rad/s，本次没有完成完整日程。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00–2.32 s，机器人四条腿的足端先后落地支撑并推动机身移动；末段腿部折低、机身下沉。画面未足以确认躯干本身触地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.2772,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.4842,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 2.16,
+   "log_step": 102,
+   "where": "操作中",
+   "evidence_zh": "观察：视频约2.00–2.32 s，正在迈腿行进的机器人腿部进一步折低，躯干靠近地面，未恢复正常支起高度即结束；log step 102的最终回合由服务在control step 117以reason=fell终止，未进入第二段指令。断点锚点取视频2.16 s的下沉过程，不将日志中的仿真1.17 s当作视频时间。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 2.16,
+    "log_step": 102,
+    "evidence_zh": "观察：视频1.92–2.32 s中机器人腿部屈折、躯干持续下沉，末帧机身明显低伏，未能维持由腿支起的行进姿态；log step 102记录服务将本回合终止为fell。画面支持支撑塌低，未显示机身翻转超过90度。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：log step 54、76、84、92、96及100记录agent自编的足端轨迹、姿态反馈和关节位置控制器；log step 102以最后一组步态参数执行最终回合，打印的前向速度从control step 50的0.32 m/s变为control step 100的0.71 m/s，目标为0.4594 m/s，随后服务以fell终止。结合视频末段的机身下沉，控制器未能同时维持行进跟踪和机身支撑；证据不足以确定具体增益、逆解或足端承载布局哪一项导致失稳。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "未提供逐步状态日志或最终回合的躯干世界高度、全部实际关节角及首次终止谓词；fell同时覆盖翻倒、关节越界和躯干过低，不能仅凭此字段认定具体触发项。log step 102打印的final up_z实际来自重力方向分量，不能直接当作机身向上轴；agent在log step 116关于高度规则或关节规则的解释属于待核实推测。视频总长2.36 s，而最终回合服务记录仿真1.17 s，未建立精确的逐步时钟映射，所有视频锚点均直接取自已检查的帧；评分器仅给出误差汇总，不能补出未运行的第二段误差。"
+ },
+ "summary_zh": "最终被评分的是第11回合，机器人已迈腿移动，但视频末段腿部折低、机身下沉，服务在control step 117以fell终止，唯一节点“按指令行进”始终未完成（0/1）。评分器两项速度误差均超过阈值，控制器未能维持稳定支撑与跟踪；缺少逐步状态，无法确定fell具体由高度、关节边界还是翻倒规则触发，且视频时钟与仿真时钟不能直接等同。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "标签与视频、日志一致：第11回合在control step 117以fell终止，唯一节点未达成；断点、平衡现象与控制根因均有证据且对不确定处已在gap_zh中说明，未发现需改内容的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:02:24Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--jreayho/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--jreayho/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "c324f31672fa65de4ca846ea593393c0897295bba04d457f37399366e2eb8555",
+   "video_sha256": "af697b08fed010cc4c72a5f16b8e5468ec8edb28cd8b5f66fcbbc0d1b28c1bd2",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--mbzvhxu.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--mbzvhxu.json
new file mode 100644
index 0000000000000000000000000000000000000000..537ffe5ac2d16d7ca418199beb02d227b027c377
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--mbzvhxu.json
@@ -0,0 +1,209 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard--mbzvhxu",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 1000,
+   "sim_steps": 1000,
+   "physics_steps": 5000,
+   "steps": 6025,
+   "elapsed_s": 1067.9,
+   "video_rendered": 1,
+   "fell": 0,
+   "worst_lin_err": 0.3604,
+   "worst_yaw_err": 0.7504,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard--mbzvhxu/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "ac6bbcd9d5603357db8ea2dbf676c516b9997cb367a734fc2cc747755afe868a"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 47,
+      "text_zh": "观察：log step 47 的机器人服务返回 reset -> 11 0，随后运行最终回合；log step 48 确认 episode=11、control_steps=1000、reason=horizon。视频从站立开始，连续行进至结尾，没有中途重置；前面调试回合的摔倒不属于本次评分轨迹。"
+     },
+     {
+      "t": 5.5,
+      "log_step": 47,
+      "text_zh": "观察：视频 0.0–5.0 s 机器人在棋盘格地面上小幅迈步平移；视频 5.5–9.0 s 继续移动，黄色头部从画面右侧逐渐转向相机，机身由腿支起。log step 7 的服务指令表要求仿真 5.01 s 后偏航角速度为 +0.6428 rad/s；log step 47 最终回合在仿真 6、7、8、9 s 打印的服务 IMU 陀螺仪偏航速度分别为 -0.35、-0.34、-0.28、-0.21 rad/s，均为反向。"
+     },
+     {
+      "t": 10.143,
+      "log_step": 48,
+      "text_zh": "观察：逐帧查看视频 9.0–10.143 s，机器人仍由足端支撑，末段迈步减弱，未见翻倒；log step 48 服务终态为完成 1000 个控制步并因 horizon 结束。评分器记录 n_actions=1000、fell=0、worst_lin_err=0.3604 m/s、worst_yaw_err=0.7504 rad/s，后两项分别超过 0.2 m/s 和 0.3 rad/s 阈值。"
+     },
+     {
+      "t": 10.143,
+      "log_step": 33,
+      "text_zh": "推断：该节点要求完整日程结束且两项跟踪误差均合格，因此本回合从未达到节点，不能因走满日程而记为完成。agent 在 log step 49 自述仅偏航失败，但其 log step 33 的 score 函数对平移误差采用分量绝对值均值，并仅使用每 25 个控制步采样的结果，区别于评分器的逐步平面误差向量模长均值；log step 47 的自算数值不能替代评分器记录。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–10.143 s，Barkour 四条腿的足端在棋盘格地面交替落足，支撑躯干并使其平移、转向；末段趋于站立。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.3604
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.7504
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 5.5,
+   "log_step": 47,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 5.5–9.0 s，机器人继续小步移动，黄色头部逐渐转向相机，未见翻倒；log step 47 最终回合的仿真 6–9 s IMU 偏航速度持续为负，而 log step 7 服务给出的第二段目标为 +0.6428 rad/s。推断：换指令后的反向转动是未修复的行进跟踪失败的明显表现，取视频 5.5 s 为导航锚点；该回合最终走满 1000 个控制步，但评分器两项误差均超阈值，节点未完成。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 5.5,
+    "log_step": 47,
+    "evidence_zh": "观察：视频 5.5–9.0 s，机器人虽保持足端支撑并移动，黄色头部却从画面右侧逐渐转向相机；末段视频 9.0–10.143 s 未见翻倒。log step 47 的最终回合服务 IMU 读数在仿真 6–9 s 给出负偏航速度，与服务指令表（log step 7）的正偏航目标相反。推断：行进中的方向跟踪偏离，属于未倒地的步态失败。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志",
+    "evidence_zh": "推断：agent 的行进控制器未能把正确的速度目标实现为机体运动。log step 33 的脚本包含正确指令表、速度及偏航反馈，并通过足端轨迹和逆解发送关节位置；log step 44、46 的修改在最终回合恢复对角迈步方案，log step 47 确认执行至 1000 个控制步，但该回合 IMU 偏航速度在仿真 6–9 s 仍持续为负，评分器记录平移和偏航误差均超标。证据支持控制实现不足，尚不能区分足端接触设计、轨迹生成或反馈参数的具体责任。"
+   }
+  ],
+  "gap_zh": "没有状态日志或逐控制步的完整传感记录，评分器也未给出各段的精确均值；不能从跟随相机与头部相机复算误差，或确定首次偏离的精确时刻及具体控制机制。agent 自算平移误差与评分器采用不同统计方法，节点和原生指标以评分器结果为准；不据 agent 关于足端打滑的自述进一步细分根因。"
+ },
+ "summary_zh": "最终评分的第 11 回合走满 1000 个控制步，视频至 10.143 s 仍由腿支撑，评分器 fell=0，但“按指令行进”节点未完成（0/1）。换指令后出现反向偏航，评分器平移误差 0.3604 m/s、偏航误差 0.7504 rad/s 均超阈值，标为全身·步态，根因归于控制。缺少逐步状态及分段精确评分，不能细分控制失效机制；agent 自算平移合格的结论未采用。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未完成（评分器 worst_lin_err 0.3604、worst_yaw_err 0.7504 超阈值，fell=0，1000 步 horizon）、断点、步态现象和控制根因均有 log 与评分器字段支持，视频时间在范围内，未发现需改动的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:03:34Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--mbzvhxu/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--mbzvhxu/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "71d2809db674d12b37c3631b1b8ec82f8c97a105435805f9f408de5e16038dbc",
+   "video_sha256": "ac6bbcd9d5603357db8ea2dbf676c516b9997cb367a734fc2cc747755afe868a",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--nksvqsy.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--nksvqsy.json
new file mode 100644
index 0000000000000000000000000000000000000000..2922abcfad9c79f763b2d2c77255a22536b2c418
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--nksvqsy.json
@@ -0,0 +1,205 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard--nksvqsy",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "graded_by_live": 1,
+   "replays": 0,
+   "finalized": 1,
+   "missing_log": 0,
+   "sim_error": 0,
+   "grader_error": 0,
+   "infra_error": 0,
+   "n_actions": 1000,
+   "sim_steps": 1000,
+   "physics_steps": 5000,
+   "steps": 1832,
+   "elapsed_s": 1695.8,
+   "video_rendered": 1,
+   "fell": 0,
+   "worst_lin_err": 0.5119,
+   "worst_yaw_err": 0.6466,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "instance_mismatch": 0,
+   "no_verdict": 0,
+   "gave_up": 0,
+   "sim_crashed_on_call": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 4,
+   "reset_budget": 10,
+   "episodes": 5,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.321,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.298,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard--nksvqsy/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "91db0fa4bbbd0a820e3410e3aef654517181e1cf65ad9fb3574f9f3387f2dee6"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.5,
+      "log_step": 45,
+      "text_zh": "观察：视频 0.00–1.43 s 的逐帧画面显示，Barkour四足机器人由足端支起躯干，仅有小幅姿态变化，没有启动交替迈步；log step 45 的脚本重置到第5回合后，重复同一组站姿关节目标600个控制步。"
+     },
+     {
+      "t": 5.0,
+      "log_step": null,
+      "text_zh": "观察：视频全程按约0.5 s抽帧，并逐帧检查视频 4.64–5.71 s，机器人仍在原地站立；后半段未出现持续侧移或明显转向。成功参考视频约 5–9.5 s 可见连续迈步和机身朝向明显改变，该参考仅用于辨认视觉线索。"
+     },
+     {
+      "t": 10.14,
+      "log_step": 46,
+      "text_zh": "观察：视频 9.00–10.14 s 的逐帧画面仍显示机器人站立，未见跌倒或行进；log step 46 的脚本继续重复站姿目标400步，服务返回第5回合 control_steps=1000、done=True、reason=horizon。评分器记录 fell=0、worst_lin_err=0.5119 m/s、worst_yaw_err=0.6466 rad/s，分别超过0.2 m/s及0.3 rad/s的阈值，完整日程的行进验收未通过。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–10.14 s 中，Barkour足端接触棋盘格地面，四腿支起躯干；实际仅保持站姿并小幅调整，没有形成行进所需的连续交替落足。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.5119,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.6466,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 0.5,
+   "log_step": 45,
+   "where": "首步前",
+   "evidence_zh": "观察：视频 0.00–1.43 s 的逐帧画面及后续视频至 10.14 s 均未见持续迈步，机器人只保持站姿；log step 45–46 的第5回合脚本仅重复固定关节目标，服务最终报告1000步到达horizon。评分器两项速度误差超标；先前测试回合已被重置，不计为本次评分轨迹的跌倒。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [
+   {
+    "cat": "没尝试",
+    "item": "没尝试",
+    "t": 0.5,
+    "log_step": 45,
+    "evidence_zh": "观察：视频 0.00–10.14 s 中机器人始终由四腿支撑站立，未启动按指令平移和转向的行进；log step 45–46 的最终回合只发送600步及400步重复站姿关节目标。因此本次评分回合没有尝试行进，不能将此前测试回合的失稳当成本回合现象。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 39,
+    "evidence_zh": "推断：log step 39 的步态测试中，agent脚本记录前向速度仅约0.11–0.12 m/s，低于指令约0.459 m/s，偏航速度从约−0.94变到+0.87 rad/s，服务在第2回合293个控制步返回reason=fell；log step 42、44 的调参测试也分别在246及108个控制步返回reason=fell。随后log step 45–46 改为仅维持站姿，视频确认最终没有行进；这些会话测试支持未能实现可用行进控制这一根因，但不能证明牵引、力矩或某个特定参数有问题。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有状态日志；最终回合仅有站姿命令、服务终态和评分器汇总指标，无法定位此前步态控制失败的具体动力学机制，故根因保留为“控制”，不细分子类。此前测试回合的reason=fell不是本次评分视频中的跌倒证据。"
+ },
+ "summary_zh": "评分第5回合中机器人保持站立并运行满1000步，但未启动按指令行进，节点最终为0/1；评分器记录平面速度误差0.5119 m/s、偏航角速度误差0.6466 rad/s，均超阈值，fell=0。断点在行进首步前，现象为没尝试；此前步态测试反复失败后，最终回合只维持站姿，根因归为控制。没有状态日志，无法进一步确定具体控制失效机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "节点未达、断点首步前/没尝试、根因控制（来自会话测试证据，依 FAIL-22 做法）均与视频、log 及评分器字段一致；视频 0–10.2 s 全程站立。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:22:12Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--nksvqsy/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--nksvqsy/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "9f6cbaad96fa7c2058010fb5cf3db93bba7f7149655dbafaf7654575a65409c4",
+   "video_sha256": "91db0fa4bbbd0a820e3410e3aef654517181e1cf65ad9fb3574f9f3387f2dee6",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--p3ejwxv.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--p3ejwxv.json
new file mode 100644
index 0000000000000000000000000000000000000000..c26e029cc8c4bba2e284a7f112c26caa8172360b
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--p3ejwxv.json
@@ -0,0 +1,189 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard--p3ejwxv",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 365,
+   "sim_steps": 365,
+   "physics_steps": 1825,
+   "steps": 494,
+   "elapsed_s": 1638.1,
+   "video_rendered": 1,
+   "fell": 0,
+   "worst_lin_err": 0.2614,
+   "worst_yaw_err": 0.2735,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 1,
+   "reset_budget": 10,
+   "episodes": 2,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.179,
+   "peak_processes": 6,
+   "peak_disk_gb": 0.26,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard--p3ejwxv/replay.mp4",
+  "duration_s": 4.88,
+  "fps": 25.0,
+  "frames": 122,
+  "sha256": "3eb954a851c5c2fad47064697476e70809d026625c3d1cf50e5efffe5e6237d7"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 2.4,
+      "log_step": null,
+      "text_zh": "观察：已检查本次视频全部122帧（视频0.00–4.84 s）；机器人从站立、轻微调整发展为迈步移动，视频2.4–4.84 s腿部交替运动，末段机身明显倾斜，最后仍由腿支起，未见翻倒。画面没有呈现完整日程结束。"
+     },
+     {
+      "t": null,
+      "log_step": 57,
+      "text_zh": "观察：log step 42重置为episode 2；log step 54、55、57连续推进该回合，log step 57最后读数为control_steps=365、done=False。评分器reward记录n_actions=365、sim_steps=365、fell=0、live_matches_replay=1、success=0；这与DAG要求完整1000个控制步不符，不能把先前episode 1的终止或本地仿真测试计入评分回合。"
+     },
+     {
+      "t": null,
+      "log_step": 59,
+      "text_zh": "推断：结合最后回合仅365/1000控制步及视频未显示完整日程，按指令行进节点始终未达到，final=false；log step 59的模型提供方过滤器中断发生在服务仍未结束回合之后。控制步计数只用于确认日程未完成，未将仿真365×0.01=3.65 s直接当成视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频2.4–4.84 s，Barkour四条腿的足端交替落在棋盘格地面，支撑躯干并随迈步移动；末段倾斜时仍有足端支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.2614
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.2735
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": null,
+   "log_step": 59,
+   "where": "操作中",
+   "evidence_zh": "观察：机器人已在视频中迈步，最后检查到的视频4.84 s仍未翻倒；log step 57现场回合停留在控制步365且done=False，log step 58报告尚余185 s墙钟预算，log step 59随即出现“This request was blocked by Gemini's filters”并结束日志。推断：中断时正在进行按指令行进，未完成完整日程；过滤器触发无法定位到某一视频帧，故t=null。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [],
+  "root_causes": [],
+  "gap_zh": "观察：inputs/有本次与参考视频及完整run.json，但没有状态日志；log step 59记载Gemini提供方过滤器中断，之前log step 57仍为done=False，评分器fell=0，未见评分回合已经不可逆判负。推断：这是结果尚未固定时的外因截断，应留待重测，不纳入智能体失败归因，现象和根因均留空。评分器worst_lin_err=0.2614 m/s高于0.2 m/s，worst_yaw_err=0.2735 rad/s低于0.3 rad/s，但这些是未走完整日程的记录，不能据此断言继续执行后仍会失败。视频有末段倾斜，未见翻倒；无法从画面确认实际关节角、躯干高度及逐步速度误差，也未建立控制步到视频钟的精确映射。"
+ },
+ "summary_zh": "评分回合中机器人已迈步移动，视频末段明显倾斜但仍由腿支撑；现场回合只执行365/1000个控制步，按指令行进未完成，最终节点0/1。log step 57仍为done=False，随后log step 59出现Gemini提供方过滤器中断，属于结果尚未固定时的外因截断，保留目录success=0而不标智能体失败现象或根因。缺少状态日志，未建立控制步到视频钟的精确映射，精确速度及终止判据不能由画面补全。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "标签合理：log step 57 现场回合 control_steps=365、done=False，log step 58 仍余185 s，log step 59 为Gemini提供方过滤器中断，结果未固定，属外因截断；节点false、现象与根因留空并写gap_zh，符合FAIL-5/D37。视频无翻倒，时间均未把智能体时间换算成视频时间。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:04:04Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--p3ejwxv/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--p3ejwxv/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "f1453e9f7c850f1603605071841bd386cf6ef73a78e272bbec68ae785cc50b08",
+   "video_sha256": "3eb954a851c5c2fad47064697476e70809d026625c3d1cf50e5efffe5e6237d7",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--shxsrs8.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--shxsrs8.json
new file mode 100644
index 0000000000000000000000000000000000000000..2e67868d25eb79ca80f3d0d2d6baf3c155a792d0
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--shxsrs8.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard--shxsrs8",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 99,
+   "sim_steps": 99,
+   "physics_steps": 495,
+   "steps": 1466,
+   "elapsed_s": 527.2,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.7971,
+   "worst_yaw_err": 0.3627,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.291,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.239,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard--shxsrs8/replay.mp4",
+  "duration_s": 2.0,
+  "fps": 25.0,
+  "frames": 50,
+  "sha256": "9dac77ada6b9ee1e6913c7527b4527df317720a6d860e063533eb7e0a573afa1"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00 s 机器人由足端支起，随后腿部动作并伴随机身移动；视频 1.40–1.96 s 机身逐渐前倾，后部抬高，黄色前部降至地面附近，未恢复正常支撑。整段没有完成两段指令的连续行进及后半段转向。"
+     },
+     {
+      "t": 1.96,
+      "log_step": 24,
+      "text_zh": "观察：log step 24 明确开始最后一次重置后的评分回合，服务返回 episode=11、control_steps=99、done=true、reason=fell、resets_left=0；评分器记录 sim_steps=99、fell=1、live_matches_replay=1。该回合未达到所需1000控制步；评分器 worst_lin_err=0.7971 m/s、worst_yaw_err=0.3627 rad/s，分别超过0.2 m/s和0.3 rad/s阈值。"
+     },
+     {
+      "t": 1.8,
+      "log_step": 24,
+      "text_zh": "推断：结合视频末段前倾失稳与最后评分回合的服务终止记录，该节点从未完成。视频为2.00 s，而99控制步对应仿真0.99 s；没有已验证的时钟映射，证据锚点直接取所检查的视频画面，不把log step 24打印的仿真时间换算为视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–1.40 s，Barkour足端与棋盘格地面接触并支撑机身，腿部尝试迈步；视频 1.60–1.96 s，后部抬高、前腿仍接地，机身向前下方倾倒。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.7971
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.3627
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 1.8,
+   "log_step": 24,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 1.60–1.96 s 机器人行进尝试中机身持续前倾，后部抬起、前部下沉，未恢复支撑；log step 24 的最后评分回合在99控制步后以 reason=fell 终止，未进入后半段命令。断点锚点取视频约1.80 s的明显失稳画面。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.8,
+    "log_step": 24,
+    "evidence_zh": "观察：视频 1.60–1.96 s，Barkour后部逐步抬高，躯干前倾、黄色前部降到地面附近，结束前没有恢复四足支起的正常姿态；log step 24 服务返回 done=true、reason=fell，评分器 fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：agent的足端轨迹与关节位置控制没有实现所需的速度跟踪和支撑稳定。log step 12 将控制改为每10个控制步批量提交再读取反馈；log step 22–23 又在逆解目标上加入按站立关节偏差估计的承载补偿，log step 24 实际以该控制器运行最后评分回合，脚本读取的速度从仿真约0.49 s的(-0.14,-0.05) m/s变为仿真约0.99 s的(-0.98,-0.03) m/s，与第一段(+0.4594,+0.2212) m/s命令相反，并收到fell终止；视频 1.60–1.96 s可见支撑失稳。现有证据支持控制失效，但不能单独证实某个增益、反馈周期或承载补偿就是原因。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有状态日志，无法确定fell究竟由躯干高度、关节边界还是翻倒判据首次触发，也无法检验agent在log step 25提出的软执行器下沉、摆动足拖地解释。视频2.00 s与99控制步对应的仿真0.99 s不同，缺少已核实的时钟映射，所有视频锚点均直接来自逐帧检查；根因保留在控制大类。"
+ },
+ "summary_zh": "评分回合尝试迈步但未完成“按指令行进”，最终0/1节点；视频约1.80 s出现明显前倾失稳，log step 24确认最后回合在99控制步后以fell终止。评分器平移和偏航误差均超阈值，日志与画面支持控制失败，但不能确定具体控制子因。没有状态日志，且视频2.00 s与该回合仿真0.99 s的映射未经核实，失败锚点采用视频画面。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "单节点未达成、断点取该节点、平衡现象与控制大类根因均有视频和评分器/服务读数支持；时钟处理谨慎，未把仿真步数换算成视频时间。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:04:14Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--shxsrs8/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--shxsrs8/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "8f01f69dcb22e73c081c1a81759b66d0c170d355fc140cf4e6d385b425f9d1ad",
+   "video_sha256": "9dac77ada6b9ee1e6913c7527b4527df317720a6d860e063533eb7e0a573afa1",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--v5f2nhs.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--v5f2nhs.json
new file mode 100644
index 0000000000000000000000000000000000000000..30229be23f1de78bc5c3399fa17aa6107e92d470
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard--v5f2nhs.json
@@ -0,0 +1,209 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard--v5f2nhs",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 123,
+   "sim_steps": 123,
+   "physics_steps": 615,
+   "steps": 2944,
+   "elapsed_s": 851.8,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.4717,
+   "worst_yaw_err": 0.501,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard--v5f2nhs/replay.mp4",
+  "duration_s": 2.48,
+  "fps": 25.0,
+  "frames": 62,
+  "sha256": "17ca1c88f853ff082f5bf7c7c100343afc7cdfe20b98dfe239f6f258bc2d055b"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00–1.44 s，Barkour四足机器人由足端支撑在棋盘格地面上，机身先下沉并有小幅腿部运动；没有完成持续平移、随后换向转弯的整段日程。"
+     },
+     {
+      "t": 1.8,
+      "log_step": null,
+      "text_zh": "观察：视频 1.80–2.44 s，机器人躯干明显倾斜，一端抬高、另一端下沉，多个足端抬离地面；末帧仍未恢复平稳支撑。"
+     },
+     {
+      "t": null,
+      "log_step": 55,
+      "text_zh": "观察：log step 55 的服务重置结果为 episode 11、resets_left 0，agent随后执行控制脚本；此前回合均被重置，不属于本次评分轨迹。"
+     },
+     {
+      "t": null,
+      "log_step": null,
+      "text_zh": "推断：inputs/run.json 的评分器记录 sim_steps=123、fell=1、live_matches_replay=1、success=0；本次评分回合未运行满所需1000个控制步，且 worst_lin_err=0.4717 m/s、worst_yaw_err=0.501 rad/s 均超过节点阈值，因此该完整日程节点始终未达成。123个控制步仅用于判断日程未完成，不转换成视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–1.76 s，Barkour的足端接触棋盘格地面承重，腿部屈伸并尝试迈步；视频 1.80–2.44 s，足端支撑变得不稳定，躯干持续倾斜。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.4717
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.501
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 1.8,
+   "log_step": 56,
+   "where": "操作中",
+   "evidence_zh": "观察：视频约 1.80 s 起，正在尝试行进的机器人机身倾斜加剧、部分足端离地，至视频 2.44 s 仍未恢复稳定支撑。log step 55 启动最后的第11回合；log step 56 的同回合脚本输出显示 IMU姿态偏差增长、接触足数减少；评分器最终记录123个控制步及 fell=1，整段行进日程未完成。断点时间直接取视频帧。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.8,
+    "log_step": 56,
+    "evidence_zh": "观察：视频 1.80–2.44 s，机器人躯干由近水平变为明显倾斜，一端下沉，多足抬离地面，无法保持稳定站立。log step 56 的脚本读取到控制步90时足接触为[0,1,0,1]，控制步110时为[0,0,0,1]；评分器记录 fell=1。画面可见失衡，但未显示完整翻转至倒扣。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 55 记录agent在最后回合运行调整后的步态控制脚本；log step 56 的脚本传感读数中，指令约为[0.46,0.22,0.02]，控制步80的平面速度约[-0.01,-0.04]，控制步110约[0.02,-0.09]，IMU重力方向的水平分量由控制步80的[0.10,-0.02]增至控制步110的[0.42,0.28]，足接触由四足减至单足。结合视频 1.80–2.44 s 的持续倾斜，所执行控制未能同时建立所需运动并稳定机身；证据支持行进控制失稳，但不足以指定控制律或参数子类。"
+   }
+  ],
+  "gap_zh": "无逐帧状态日志；评分器的 fell 汇总字段未区分躯干翻倒、实际关节越界和躯干过低，视频也不能量出这些精确阈值，因此不声称触发了哪一条底层终止条件。日志中的控制脚本修改仅保留不完整的调用文本，无法核实具体控制律、参数或推扰发生时刻；agent在 log step 58 自述“第一次推扰中跌倒”不能单独证明推扰是原因。本视频时长2.48秒与123个控制步对应的仿真1.23秒不同，未提供明确播放映射，所有视频锚点均直接读帧。"
+ },
+ "summary_zh": "评分的第11回合只完成起始支撑及短暂腿部运动，“按指令行进”未达成，最终0/1；此前跑满日程的测试已被重置。机器人在视频约 1.80 s 起失衡且未恢复，评分器记录123个控制步、fell=1，两项速度误差也超阈值，日志支持行进控制失稳。缺少逐帧状态及完整控制脚本，无法确定具体终止判据或控制参数原因。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成、断点取视频约1.8 s失衡、平衡现象与控制根因均有帧和评分器字段(fell=1、123步、两项误差超阈值)支持，未发现需改内容的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:22:32Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--v5f2nhs/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard--v5f2nhs/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "d226dd10381509e08149f6f7eec39875fcd75052962513477efe90c50b1a18b1",
+   "video_sha256": "17ca1c88f853ff082f5bf7c7c100343afc7cdfe20b98dfe239f6f258bc2d055b",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--3jnmac4.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--3jnmac4.json
new file mode 100644
index 0000000000000000000000000000000000000000..259f59de8acea4e42c813380f998ed593c7508a7
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--3jnmac4.json
@@ -0,0 +1,187 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard-realtime--3jnmac4",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "finalized": 1,
+   "missing_log": 0,
+   "infra_error": 0,
+   "grader_error": 0,
+   "sim_error": 0,
+   "sim_crashed_on_call": 0,
+   "n_actions": 0,
+   "sim_steps": 0,
+   "calls": 13,
+   "rows": 0,
+   "resets": 0,
+   "reset_budget": 10,
+   "sim_startups": 1,
+   "elapsed_s": 261.0,
+   "video_rendered": 1,
+   "realtime": 1,
+   "slowdown": 1.0,
+   "clock_s": 0.0,
+   "max_lag_s": 0.0,
+   "stalls": 0,
+   "stalled_s": 0.0,
+   "stalled_share": 0.0,
+   "held_steps": 0,
+   "late_steps": 0,
+   "requests_refused": 0,
+   "request_busy_s": 0.009,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.175,
+   "peak_processes": 6,
+   "peak_disk_gb": 0.249,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard-realtime--3jnmac4/replay.mp4",
+  "duration_s": 0.04,
+  "fps": 25.0,
+  "frames": 1,
+  "sha256": "db7c3bad0cdaa29b631ba80a4dc0951a25e850225f7b2dea17374a001258cc6c"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：本次视频全长0.04 s、只有视频0.00 s的一帧；左侧跟随视角显示Barkour处于初始站姿，右侧头部视角显示棋盘格地面，没有可供观察的迈步、平移或转向过程。"
+     },
+     {
+      "t": null,
+      "log_step": 14,
+      "text_zh": "观察：log step 14中机器人服务返回running: False；该次调用为get_obs(seconds=0.0)。log step 3的服务状态为control_steps=0、planned_steps=0、running=false；评分器最终记录sim_steps=0、n_actions=0、rows=0、clock_s=0.0。"
+     },
+     {
+      "t": null,
+      "log_step": 15,
+      "text_zh": "推断：log step 15记录Gemini过滤器拦截，此前全部机器人请求仅查询状态或零时长观测，未提交动作；结合视频0.00 s初始画面及评分器sim_steps=0，可确定完整1000控制步指令日程未执行，节点从未达到。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "不适用",
+      "how": "不适用",
+      "note_zh": "观察：评分器记录n_actions=0、rows=0、sim_steps=0，log step 14仍为running: False；仅有视频0.00 s初始画面，没有交付行进轨迹，按无交付规则标不适用。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": null
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "unknown",
+   "t": null,
+   "log_step": 15,
+   "where": "首步前",
+   "evidence_zh": "观察：log step 15的最后消息为“This request was blocked by Gemini's filters”，未附动作调用；log step 14的机器人服务仍返回running: False，评分器记录n_actions=0、sim_steps=0。推断：模型提供方过滤器在回合启动和轨迹交付之前截断会话，无法定位机器人执行中的失败步骤，按外部中断规则标unknown／首步前。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [],
+  "root_causes": [],
+  "gap_zh": "观察：log step 15记录Gemini提供方过滤器拦截，log step 14仍未启动回合；inputs/replay.mp4只有视频0.00 s一帧，无状态日志，评分器未报告worst_lin_err和worst_yaw_err。推断：没有已执行的行进尝试可供判定物理失败现象或机器人根因，按CASES.md FAIL-5作为外部中断留空两层，不能标为没尝试或控制失败；目录中的success=0保留为原始评分结果，不将本次纳入代理失败归因。"
+ },
+ "summary_zh": "观察：本次仅有视频0.00 s初始站姿，评分器记录零动作、零仿真步，“按指令行进”未达到，最终节点0/1。观察：log step 14的服务仍未运行，log step 15随即记录Gemini过滤器拦截；推断：这是首步前的外部中断，断点为unknown，现象与根因留空，不作机器人失败归因。观察：无状态日志且两项速度误差均未报告，无法评价行进控制质量。",
+ "conf": 10,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 9,
+    "summary_zh": "标注符合FAIL-5：Gemini过滤器在未提交任何动作时截断会话，节点未达到、断点unknown/首步前、两层为空并附gap_zh，视频仅一帧初始站姿，与评分器n_actions=0、sim_steps=0一致。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:22:37Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--3jnmac4/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--3jnmac4/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "3318f2942f87e21b02ecd16e53f96d978f2a1bc048a07cb78e5801f127e0a3da",
+   "video_sha256": "db7c3bad0cdaa29b631ba80a4dc0951a25e850225f7b2dea17374a001258cc6c",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--5jjbxvh.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--5jjbxvh.json
new file mode 100644
index 0000000000000000000000000000000000000000..8dc8c5b7e62dce94c5ed384cbec1460cfb3b659b
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--5jjbxvh.json
@@ -0,0 +1,225 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard-realtime--5jjbxvh",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "finalized": 1,
+   "missing_log": 0,
+   "infra_error": 0,
+   "grader_error": 0,
+   "sim_error": 0,
+   "sim_crashed_on_call": 0,
+   "n_actions": 578,
+   "sim_steps": 7053,
+   "calls": 9544,
+   "rows": 57622,
+   "resets": 10,
+   "reset_budget": 10,
+   "sim_startups": 1,
+   "elapsed_s": 439.5,
+   "video_rendered": 1,
+   "realtime": 1,
+   "slowdown": 1.0,
+   "clock_s": 70.44,
+   "max_lag_s": 0.0035,
+   "stalls": 0,
+   "stalled_s": 0.0,
+   "stalled_share": 0.0,
+   "held_steps": 0,
+   "late_steps": 5,
+   "requests_refused": 10,
+   "request_busy_s": 3.981,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.148,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.24,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard-realtime--5jjbxvh/replay.mp4",
+  "duration_s": 5.8,
+  "fps": 25.0,
+  "frames": 145,
+  "sha256": "71d598d6c7b3b7e59d8116cd6aa6c0cb3cfc4b40cc828cc086982dcaa40ea844"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 2.8,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0–5.0 s，Barkour四足机器人交替抬腿、落足并在棋盘格地面移动；视频 2.0 s 机身近侧向，视频 2.8–3.0 s 转成较正对跟随相机，视频 4.0 s 又回到近侧向，行进朝向明显摆动。"
+     },
+     {
+      "t": 5.76,
+      "log_step": 27,
+      "text_zh": "观察：视频 4.80–5.76 s 逐帧仍可见机器人迈步及腿部支撑，录像随即结束，未展示完整10秒日程；log step 27 中仿真服务状态为episode=11、control_steps=578、done=True、reason=failed，且与评分器n_actions=578一致。"
+     },
+     {
+      "t": 5.76,
+      "log_step": 27,
+      "text_zh": "推断：log step 2 的服务规格给出 control_dt=0.01 s，log step 27 最后第11回合的578个控制步对应仿真5.78 s，与 inputs/video.json 的视频总长5.8 s 仅在时长上吻合；视频中看不到可辨认的指令切换事件，因此没有独立可见事件核对控制步到视频时间的映射，不据此给出精确的视频换指令时刻。该回合未运行满1000步，故完整日程节点从未成立。成功参考视频后半程持续侧移、转向并运行到参考视频约10秒，仅用来比较完成外观。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：本次视频 0.0–5.76 s，Barkour的四条腿足端交替落在棋盘格地面上，承重并推进躯干；视频末帧仍可见腿部接地支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": null
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 5.76,
+   "log_step": 27,
+   "where": "操作中",
+   "evidence_zh": "观察：log step 27 中机器人服务的 final 状态给出 episode=11、control_steps=578、done=True、reason=failed，与评分器 n_actions=578 一致；视频5.0–5.76 s 仍可见机器人迈步且机身由腿部支起，随后录像结束，未见躯干翻倒。推断：以已检查的视频末帧5.76 s 作为“按指令行进”失败终止附近的粗略导航锚点，不再用视频2.8 s 的朝向摆动定位终止；按服务 control_dt=0.01 s，终止为仿真5.78 s，较指令日程在仿真5.01 s 的切换晚约0.77 s，但视频时间映射仅有时长吻合，具体终止触发条件无法确定。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 5.76,
+    "log_step": 27,
+    "evidence_zh": "观察：视频5.0–5.76 s，Barkour仍迈步并由腿部支起，录像在完整日程完成前结束；log step 27 的服务状态确认第11回合在578个控制步以 reason=failed 终止。视频2.0 s 机器人近侧向，视频2.8–3.0 s 转为较正对跟随相机，视频4.0 s 又转回，这是较早的方向摆动观察，不作为终止时刻。推断：按可见行进及未完成整段日程标为全身·步态；现有证据不能将其具体化为翻倒或失去支撑。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "log_step": 27,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 5 的控制器代码以周期腿部目标叠加机体线速度和偏航角速度反馈，log step 17 增加关节跟踪补偿；log step 26–27 修改并启用横滚补偿，log step 27 的最终参数相较 log step 21 降低了部分反馈增益。log step 28 中 agent 自述这些最终改动未经测试，此自述只说明其调试判断，不证明具体参数错误。视频2.0–4.0 s 可见行进朝向摆动；log step 27 服务记录的最终回合在578步失败，按控制周期和指令日程推得终止在仿真5.01 s 换指令后约0.77 s。结合控制器结构、可见摆动与换指令后的提前终止，将反馈控制未稳定完成跟踪作为根因推断；不采用 agent 脚本的误差统计，也不能确定某个增益或横滚补偿符号导致了终止。"
+   }
+  ],
+  "gap_zh": "观察：inputs/没有状态日志；run.json 的 reward 未报告 worst_lin_err、worst_yaw_err 或具体终止触发条件，故两项 native_metrics 值保留 null，不能确定是否触发关节越界、躯干过低或翻倒阈值。agent 在 log step 28 自述“倒地”，但视频最后一秒至视频5.76 s 仍可见机身由腿部支起，画面不足以确认摔倒，不采用该自述作为平衡现象。推断：578步按控制周期对应仿真5.78 s，与视频总长5.8 s 仅在时长上吻合；视频没有可辨认的指令切换事件，故没有独立事件验证其时间映射。反馈控制归因有控制器代码和可见摆动支持，但换指令后的具体终止机制、各参数和推扰的影响均无法隔离。",
+  "unmet": {
+   "joystick_schedule_followed": false
+  }
+ },
+ "summary_zh": "评分采用最后第11回合：机器人以足端支撑迈步移动，但未完成完整“按指令行进”日程，节点最终为0/1；断点粗锚在视频5.76 s 末段，服务记录578步后以 failed 终止，按控制周期及日程推得终止在仿真5.01 s 换指令后约0.77 s。视频2.0–4.0 s 的朝向摆动只作为次要观察，结合控制器代码推断反馈跟踪未稳，不能确认具体不当参数。因无状态日志、原生误差和具体终止字段，视频未证实倒地且时钟映射仅有时长吻合，关键时刻对应关系及根因仍含推断，置信度降为5。",
+ "conf": 5,
+ "review": {
+  "status": "agreed",
+  "rounds": 2,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "fix",
+    "conf": 6,
+    "summary_zh": "节点值（未达成）、contact_actual 和根因方向基本可接受；但断点与现象锚在视频 2.8 s 的朝向摆动，而服务读数显示回合在换指令（约 5.01 s）后约 0.77 s、视频约 5.76 s 以 reason=failed 终止，这才是任务真正失败的位置，需要改锚点并澄清证据。",
+    "n_issues": 3,
+    "n_major": 1
+   },
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 2,
+    "verdict": "agree",
+    "conf": 6,
+    "summary_zh": "第二轮修改后节点未达成、断点锚在服务读数的 578 步终止（视频约 5.76 s）、时钟映射标为推断、根因证据不再依赖脚本统计，均与视频和 log step 27 一致；未发现需改内容的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:26:20Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--5jjbxvh/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--5jjbxvh/04_review_r1"
+   },
+   {
+    "stage": "reconcile",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--5jjbxvh/05_reconcile_r1"
+   },
+   {
+    "stage": "traj_review",
+    "round": 2,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--5jjbxvh/04_review_r2"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "96e9a3e8b4e964673dcceb3c9d893bc14b694cb5f4c52c702bfa9ef66f571553",
+   "video_sha256": "71d598d6c7b3b7e59d8116cd6aa6c0cb3cfc4b40cc828cc086982dcaa40ea844",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--bjj83yu.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--bjj83yu.json
new file mode 100644
index 0000000000000000000000000000000000000000..d162de501b640ada06827715ba7f7cdbbecabe59
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--bjj83yu.json
@@ -0,0 +1,202 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard-realtime--bjj83yu",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "finalized": 1,
+   "missing_log": 0,
+   "infra_error": 0,
+   "grader_error": 0,
+   "sim_error": 0,
+   "sim_crashed_on_call": 0,
+   "n_actions": 127,
+   "sim_steps": 7188,
+   "calls": 3715,
+   "rows": 3693,
+   "resets": 10,
+   "reset_budget": 10,
+   "sim_startups": 1,
+   "elapsed_s": 468.2,
+   "video_rendered": 1,
+   "realtime": 1,
+   "slowdown": 1.0,
+   "clock_s": 71.808,
+   "max_lag_s": 0.0027,
+   "stalls": 0,
+   "stalled_s": 0.0,
+   "stalled_share": 0.0,
+   "held_steps": 3510,
+   "late_steps": 0,
+   "requests_refused": 4,
+   "request_busy_s": 1.312,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard-realtime--bjj83yu/replay.mp4",
+  "duration_s": 2.56,
+  "fps": 25.0,
+  "frames": 64,
+  "sha256": "f80beb053ccdd7d19cc17222eddca61bb5390e38ab2d9fd8881973a4a2820c30"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.8,
+      "log_step": 55,
+      "text_zh": "观察：视频 0.00–0.80 s，Barkour机器人由足端支撑并交替迈腿，机身随步态起伏；log step 55显示最后一次重置进入episode 11，控制器开始执行本回合。"
+     },
+     {
+      "t": 2.52,
+      "log_step": 56,
+      "text_zh": "观察：视频 2.08–2.52 s，机器人机身持续倾斜、一侧降低，部分腿抬高或外伸，未恢复起始的腿部支撑姿态；log step 56的机器人服务状态为episode 11、control_steps=127、done=True、reason=failed，评分记录n_actions=127。"
+     },
+     {
+      "t": 2.52,
+      "log_step": 56,
+      "text_zh": "推断：最后回合仅执行127个控制步即终止，未完成节点要求的1000步及后半段指令，因此全程没有满足完整日程节点；会话中另一次1000步测试（log step 28–32）不属于这段评分视频。视频锚点按帧号/25取得，没有把控制步或agent工作时间换算成视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–2.00 s，Barkour机器人的足端交替落在棋盘格地面上，支撑机身并迈步；视频 2.08–2.52 s机身倾斜降低，足端支撑未能维持起始姿态。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": null
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 2.08,
+   "log_step": 56,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 2.08–2.52 s，正在迈步的Barkour机器人机身逐渐侧倾并降低，末帧仍未恢复；log step 56的机器人服务确认最后回合episode 11仅到control_steps=127便以failed终止，未进入第二段指令，更未完成完整行进日程。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 2.08,
+    "log_step": 56,
+    "evidence_zh": "观察：视频 2.08–2.52 s，机器人机身明显侧倾、靠近地面，一侧腿抬起，未重新建立起始的支撑姿态；log step 56打印的最后回合控制步125重力方向为[-0.483,0.413,-0.772]，随后服务在控制步127判为failed。这里标记可见的失衡，不把它当作已证实超过90度翻倒。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：agent的关节步态控制器未能在跟随首段速度指令时维持机体稳定。log step 55–56的最后回合输出中，指令始终为[0.459,0.221,0.016]，实际前向速度却从控制步75的-0.448变到控制步100的0.263、控制步125的-0.537，重力方向也从[-0.218,0.076,-0.973]发展为[-0.483,0.413,-0.772]；视频 2.08–2.52 s可见机身倾斜降低。日志支持步态控制失稳，但未给出完整控制器代码及逐步目标，不能确定是接触布局、反馈律还是某个参数造成。",
+    "source_zh": "日志 / 视频"
+   }
+  ],
+  "gap_zh": "没有状态日志，最后回合的服务只报告failed，未给出触发终止的具体关节、躯干高度或翻倒谓词，故不判定是哪条约束触发。评分字段未报告worst_lin_err与worst_yaw_err，均保留null；会话测试的误差不能替代最后回合的原生指标。控制周期为0.01 s、最后回合127步，而视频长2.56 s，二者没有已核实的逐步映射，视频锚点均直接取自所检查帧。控制根因仅能确定到大类。"
+ },
+ "summary_zh": "观察：评分视频中的Barkour机器人已开始迈步，但视频 2.08–2.52 s机身明显倾斜降低，log step 56确认最后回合只执行127个控制步便失败，“按指令行进”节点最终为0/1。推断：关节步态控制未能维持机体稳定，导致完整指令日程未完成；先前的1000步测试不计入本回合。没有状态日志或明确的终止谓词，无法确认具体触发约束与控制子类，两项原生误差指标也未报告。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "标签与证据一致：评分视频（2.56 s）中机器人迈步后机身侧倾降低，log step 56 服务报告 episode 11 在 control_steps=127 以 failed 终止，节点 joystick_schedule_followed 为 false；断点、平衡现象和停在“控制”大类的根因均有证据支持，时钟未做换算。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:24:01Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--bjj83yu/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--bjj83yu/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "3ef3ffd43af5719f9de53a462b388a18682d12662a95be3a7a4180f11984673f",
+   "video_sha256": "f80beb053ccdd7d19cc17222eddca61bb5390e38ab2d9fd8881973a4a2820c30",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--hfgptfp.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--hfgptfp.json
new file mode 100644
index 0000000000000000000000000000000000000000..04d1fc7a5183153d626fcafad877bca145630f9f
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--hfgptfp.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard-realtime--hfgptfp",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "finalized": 1,
+   "missing_log": 0,
+   "infra_error": 0,
+   "grader_error": 0,
+   "sim_error": 0,
+   "sim_crashed_on_call": 0,
+   "n_actions": 80,
+   "sim_steps": 3662,
+   "calls": 326,
+   "rows": 2913,
+   "resets": 10,
+   "reset_budget": 10,
+   "sim_startups": 1,
+   "elapsed_s": 714.6,
+   "video_rendered": 1,
+   "realtime": 1,
+   "slowdown": 1.0,
+   "clock_s": 36.526,
+   "max_lag_s": 0.0032,
+   "stalls": 0,
+   "stalled_s": 0.0,
+   "stalled_share": 0.0,
+   "held_steps": 2802,
+   "late_steps": 0,
+   "requests_refused": 28552,
+   "request_busy_s": 1.55,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.2,
+   "peak_processes": 13,
+   "peak_disk_gb": 0.273,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard-realtime--hfgptfp/replay.mp4",
+  "duration_s": 1.64,
+  "fps": 25.0,
+  "frames": 41,
+  "sha256": "a039d7d420a3a745b1e607407a10322d4445bacd34ccc84bad050adda643b558"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00 s，Barkour四足机器人由足端支起，尚未完成行进日程；完整日程是该节点的验收条件，开局站立不计为到达。"
+     },
+     {
+      "t": 1.2,
+      "log_step": null,
+      "text_zh": "观察：逐帧检查视频 0.00–1.60 s（全部41帧），机器人主要在原处调整腿部；视频 0.72、1.20、1.56 s 可见腿部逐渐弯折，机身降低，末段未重新撑高，也未出现完整平移和后半段明显转向。"
+     },
+     {
+      "t": null,
+      "log_step": 45,
+      "text_zh": "观察：log step 44 返回最后一次重置、resets_left=0；log step 45 的机器人服务状态返回 episode=11、control_steps=80、done=True、reason=failed，log step 47 再次确认相同终态。服务状态是现场回合读数；该回合未运行满1000个控制步，不能通过完整日程验收。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–1.60 s，Barkour足端接触棋盘格地面承重，腿关节发生调整并逐渐弯折；末段机身降低。可见的是足端与地面的支撑交互，未据阴影判定躯干触地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": null
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 1.2,
+   "log_step": 45,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 0.72–1.56 s，机器人腿部逐渐弯折、机身降低，未恢复到开局支撑高度；行进与转向日程未完成。log step 45 的现场服务记录最后回合 episode=11 在80个控制步终止为 failed，log step 47 确认未恢复。视频 1.20 s 为可见支撑塌低的粗略导航锚点，未由日志工作时间换算。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.2,
+    "log_step": null,
+    "evidence_zh": "观察：对比视频 0.00、0.72、1.20、1.56 s，机器人承重腿越来越弯，足端附近的腿段接近地面，机身明显降低；末段未重新撑起。推断：机器人未保持足够的腿部承重支撑，属于支撑塌低的平衡失败；画面未显示机身翻转超过90度，不将其描述为侧翻。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": null,
+    "log_step": 42,
+    "evidence_zh": "观察：log step 42–43 的控制脚本通过 joint_position 提交腿关节目标，含外展正反探测及后续步态、姿态修正逻辑；视频 0.72–1.56 s 显示腿部调整后支撑高度持续降低。log step 45 确认最后现场回合仅运行80个控制步即失败；log step 46 的 EpisodeOver 异常明确发生于服务已判 failed 后。推断：该回合的腿部控制未能维持承重姿态，根因停留在控制层；不能把回合结束后的异常当成引发塌低的原因，也不能据此前调试回合或agent最终自述确定某一增益、抬腿方向或反馈项有误。",
+    "source_zh": "日志 / 视频"
+   }
+  ],
+  "gap_zh": "观察：inputs/提供了评分视频、完整会话记录及成功参考视频，但没有状态日志；reward未报告worst_lin_err或worst_yaw_err，故两项原生指标均为null。最后现场回合的服务状态只给reason=failed，没有明确是哪条终止条件触发，也没有该回合逐步高度、实际关节角或速度记录。log step 3 的control_dt=0.01 s与log step 45 的80个控制步对应仿真 0.80 s，而评分视频长度为1.64 s；未建立经可见事件核对的两者映射，因此所有视频锚点直接取自帧，未将仿真时间套入视频。log step 46 的异常位于探测提交处，不能确认最终脚本已进入主步态循环；log step 53 所称摆腿俯仰原因仍是agent自述，缺少最后回合数据验证。",
+  "unmet": {
+   "joystick_schedule_followed": false
+  }
+ },
+ "summary_zh": "最后评分回合中，Barkour足端接地并调整腿部，但随后腿部弯折、机身塌低，未完成“按指令行进”，节点最终0/1。断点在该步骤操作中，现象为平衡失败，根因仅支持宽泛的控制问题；服务在log step 45确认最后回合80/1000控制步即失败。没有状态日志或最后回合的精确终止读数，两项速度误差指标未报告，控制步与视频时间的映射也未核实。",
+ "conf": 7,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点0/1、断点（操作中）、平衡现象与停在“控制”的根因均与视频和服务读数一致；视频 0.00→1.56 s 可见腿弯折、机身降低，log step 45/47 的服务读数确认80步failed。未发现需要修改的内容。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:23:58Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--hfgptfp/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--hfgptfp/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "2a06b143ff638b265d892efdb7f09e63402a5c52dcb306567972761c64b80b6a",
+   "video_sha256": "a039d7d420a3a745b1e607407a10322d4445bacd34ccc84bad050adda643b558",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--kcephp9.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--kcephp9.json
new file mode 100644
index 0000000000000000000000000000000000000000..7b3a2fde6918ea54fe9a1dda0725aabe23bb0428
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--kcephp9.json
@@ -0,0 +1,202 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard-realtime--kcephp9",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "finalized": 1,
+   "missing_log": 0,
+   "infra_error": 0,
+   "grader_error": 0,
+   "sim_error": 0,
+   "sim_crashed_on_call": 0,
+   "n_actions": 324,
+   "sim_steps": 3834,
+   "calls": 12135,
+   "rows": 8913,
+   "resets": 10,
+   "reset_budget": 10,
+   "sim_startups": 1,
+   "elapsed_s": 1102.1,
+   "video_rendered": 1,
+   "realtime": 1,
+   "slowdown": 1.0,
+   "clock_s": 38.244,
+   "max_lag_s": 0.0033,
+   "stalls": 0,
+   "stalled_s": 0.0,
+   "stalled_share": 0.0,
+   "held_steps": 1034,
+   "late_steps": 0,
+   "requests_refused": 32,
+   "request_busy_s": 3.254,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.388,
+   "peak_processes": 10,
+   "peak_disk_gb": 1.059,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard-realtime--kcephp9/replay.mp4",
+  "duration_s": 4.36,
+  "fps": 25.0,
+  "frames": 109,
+  "sha256": "a94769a7d04ec5284c712ea18c8d4e51fc3db9e1b43a3df2ea31484b9dafa983"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.8,
+      "log_step": null,
+      "text_zh": "观察：视频0.0–1.6 s，跟随视角中Barkour机器人反复迈腿、足端接触棋盘格地面并移动，已开始行进；起始站立和短时迈步不满足完整日程的节点判据。"
+     },
+     {
+      "t": 3.4,
+      "log_step": null,
+      "text_zh": "观察：视频3.2–3.76 s，机器人机身明显倾斜；视频3.8–4.32 s，机身继续下沉，末帧呈低伏姿态，没有恢复正常站立行进。"
+     },
+     {
+      "t": null,
+      "log_step": 113,
+      "text_zh": "观察：log step 113的机器人服务status返回episode=11、control_steps=324、done=true、reason=failed、resets_left=0；log step 115及120再次确认同一终止状态。该评分回合未运行到要求的1000个控制步，故完整日程节点从未成立；此前回合的测试结果不计入。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "视频0.0–3.2 s可见Barkour四条腿反复屈伸，足端交替落地承重并推动机身行进；视频3.4–4.32 s机身倾斜下沉，正常腿部支撑未维持。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": null
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 3.4,
+   "log_step": 111,
+   "where": "操作中",
+   "evidence_zh": "观察：视频3.2–3.76 s，正在迈步的Barkour机器人出现明显机身倾斜，随后在视频3.8–4.32 s下沉至低伏姿态，未恢复正常行进。log step 111启动最后一次步态控制后收到episode finished、reason failed；log step 113确认该回合为episode 11，止于control_steps=324，未完成1000步日程。"
+  },
+  "unmet": {
+   "joystick_schedule_followed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 3.4,
+    "log_step": null,
+    "evidence_zh": "观察：视频3.2–3.76 s，机器人机身明显倾斜，腿部未能使机身保持平稳；视频3.8–4.32 s，机身逐渐下沉，最后低伏在地面附近。这一失稳在视频结束前没有修复。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 97展示agent的步态脚本以控制步更新相位、由速度误差调整步幅，并以IMU重力及角速度修正膝关节目标；log step 111修改俯仰反馈项后运行最后一回合，仍收到失败终止。结合视频3.2–4.32 s的倾斜和下沉，实际步态控制未能维持机身稳定，归为控制；没有足够证据将最终失稳唯一归于某个增益、符号或支撑布局。"
+   }
+  ],
+  "gap_zh": "没有状态日志。log step 113的服务状态只报告reason=failed，未给出最后回合的关节角、躯干高度、向上轴分量或终止子原因，不能确认具体是哪条终止规则触发；run.json的reward也没有worst_lin_err或worst_yaw_err，二者保留null。服务control_steps=324按0.01秒控制周期对应仿真3.24 s，但视频时长为4.36 s，最后可见帧为视频4.32 s，未提供经核实的时钟映射；所有视频锚点均直接取自画面，未把仿真3.24 s当作视频时间。早先回合的数值与agent关于失稳机理的自述不足以确定最终回合的具体控制子类。"
+ },
+ "summary_zh": "最后评分回合已开始迈步行进，但在视频约3.4 s出现明显倾斜，随后下沉，按指令行进节点未达成（0/1）；log step 113确认episode 11在控制步324失败，未运行完1000步日程。步态控制未能维持机身平衡，具体控制子类证据不足；参考成功视频可见机器人持续支撑行进并在后半段转向，本次未完成该过程。没有状态日志或两项速度误差评分值，且视频时长与控制步对应的仿真时长不同，无法确定精确终止条件和时钟映射。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成、断点、平衡现象和停在“控制”的根因均有视频与服务读数支持；时钟未把控制步换算成视频时间，未发现需改内容的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:22:38Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--kcephp9/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--kcephp9/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "7aa559700ec4e92cd5395177ed4d2ff9739784b25fc52f9cdefb573378b6e537",
+   "video_sha256": "a94769a7d04ec5284c712ea18c8d4e51fc3db9e1b43a3df2ea31484b9dafa983",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--tebn89a.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--tebn89a.json
new file mode 100644
index 0000000000000000000000000000000000000000..d5a05ec0606a135d200590ac6884ff526beb267a
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--tebn89a.json
@@ -0,0 +1,183 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard-realtime--tebn89a",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "live_success": 1,
+   "replay_success": 1,
+   "deterministic": 1,
+   "finalized": 1,
+   "missing_log": 0,
+   "infra_error": 0,
+   "grader_error": 0,
+   "sim_error": 0,
+   "sim_crashed_on_call": 0,
+   "n_actions": 1000,
+   "sim_steps": 5131,
+   "calls": 10565,
+   "rows": 10556,
+   "resets": 6,
+   "reset_budget": 10,
+   "sim_startups": 1,
+   "elapsed_s": 1010.0,
+   "video_rendered": 1,
+   "realtime": 1,
+   "slowdown": 1.0,
+   "clock_s": 51.255,
+   "max_lag_s": 0.0075,
+   "stalls": 0,
+   "stalled_s": 0.0,
+   "stalled_share": 0.0,
+   "held_steps": 159,
+   "late_steps": 0,
+   "requests_refused": 7,
+   "request_busy_s": 3.527,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard-realtime--tebn89a/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "2d816ba833ffd899746638326d80211ff927554bcf235d370f1f98e10b108014"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [
+     [
+      10.14,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 2.0,
+      "log_step": null,
+      "text_zh": "观察：已查看本次视频全部143帧；视频0.00–5.00 s，Barkour四足机器人持续交替迈腿，棋盘格地面相对机器人移动，躯干由腿支起。"
+     },
+     {
+      "t": 6.5,
+      "log_step": null,
+      "text_zh": "观察：视频5.14–6.79 s，机器人继续迈步，头部和躯干的朝向逐渐改变；视频7.00–10.14 s转向持续，头部由朝画面右侧逐渐转向左前方。视频9.14–10.14 s逐帧可见机器人仍由腿支撑，未见翻倒。"
+     },
+     {
+      "t": null,
+      "log_step": 53,
+      "text_zh": "观察：log step 52中agent脚本重置并启动最后一次控制回合；log step 53中机器人仿真服务的status（现场回合状态接口）报告episode=7、control_steps=1000、done=True、reason=success。inputs/run.json的评分器字段同时记录success=live_success=replay_success=1、deterministic=1、n_actions=1000；此前调试回合的失败不属于本次评分视频。"
+     },
+     {
+      "t": 10.14,
+      "log_step": 53,
+      "text_zh": "推断：视频完整行进、后半段转向及结束时保持支撑，与最终现场回合和评分器的成功结果一致，故完整指令日程及其误差验收已通过；以已检查的视频末帧10.14 s作为节点达成的近似锚点。没有将agent工作时间换算为视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.50–10.14 s，Barkour机器人四条腿的足端交替落在棋盘格地面上，支撑躯干并完成平移和转向。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": null
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": null,
+ "summary_zh": "本次评分轨迹在视频约10.14 s完成“按指令行进”，最终1/1节点成立；机器人前半段平移，后半段持续转向，直到视频末帧仍由腿支撑。log step 53的服务状态确认最终第7回合运行满1000个控制步并成功，评分器的现场与回放结果均为成功。没有状态日志，评分器reward也未单独报告worst_lin_err和worst_yaw_err，因此精确原生误差值保留为null；log step 53中agent脚本自行统计的误差不替代评分器原生指标。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "评分成功（success=1，live/replay均为1），目标节点final=true、无失败层，与视频末帧机器人仍由腿支撑一致；contact_actual为腿/躯干支撑，合理。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:23:28Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--tebn89a/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--tebn89a/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "1bfd20b9311400a702b212990d19b4d81a478c7809ba65d5b0b514d51ced90ef",
+   "video_sha256": "2d816ba833ffd899746638326d80211ff927554bcf235d370f1f98e10b108014",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--vyhemke.json b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--vyhemke.json
new file mode 100644
index 0000000000000000000000000000000000000000..80a2ced70efd1800f93d4e31bbecf959f79b08c4
--- /dev/null
+++ b/phase2/mujoco-playground/barkour_joystick/trajectories/mujoco_playground_barkour_joystick-standard-realtime--vyhemke.json
@@ -0,0 +1,208 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_barkour_joystick-standard-realtime--vyhemke",
+ "taskKey": "mujoco-playground/barkour_joystick",
+ "dag_sha256": "e24d0f402dbcad765944ffad94e7b301b234603065e44e6cf6bb76b5bbdcfb00",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "finalized": 1,
+   "missing_log": 0,
+   "infra_error": 0,
+   "grader_error": 0,
+   "sim_error": 0,
+   "sim_crashed_on_call": 0,
+   "n_actions": 1000,
+   "sim_steps": 10128,
+   "calls": 9156,
+   "rows": 9178,
+   "resets": 10,
+   "reset_budget": 10,
+   "sim_startups": 1,
+   "elapsed_s": 782.2,
+   "video_rendered": 1,
+   "realtime": 1,
+   "slowdown": 1.0,
+   "clock_s": 101.191,
+   "max_lag_s": 0.003,
+   "stalls": 0,
+   "stalled_s": 0.0,
+   "stalled_share": 0.0,
+   "held_steps": 950,
+   "late_steps": 0,
+   "requests_refused": 18,
+   "request_busy_s": 4.207,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.202,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.286,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_barkour_joystick-standard-realtime--vyhemke/replay.mp4",
+  "duration_s": 10.214,
+  "fps": 14.0,
+  "frames": 143,
+  "sha256": "8c8fa91a9b6a395a850d896042f6a9519f5bd087aaf2552171e2cd7e6507b3fc"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "joystick_schedule_followed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.0 s机器人以四足支撑站在棋盘格地面；节点要求完整日程及两段跟踪验收，开局站立不代表已完成。"
+     },
+     {
+      "t": 2.5,
+      "log_step": null,
+      "text_zh": "观察：视频0.5–3.5 s机器人腿部交替运动，足端接触地面支撑机身；黄色前端由画面右侧逐渐转向相机，前半段已发生明显偏航。"
+     },
+     {
+      "t": 5.5,
+      "log_step": null,
+      "text_zh": "观察：视频5.0–10.143 s机器人继续由腿支起，机身方向仅小幅变化，未出现持续的大幅转向，最后一秒及末帧也没有翻倒。成功参考视频5.5–10.0 s可见机身持续转向；该参考仅用于识别可见动作。"
+     },
+     {
+      "t": 10.143,
+      "log_step": null,
+      "text_zh": "推断：本次run.json的reward记录n_actions=1000、success=0且infra_error=0；结合视频前半段明显偏转、后半段转向不足，完整按指令行进的节点未达到。评分字段未提供两段误差均值，不能把走满日程或未倒地等同于跟踪达标，也不把前次测试的自算误差当成本次评分值。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.0–10.143 s，Barkour四条腿的足端接触棋盘格地面，腿部交替活动并支撑躯干；没有手臂或抓手参与。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": null
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "joystick_schedule_followed",
+   "t": 2.0,
+   "log_step": null,
+   "where": "操作中",
+   "evidence_zh": "观察：视频1.5–3.5 s，机器人在第一段指令期间已从侧向相机的姿态明显转向相机；视频5.5–10.143 s又未呈现后半段所需的持续转向。该行进步骤一直未完成，断点以视频2.0 s的明显偏转作为粗略导航锚点，不表示精确误差阈值在此刻首次被突破。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 2.0,
+    "log_step": null,
+    "evidence_zh": "观察：视频1.5–5.0 s，机器人边动腿边明显改变朝向；视频5.5–10.143 s，机身却基本维持同一朝向，仅有小幅摆动，未像成功参考视频后半段那样持续旋转。推断：结合log step 85所打印的共同指令日程（第一段偏航指令接近零，第二段为正向持续转动），可见本次行进的转向时序与幅度不符；全程未见失去支撑或倒地，故标步态。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 109,
+    "source_zh": "日志",
+    "evidence_zh": "推断：agent在log step 109、111检查的是前一次测试的遥测，显示偏航反复正负摆动且第二段侧移和转向跟踪不足；随后在log step 112改写步态及偏航补偿控制器，log step 114启动最终回合。最终视频1.5–3.5 s仍有明显非指令转向，视频5.5–10.143 s转向不足，支持行进控制未实现所要求的速度与偏航跟踪。前次测试只作为控制问题的会话证据，不作为最终节点或评分值；具体是增益、运动学还是足地作用造成，现有证据不能细分。"
+   }
+  ],
+  "gap_zh": "无状态日志；本次reward未报告worst_lin_err、worst_yaw_err、分段误差或终止条件细项，最终回合的完整服务返回也未收录于log step 114之后，因此原生指标值留null，不能精确定位误差首次超标时刻或控制子类型。log step 109–111的遥测属于此前测试，log step 120的完成情况是agent自述，均不替代最终评分回合的观测。",
+  "unmet": {
+   "joystick_schedule_followed": false
+  }
+ },
+ "summary_zh": "评分记录显示最终回合执行1000个控制动作，视频中机器人始终由腿支撑，但前半段明显偏转、后半段持续转向不足，“按指令行进”最终为0/1。断点在该步骤操作中，现象为步态，结合会话中的偏航调试记录将根因归于控制。缺少本次分段误差及状态日志，两个原生指标值未知，无法进一步定位控制子类型。",
+ "conf": 7,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成、断点、步态现象与停在“控制”的根因均有视频与评分字段（success=0、n_actions=1000）支持；视频全程未倒地，前半段明显偏航，后半段转向不足，与标签一致。log step 85的日程与会话遥测被标为推断/会话证据，未当作评分值，未发现需改内容的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:23:48Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--vyhemke/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_barkour_joystick-standard-realtime--vyhemke/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "c81e8bdf5ebe91f7159ad1e1ec4330809b6d2c8235c2df9ebda6ba135e61b863",
+   "video_sha256": "8c8fa91a9b6a395a850d896042f6a9519f5bd087aaf2552171e2cd7e6507b3fc",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--7b5rrvr.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--7b5rrvr.json
new file mode 100644
index 0000000000000000000000000000000000000000..78573b3f796e3a3c4e93662a2ff57ef7ded5b195
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--7b5rrvr.json
@@ -0,0 +1,193 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-privileged--7b5rrvr",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 341,
+   "horizon": 500,
+   "fell": 1,
+   "fell_at_step": 26.0,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.0554,
+   "end_pose_angle_deg": 96.87,
+   "video_rendered": 1,
+   "agent_sessions": 53,
+   "agent_resets": 83,
+   "agent_sim_steps": 35445,
+   "agent_replays": 9,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-privileged--7b5rrvr/replay.mp4",
+  "duration_s": 6.706,
+  "fps": 17.0,
+  "frames": 114,
+  "sha256": "204697cfda223d28020206276859530241a0d05a49f60e5dd1d240fa376828d3"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00 s 的track与side并排画面显示Go1四足姿态，躯干近水平，未处于头朝下的竖直倒立。"
+     },
+     {
+      "t": 0.53,
+      "log_step": null,
+      "text_zh": "观察：逐帧检查视频 0.00–1.00 s，Go1短暂抬起后躯，但躯干始终明显偏离竖直；视频约 0.53 s 前腿折低、腿段落地，随后躯干下沉，未建立可持续的前脚倒立。"
+     },
+     {
+      "t": 6.65,
+      "log_step": 119,
+      "text_zh": "观察：视频 1.0–6.65 s 的检查帧中Go1一直低伏，视频 5.65–6.65 s 逐帧检查也未见恢复；log step 119 对最终交付文件的重放返回success=false、fell_at=26、pose_angle_deg=96.87、imu_height=0.0554，与评分器终态字段一致。未出现符合倒立判据的连续保持窗口。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–0.53 s，Go1脚部与地面接触，在抬后躯、前倾尝试中以前脚接地；随后前腿折低、腿段及躯干落到地面，视频 0.8–6.65 s 维持低伏接地。未形成持续的竖直倒立支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better",
+   "value": 96.87
+  },
+  {
+   "name": "imu_height",
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better",
+   "value": 0.0554
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "front_leg_handstand",
+   "t": 0.53,
+   "log_step": 119,
+   "where": "操作中",
+   "evidence_zh": "观察：视频约 0.53 s，Go1正在将四足姿态转为前腿倒立，前腿却折低、腿段落地，视频约 0.8 s 躯干已低伏，直到视频 6.65 s 未恢复。log step 119 的最终交付重放及评分器字段均记录fell_at=26且success=false；控制步编号仅作日志证据，视频锚点取自逐帧观察。"
+  },
+  "unmet": {
+   "front_leg_handstand": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 0.53,
+    "log_step": 119,
+    "evidence_zh": "观察：视频 0.24–0.53 s，Go1后躯抬起但前腿支撑折低，前腿腿段落到地面；视频 0.65–1.0 s 躯干继续下沉，之后直到视频 6.65 s 都低伏在地，未站成倒立。log step 119 的最终文件重放返回fell_at=26，评分器fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "轨迹或逆解",
+    "log_step": 111,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 46 的交付生成脚本依次使用固定起跳目标、24个控制步的线性插值和恒定保持目标；log step 111 将保持目标替换为单独验证过的静态倒立目标，但重放最终文件仍返回fell_at=26、success=false，log step 119 再次得到相同终态。log step 113 的成功保持测试先直接写入倒立qpos并清零qvel，并未验证从规定初态起立。结合视频 0.24–0.8 s 的前腿折倒，可判断交付的起立过渡轨迹未能将规定初态带入稳定倒立；单独成立的静态姿态不能补足这段运动。"
+   }
+  ],
+  "gap_zh": "本次与参考运行均无状态日志；无法逐控制步核对接触传感器、IMU角度和高度，也无法确定首次跌倒具体由哪项终止传感条件触发。视频足以判断始终未形成倒立保持；根因依据已交付脚本及其重放结果推断，未进一步归因为某个增益、摩擦或力参数。"
+ },
+ "summary_zh": "Go1尝试从四足姿态抬后躯并转为前腿倒立，但视频约 0.53 s 前腿折倒，随后低伏至片尾，“前腿倒立”始终未达成（终态0/1）。交付的起立过渡轨迹未将机器人带入稳定倒立；log step 119 的最终重放和评分器一致报告失败，终态夹角96.87度、IMU高度0.0554米，单独写入倒立状态的保持测试不属于本次评分轨迹。缺少状态日志，无法确认首次失败的具体接触传感条件。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成（0/1）、断点（front_leg_handstand，约0.53 s，对应fell_at=26×0.02 s）、平衡现象与轨迹/逆解根因均与视频和log step 119重放结果（success=false, fell_at=26, 96.87°, 0.0554 m）一致；视频约0.5 s后躯干低伏直到片尾。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:04:18Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--7b5rrvr/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--7b5rrvr/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "cbe8ec98f6cfc7057fc7d6a3d51a12fd24b532453f58babe0450191abb91903f",
+   "video_sha256": "204697cfda223d28020206276859530241a0d05a49f60e5dd1d240fa376828d3",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--7bm4aag.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--7bm4aag.json
new file mode 100644
index 0000000000000000000000000000000000000000..863275a3235009f3b7ed656d7c2d3c7828f35ee0
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--7bm4aag.json
@@ -0,0 +1,173 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-privileged--7bm4aag",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 135,
+   "horizon": 500,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.5442,
+   "end_pose_angle_deg": 10.24,
+   "video_rendered": 1,
+   "agent_sessions": 2,
+   "agent_resets": 4,
+   "agent_sim_steps": 1350,
+   "agent_replays": 2,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-privileged--7bm4aag/replay.mp4",
+  "duration_s": 2.72,
+  "fps": 25.0,
+  "frames": 68,
+  "sha256": "188eb95a52ecb3bbe903201601e19ecb88e7a3f179942990afd12272ef16ac65"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [
+     [
+      2.68,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.00 s的并排视图中，Go1躯干近水平、四脚位于地面，尚未进入前腿倒立。逐帧检查视频0.04–0.68 s可见机器人屈腿、腾起并将后躯向上转，随后躯干转为近竖直、头端朝下。"
+     },
+     {
+      "t": 0.88,
+      "log_step": null,
+      "text_zh": "观察：视频0.68–0.88 s机器人由腾起后的近竖直姿态逐渐伸出前腿，视频0.88 s前脚已到地面支撑位置、后脚高于躯干；逐帧检查视频0.92–2.68 s，机器人保持头朝下、躯干近竖直和双后脚离地，前脚在地面附近作小幅支撑调整，未见倒地。"
+     },
+     {
+      "t": 2.68,
+      "log_step": 25,
+      "text_zh": "观察：log step 25记录agent修正生成脚本后重新保存的最终交付文件为135行动作，Session重放和独立CLI重放均报告success=true、steps=135、time_s=2.7、fell_at=null、pose_angle_deg=10.24、imu_height=0.5442；inputs/run.json评分器字段同样为success=1、replay_success=1、deterministic=1、fell=0。视频2.64 s和2.68 s均显示前腿支撑倒立姿态。"
+     },
+     {
+      "t": 2.68,
+      "log_step": 25,
+      "text_zh": "推断：依本次评分规则，成功结果确认末尾连续100个控制步满足夹角、高度和后脚离地条件。节点包含完整保持窗口，故将完成锚点近似放在视频2.68 s末帧，而非首次转为近竖直的时刻；log step 25的135个控制步×0.02 s=仿真2.70 s与视频2.72 s总长及视频末尾倒立相符，仅用于核对交付片段。没有逐步状态记录，无法精确确定最早合格采样和最早完成时刻。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：Go1以脚与地面作用起立，转起过程中有短暂腾空；视频0.88–2.68 s前脚处于地面支撑位置、双后脚悬空，前腿与躯干协同保持倒立。并排视图中两前脚有重叠，不能逐步量化各脚承重。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better",
+   "value": 10.24
+  },
+  {
+   "name": "imu_height",
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better",
+   "value": 0.5442
+  }
+ ],
+ "failure": null,
+ "summary_zh": "Go1由四足姿态转起，最终完成前腿倒立节点（1/1）；视频末尾保持头朝下、躯干近竖直、后脚离地，前脚位于地面支撑位置。最终交付的135步轨迹在log step 25经重新生成和独立重放成功，评分器确认末尾连续100步合格、无历史翻倒，终态夹角10.24°、IMU高度0.5442米。无逐步状态日志，精确阈值和连续计数依据评分结果，完成时刻近似标在视频2.68 s；oracle元数据的runId与本次相同，未作为独立参考证据。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "视频显示Go1由四足起立至头朝下倒立并保持到末尾，评分器成功、无翻倒，节点final=true、时间锚点2.68 s（含100步保持窗口）合理，接触标注为腿/躯干·支撑，无失败层，未发现错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:09:19Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--7bm4aag/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--7bm4aag/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "a51668b0ba8538cab1aa9384f8926bdd61981672f386add1b7a2aaeaac8bda7a",
+   "video_sha256": "188eb95a52ecb3bbe903201601e19ecb88e7a3f179942990afd12272ef16ac65",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--efkdv8w.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--efkdv8w.json
new file mode 100644
index 0000000000000000000000000000000000000000..28b0d873d0721ec1aa0139390669077b9a062b41
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--efkdv8w.json
@@ -0,0 +1,192 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-privileged--efkdv8w",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 500,
+   "horizon": 500,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.2465,
+   "end_pose_angle_deg": 113.91,
+   "video_rendered": 1,
+   "agent_sessions": 666,
+   "agent_resets": 1335,
+   "agent_sim_steps": 35180,
+   "agent_replays": 2,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 11.554,
+   "peak_processes": 12,
+   "peak_disk_gb": 0.241,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-privileged--efkdv8w/replay.mp4",
+  "duration_s": 10.5,
+  "fps": 12.0,
+  "frames": 126,
+  "sha256": "83d7953eb59b17a73010df5d629127364bbfc986012bfdead185c54094b487b5"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0 s 的track与side并排视图中，Go1以四足姿态接地，躯干近水平；视频约 0.7–1.2 s 腿部调整后，头端抬高、后部下沉，未转成头朝下的近竖直姿态。"
+     },
+     {
+      "t": 1.0,
+      "log_step": null,
+      "text_zh": "观察：视频 1.0–10.0 s 的逐段画面及视频 9.5–10.42 s 的逐帧末段中，Go1持续保持头高尾低的倾斜姿态，后脚仍在地面，未见后腿抬至身体上方；成功参考视频约 0.6–2.68 s 中近竖直躯干、头朝下和后腿高悬的可见特征在本次均未出现。"
+     },
+     {
+      "t": 10.4167,
+      "log_step": 109,
+      "text_zh": "观察：评分器字段end_pose_angle_deg=113.91、end_imu_height=0.2465，分别不满足不超过30°和至少0.40米的要求；log step 108保存500行动行后，log step 109及113对该交付文件重放，均返回success=false、pose_angle_deg=113.91、imu_height=0.2465、fell_at=null，与视频末态一致。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–1.2 s，Go1前后脚与地面接触支承机体，前腿调整、后腿弯曲使躯干头端抬高；视频 1.2–10.42 s 仍以前后脚接地支撑，未形成后脚悬空的前脚倒立支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "value": 113.91,
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better"
+  },
+  {
+   "name": "imu_height",
+   "value": 0.2465,
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "front_leg_handstand",
+   "t": 1.0,
+   "log_step": null,
+   "where": "操作中",
+   "evidence_zh": "观察：视频约 0.7–1.2 s，Go1调整腿部时头端上扬、后部下沉，随后停留在该倾斜姿态直到视频 10.42 s；本次没有建立头朝下、后腿在上方的倒立状态。评分器记录fell=0，失败末态角度113.91°、IMU高度0.2465米。"
+  },
+  "unmet": {
+   "front_leg_handstand": false
+  },
+  "phenomena": [
+   {
+    "cat": "运动",
+    "item": "位姿不准",
+    "t": 1.0,
+    "log_step": null,
+    "evidence_zh": "观察：视频约 1.0 s 起，Go1头端高于后部、躯干仍接近水平而略倾斜，后腿折在地面附近；视频 9.5–10.42 s 仍是此姿态，没有转成头朝下、躯干近竖直、后腿高悬的目标倒立。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：agent在log step 108用俯仰路径及腿部关节控制器生成并保存500行轨迹，目标俯仰参数thd=78°，但该文件在log step 109及113的两次重放均停在pose_angle_deg=113.91、imu_height=0.2465、success=false；与视频约 1.0–10.42 s 的头高尾低停滞一致，说明交付的关节运动未实现起倒立的姿态转换。现有记录不足以将问题细分到某个增益、逆解或反馈环节。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有逐控制步的状态日志，无法精确复核IMU角度、高度、脚触地传感器和连续100步窗口；视频中的持续姿态偏差及评分器末态字段已足以判节点未达。控制失效的具体机制未被独立验证；agent在log step 121自述的会话试验过度翻转和制动不足不作为本次交付轨迹发生翻倒的证据。"
+ },
+ "summary_zh": "Go1尝试调整腿部进入前腿倒立，但在视频约 1.0 s 形成头高尾低、后脚仍接地的倾斜姿态，并持续至视频 10.42 s，前腿倒立节点始终未达（0/1）。交付控制轨迹未实现目标姿态转换；评分器记录未翻倒，末态角度113.91°、IMU高度0.2465米均不合格。缺少逐控制步状态日志，无法进一步定位控制机制或复核传感器连续窗口。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达（0/1）、断点 front_leg_handstand、位姿不准 + 控制（未细分）均有视频、评分器字段（end_pose_angle_deg=113.91、end_imu_height=0.2465）和 log step 108/109/113 支持，未发现需改动的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:10:53Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--efkdv8w/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--efkdv8w/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "ec285becb26769726f8cbab3d17b5308be4795bf93a478db55b627daf9667f4f",
+   "video_sha256": "83d7953eb59b17a73010df5d629127364bbfc986012bfdead185c54094b487b5",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--etmewbj.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--etmewbj.json
new file mode 100644
index 0000000000000000000000000000000000000000..99e51a6e29f44d897085def6173559cb22f5462d
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--etmewbj.json
@@ -0,0 +1,178 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-privileged--etmewbj",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 108,
+   "horizon": 500,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.5046,
+   "end_pose_angle_deg": 17.3,
+   "video_rendered": 1,
+   "agent_sessions": 21,
+   "agent_resets": 544,
+   "agent_sim_steps": 51130,
+   "agent_replays": 6,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.937,
+   "peak_processes": 8,
+   "peak_disk_gb": 0.252,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-privileged--etmewbj/replay.mp4",
+  "duration_s": 2.2,
+  "fps": 25.0,
+  "frames": 55,
+  "sha256": "ff2ace2b61cf756c36503e7f86540b73089f35e29a832cdafd49dd56d8192d9d"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [
+     [
+      0.18,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00–0.04 s 的两路画面中，Go1躯干近水平、四脚位于地面，尚未进入倒立；视频 0.08–0.16 s 机器人以前脚支撑向头端转起，后脚逐渐离地。"
+     },
+     {
+      "t": 0.2,
+      "log_step": null,
+      "text_zh": "观察：视频 0.20 s 起，Go1头端朝下、躯干接近竖直、后腿位于身体上方，前脚位于机体下方支撑；已逐帧检查视频 0.20–2.16 s，包括末秒，机器人保持此姿态至末帧，未见倒地或后脚落地。"
+     },
+     {
+      "t": 0.18,
+      "log_step": null,
+      "text_zh": "推断：依据视频 0.16 s 与视频 0.20 s 的起立过渡帧，将持续合格姿态窗口起点近似标为视频 0.18 s（误差约一个视频帧）；按state节点持续窗口的区间惯例回记起点，完整100个控制步的验收在视频末尾完成。评分器reward给出success=1、replay_success=1、deterministic=1、n_actions=108、fell=0，支持该窗口连续合格；无逐步状态日志，不能由画面单独精确测量阈值。"
+     },
+     {
+      "t": 2.16,
+      "log_step": null,
+      "text_zh": "观察：视频 2.16 s 末帧中，Go1仍以前脚支撑倒立，后脚悬空；inputs/run.json的评分器字段end_pose_angle_deg=17.3、end_imu_height=0.5046，分别满足不超过30度和至少0.40米的终态要求。"
+     },
+     {
+      "t": null,
+      "log_step": 75,
+      "text_zh": "观察：log step 75 的最终生成脚本重置后执行起立及保持动作，首次check返回成功便结束动作数组，并调用replay检查后保存；log step 76 调用该脚本和回放。较早log step 65所报告的115行动测试不是评分器记录的108行动交付，未据此前次测试标注本视频节点。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00 s 起立前Go1四脚支承于地面；视频约 0.20–2.16 s，机器人以前脚与地面的接触支撑头朝下的机体，双后脚抬离地面。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better",
+   "value": 17.3
+  },
+  {
+   "name": "imu_height",
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better",
+   "value": 0.5046
+  }
+ ],
+ "failure": null,
+ "summary_zh": "Go1从四足站立转为前脚支撑倒立，视频约 0.18 s 起的姿态窗口保持至视频 2.16 s 末帧，前腿倒立节点终态完成（1/1）。评分器确认连续100个控制步合格、无历史翻倒，末态前向夹角17.3度、IMU高度0.5046米。未提供逐步状态日志，姿态窗口起点为视觉近似，精确角度、高度及连续窗口验收依据评分器记录。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "前腿倒立节点终态完成、区间起点约0.18 s、contact_actual（腿/躯干 支撑）与帧一致；无失败层，符合成功运行。时钟均为视频时间，validator通过。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:02:19Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--etmewbj/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--etmewbj/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "de30815becac1744ade0cefb80032e79a9987466d1f51e319e86fc7dd10adf28",
+   "video_sha256": "ff2ace2b61cf756c36503e7f86540b73089f35e29a832cdafd49dd56d8192d9d",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--kyknckd.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--kyknckd.json
new file mode 100644
index 0000000000000000000000000000000000000000..491aeca3fe7d497cd18b7244f96a70a28fde2632
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--kyknckd.json
@@ -0,0 +1,173 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-privileged--kyknckd",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 116,
+   "horizon": 500,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.5252,
+   "end_pose_angle_deg": 1.42,
+   "video_rendered": 1,
+   "agent_sessions": 607,
+   "agent_resets": 613,
+   "agent_sim_steps": 71615,
+   "agent_replays": 5,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 18.916,
+   "peak_processes": 7,
+   "peak_disk_gb": 0.255,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-privileged--kyknckd/replay.mp4",
+  "duration_s": 2.36,
+  "fps": 25.0,
+  "frames": 59,
+  "sha256": "1eaec060d5bf2595bfbe5a76563c7b2459b83009180db123beefa5f5713e0f1a"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [
+     [
+      2.32,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.00 s中Go1四足着地、躯干近水平，尚未满足倒立；视频0.08–0.32 s中机器人后腿离地，躯干向头端下方转起。"
+     },
+     {
+      "t": 0.48,
+      "log_step": null,
+      "text_zh": "观察：视频0.36–0.48 s的track与side并排画面中，Go1头端朝下、躯干近竖直，双前脚落在地面支撑，后腿举向身体上方；视频0.60–2.32 s持续保持该姿态，已逐帧检查视频1.32–2.32 s，未见倒下或后脚重新落地。"
+     },
+     {
+      "t": 2.32,
+      "log_step": 157,
+      "text_zh": "观察：评分器reward报告success=1、replay_success=1、deterministic=1、n_actions=116、fell=0，终态前向轴夹角1.42度、IMU高度0.5252米；log step 157重新生成最终交付文件后，基准回放工具同样返回116步、仿真2.32 s、success=true，终态哈希与log step 140、141一致。"
+     },
+     {
+      "t": 2.32,
+      "log_step": 139,
+      "text_zh": "推断：log step 138的最终生成脚本在连续100个合格控制步后停止，log step 139报告保存116步，且log step 157对该交付文件的回放成功。按DAG控制周期0.02秒，116×0.02对应视频2.32 s；已在视频2.32 s末帧的竖直倒立处核对，故将包含持续窗口的节点完成时刻标为视频2.32 s，而非刚转起时。无逐步状态日志，连续窗口由成功验收与视频保持过程共同支持。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00–0.04 s中Go1四脚接触地面；转起时后脚离地，视频0.40–2.32 s中左右前脚接触地面并支撑竖直躯干，后腿保持抬起。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "value": 1.42,
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better"
+  },
+  {
+   "name": "imu_height",
+   "value": 0.5252,
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better"
+  }
+ ],
+ "failure": null,
+ "summary_zh": "Go1从四足站立转起，约在视频0.36 s形成近竖直的头朝下姿态，之后以前脚支撑并保持至视频2.32 s；包含连续100控制步保持要求的前腿倒立节点在视频2.32 s完成，最终1/1。评分器确认确定性回放成功、无翻倒，终态夹角1.42度、IMU高度0.5252米；缺少逐步状态日志，精确连续窗口依赖最终交付文件的成功验收，视频提供姿态与接触支持。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "节点front_leg_handstand达成、final=1与评分器success/fell=0一致；视频中姿态与oracle一致（头朝下、前脚支撑、后腿抬起），2.32 s按116步×0.02 s换算并在末帧核对，无失败层，未见错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:07:24Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--kyknckd/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--kyknckd/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "05f6fbf1c883227c478fbb9e63e518997c2e1519471b3c755cdf4abdfde44d2b",
+   "video_sha256": "1eaec060d5bf2595bfbe5a76563c7b2459b83009180db123beefa5f5713e0f1a",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--p6ijlye.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--p6ijlye.json
new file mode 100644
index 0000000000000000000000000000000000000000..28ca6cc0ab5a87b4b562f6ce9e3e3a7cceb04df8
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--p6ijlye.json
@@ -0,0 +1,173 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-privileged--p6ijlye",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 116,
+   "horizon": 500,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.5338,
+   "end_pose_angle_deg": 12.71,
+   "video_rendered": 1,
+   "agent_sessions": 16,
+   "agent_resets": 19,
+   "agent_sim_steps": 7840,
+   "agent_replays": 3,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-privileged--p6ijlye/replay.mp4",
+  "duration_s": 2.36,
+  "fps": 25.0,
+  "frames": 59,
+  "sha256": "94b4948be309cbb3b352a5ddc6b162f7da15afddc45d64894600c4579d9f13b3"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [
+     [
+      0.34,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.00 s的track与side并排画面中，Go1四足站立、躯干近水平，尚未倒立；视频0.12–0.32 s机器人将后腿抬起，躯干逐渐转向头朝下。"
+     },
+     {
+      "t": 0.36,
+      "log_step": null,
+      "text_zh": "观察：逐帧查看本次视频0.00–2.32 s；视频0.36 s机器人头端朝下、躯干近竖直，前脚位于地面支撑位置，后脚悬空。视频0.40–2.32 s机器人继续用前腿调整支撑，后脚未见落地，末秒及视频2.32 s末帧仍保持倒立。成功参考视频0.48 s及2.40–2.68 s呈现相同的头朝下、后腿悬空姿态，仅用于辨识目标外观。"
+     },
+     {
+      "t": 2.32,
+      "log_step": 17,
+      "text_zh": "观察：log step 17中agent重新生成并保存交付轨迹后，评分器replay命令报告steps=116、time_s=2.32、success=true、fell_at=null、pose_angle_deg=12.71、imu_height=0.5338；inputs/run.json的reward同时给出n_actions=116、success=1、replay_success=1、deterministic=1、fell=0及相同终态角度和高度。log step 16的213步重放属于另一会话测试，不用作本次节点进度。"
+     },
+     {
+      "t": 0.34,
+      "log_step": 17,
+      "text_zh": "推断：DAG要求末尾100个连续合格控制步，交付116步且成功，故至少第17–116个控制步满足姿态条件；按每控制步0.02秒，第17步为仿真0.34 s。log step 17的116步对应仿真2.32 s，与视频2.32 s末帧核对一致，视频0.32–0.36 s也正呈现转入近竖直倒立，因此以视频约0.34 s作为合格保持区间起点（位于相邻帧之间，非传感器实测边界）。该区间在末尾经连续窗口验收后回溯标出；无逐步状态日志，不能独立核验每个0.02秒采样。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00 s Go1四脚接地；转起后，视频0.16–2.32 s前脚在地面承担支撑、前腿屈伸调整机体姿态，双后脚抬离地面。双前脚并非每帧均能分辨同时接地，未据此声称持续双脚承重或给出接触力。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better",
+   "value": 12.71
+  },
+  {
+   "name": "imu_height",
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better",
+   "value": 0.5338
+  }
+ ],
+ "failure": null,
+ "summary_zh": "Go1从四足站立转为以前脚支撑、头朝下且后腿悬空的倒立，视频约0.34 s起进入最终合格保持区间，并持续至视频2.32 s末帧，唯一节点最终完成。评分器确认116步交付成功且无历史翻倒，末态前向夹角12.71°、IMU高度0.5338米，均满足阈值；本次无失败断点。缺少逐步状态日志，连续100步的精确角度、高度与接触判据由成功评分确认，视频用于核对姿态和接触外观，区间起点为近似推断。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "唯一节点front_leg_handstand最终完成、无失败断点，与评分器116步成功、视频末帧倒立一致；0.34 s起点已明确标为由100步窗口推断。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:02:57Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--p6ijlye/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--p6ijlye/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "dae45ebd330974e46e8e037f4ccef2d6496e6f4c405b36e633abfd0335eaaf3f",
+   "video_sha256": "94b4948be309cbb3b352a5ddc6b162f7da15afddc45d64894600c4579d9f13b3",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--trejvr7.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--trejvr7.json
new file mode 100644
index 0000000000000000000000000000000000000000..b702b9fee471c710f70f3d4356252281aadb9226
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-privileged--trejvr7.json
@@ -0,0 +1,187 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-privileged--trejvr7",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "n_actions": 64,
+   "horizon": 500,
+   "fell": 1,
+   "fell_at_step": 64.0,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.4344,
+   "end_pose_angle_deg": 22.51,
+   "grader_error": 0,
+   "video_rendered": 1,
+   "agent_sessions": 3066,
+   "agent_resets": 3080,
+   "agent_sim_steps": 483135,
+   "agent_replays": 4,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 1,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 9.883,
+   "peak_processes": 10,
+   "peak_disk_gb": 0.831,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 7,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0,
+   "over_oom": 1
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-privileged--trejvr7/replay.mp4",
+  "duration_s": 1.32,
+  "fps": 25.0,
+  "frames": 33,
+  "sha256": "8c659eb60755e5752e19822fb946266b1e6caee342b70fa5ab643f81ab26f54c"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.00 s的track与side并排画面中，Go1为四足站姿；视频0.08–0.52 s机器人蹬地腾起并将头端转向下方，视频0.60–1.04 s可见前脚落地支撑、后腿悬空，躯干接近竖直。"
+     },
+     {
+      "t": 1.28,
+      "log_step": 71,
+      "text_zh": "推断：逐帧检查视频0.00–1.28 s，整段运动也不足节点要求的100个控制步（首末跨度1.98 s），不能把短暂倒立记为完成。log step 71对已保存交付文件的重放报告steps=64、time_s=1.28、success=false、fell_at=64；评分器同样记录n_actions=64、fell=1、deterministic=1。视频1.12–1.28 s躯干继续转动、头端下降，未恢复稳定保持。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00–0.08 s Go1脚部接触地面蹬起；腾空后，视频约0.60–1.28 s左右前脚在地面支撑机体，后脚悬空；左右前脚承重分配及前腿非脚部是否触地无传感记录，不能精确区分。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "value": 22.51,
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better"
+  },
+  {
+   "name": "imu_height",
+   "value": 0.4344,
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "front_leg_handstand",
+   "t": 1.28,
+   "log_step": 71,
+   "where": "操作中",
+   "evidence_zh": "观察：视频1.08–1.28 s Go1在前脚撑地、后脚悬空时躯干继续向另一侧倾转，头端逐渐下降，末帧尚未完全躺倒；log step 71交付文件重放在控制步64报告fell_at=64、success=false，评分器记录fell=1。推断：该控制步按任务给定0.02 s步长对应视频1.28 s，已与末段姿态变化及末帧核对；未完成保持时已触发历史翻倒判据，断点属于前腿倒立操作中。"
+  },
+  "unmet": {
+   "front_leg_handstand": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.28,
+    "log_step": 71,
+    "evidence_zh": "观察：视频1.04 s后Go1的后腿再次摆动，视频1.12–1.28 s躯干越过近竖直姿态继续倾转、头端降低，前脚支撑未使姿态稳定；log step 71对交付文件的重放报告fell_at=64，评分器记录fell=1。此处为倒立过程失衡，画面结束前未显示完全倒地。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志 / 视频 / 评分器字段",
+    "evidence_zh": "推断：log step 70列出的交付动作和log step 71重建脚本一致：机器人先执行20个控制步的蹬起目标，再固定前腿目标、分段改变后腿目标，最后12个控制步重复同一组目标。视频约0.60–1.04 s前脚支撑时躯干近竖直，随后在视频1.12–1.28 s继续倾转；log step 71重放与评分器均确认控制步64翻倒。该开环关节目标序列未能稳定转起后的机体姿态，支持控制层归因；不足以证明特定增益、接触几何或逆解问题。"
+   }
+  ],
+  "gap_zh": "无逐步状态日志，无法确定控制步64的fell由躯干上轴低于阈值还是前腿非脚部触地触发，也不能精确恢复角度、足部接触和高度的逐步变化；因此不细分控制子类。视频仍足以确认实际尝试和短暂前脚支撑，64步交付长度及重放记录足以否定100步连续保持。"
+ },
+ "summary_zh": "Go1完成蹬地转起，并在视频约0.60–1.04 s以前脚支撑、后腿悬空呈近竖直姿态，但未达到前腿倒立的100步连续保持，节点最终0/1。视频1.12–1.28 s机体继续倾转，交付文件重放在控制步64记录翻倒；开环关节目标序列未能稳定姿态，根因标为控制。缺少逐步状态与接触传感记录，无法确定翻倒的具体触发条件或控制子类。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成、断点（前腿倒立，操作中）、平衡/控制归因与证据均与视频和评分器字段一致，未发现需要修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:09:39Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--trejvr7/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-privileged--trejvr7/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "f53d65d008cb4379b4e018c163d5133378b607645ada4734d1505550ebbd6b3b",
+   "video_sha256": "8c659eb60755e5752e19822fb946266b1e6caee342b70fa5ab643f81ab26f54c",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--emkqvnw.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--emkqvnw.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9bbb376c81152ccae4781c343392648378c8433
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--emkqvnw.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-standard--emkqvnw",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 31,
+   "sim_steps": 31,
+   "physics_steps": 155,
+   "steps": 54,
+   "elapsed_s": 846.0,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.28,
+   "end_pose_angle_deg": 15.97,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-standard--emkqvnw/replay.mp4",
+  "duration_s": 0.64,
+  "fps": 25.0,
+  "frames": 16,
+  "sha256": "81c93d5bdd3b363a55550af6866eaee3d61fc49360a9551784216b33a8fe5c64"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 74,
+      "text_zh": "观察：log step 74重置到episode 11、resets_left=0；任务文本说明只评分最后一局。视频0.00–0.60 s全部16帧为同一段从四足站立向头朝下俯转的运动，没有重置；评分器记录episodes=11、n_actions=31、live_matches_replay=1。"
+     },
+     {
+      "t": 0.6,
+      "log_step": 81,
+      "text_zh": "观察：视频0.36–0.60 s机器人后躯升起、头端持续降低，前腿仍折在机体下方，末帧头端和前腿关节已贴近地面；没有形成参考视频0.60–2.60 s所示的机体抬高、前腿伸出支撑的倒立。评分器报告end_pose_angle_deg=15.97、end_imu_height=0.28 m、fell=1；角度虽进入30度阈值，高度未达0.40 m。"
+     },
+     {
+      "t": 0.6,
+      "log_step": 81,
+      "text_zh": "推断：整段视频仅0.64 s，且初态为四足站立，不能完成连续100个控制步的保持窗口；结合末态高度不足与服务在log step 81返回done=True、reason=fell，前腿倒立节点从未达到。log step 83–85才留下的未测试策略不属于这段评分运动。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00–0.32 s，Go1的前后脚与地面接触支承，机器人屈伸后腿并向头端俯转；视频0.40–0.60 s后脚抬离地面、前脚在低位接触地面，前腿保持折叠，未建立能抬高机体的稳定支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "value": 15.97,
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better"
+  },
+  {
+   "name": "imu_height",
+   "value": 0.28,
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "front_leg_handstand",
+   "t": 0.56,
+   "log_step": 81,
+   "where": "操作中",
+   "evidence_zh": "观察：视频0.48–0.60 s机器人在前腿仍折叠的情况下继续向头端俯转，头端降至地面附近，未能以前脚将机体撑高；log step 81服务在control_steps=31返回done=True、reason=fell，此后没有评分运动。推断：断点是正在执行的前腿倒立。control period=0.02 s，31步约为仿真0.62 s，与视频0.60 s末帧和0.64 s片长相符，时间锚点取自所见帧。"
+  },
+  "unmet": {
+   "front_leg_handstand": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 0.56,
+    "log_step": 81,
+    "evidence_zh": "观察：视频0.48–0.60 s，Go1头端快速下沉，前腿折在低位，机体未被前脚稳定托起，末帧头端与前腿关节贴近地面；同局服务在log step 81返回fell，评分器fell=1、末态IMU高度仅0.28 m。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：前腿展开与机体俯转的控制未协调好，未能在头端降低前建立足够高的支撑。依据是log step 76–77的脚本连续提交前大腿1.3 rad、前膝-1.8 rad目标，log step 79改为前大腿-0.3 rad、前膝-2.7 rad时服务报告俯仰角速度8.23 rad/s；log step 81再将前膝目标改为-1.8 rad，但终止时实际两前膝仍为-2.81与-2.45 rad、俯仰角速度7.94 rad/s。视频0.40–0.60 s也显示机体继续俯转而前腿未撑开；这些证据支持控制层归因，但不足以细分接触设计、轨迹或反馈问题。",
+    "source_zh": "日志 / 视频"
+   }
+  ],
+  "gap_zh": "没有状态日志或逐控制步的前腿非脚部接触传感器记录；末帧前腿与头端局部重叠，无法精确确定首先触地的腿段，也不能进一步区分控制根因的子类。视频已覆盖全段，节点未完成和失衡终止有画面、服务结果与评分器字段共同支持。"
+ },
+ "summary_zh": "评分采用第11局：Go1从四足站立转为头朝下，后腿抬起，但前腿未撑高机体，前腿倒立节点未达到（0/1）。视频约0.56 s失衡下沉，服务随后以fell终止；末态角度15.97度符合角度阈值，但IMU高度仅0.28 m，且未形成连续保持窗口，根因可归于控制协调失败。缺少逐步状态与腿段接触记录，无法确定首先触地的部位或细分控制原因。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "标签与视频、日志一致：评分局为第11局，视频0.64 s全程俯转，末帧头端贴地、机体近竖直但高度仅0.28 m，服务返回fell；节点未达到、断点在front_leg_handstand、平衡现象与控制根因均有证据支持，未发现需改内容的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:20:42Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--emkqvnw/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--emkqvnw/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "14bae1e4099a260cc775c05477e164c0b596721644a29d23d79599f06f6eed2e",
+   "video_sha256": "81c93d5bdd3b363a55550af6866eaee3d61fc49360a9551784216b33a8fe5c64",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--fpg6bcp.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--fpg6bcp.json
new file mode 100644
index 0000000000000000000000000000000000000000..484bd43603e26b21228e2121f14bbb7e9bdca5d1
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--fpg6bcp.json
@@ -0,0 +1,206 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-standard--fpg6bcp",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 76,
+   "sim_steps": 76,
+   "physics_steps": 380,
+   "steps": 441,
+   "elapsed_s": 1766.5,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.1678,
+   "end_pose_angle_deg": 49.14,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 6,
+   "reset_budget": 10,
+   "episodes": 7,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.434,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.273,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-standard--fpg6bcp/replay.mp4",
+  "duration_s": 1.56,
+  "fps": 25.0,
+  "frames": 39,
+  "sha256": "2e5a6c3ca32f839671685a0f9158520399bdb957036efc49d005ce5f0edaec8e"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00–0.64 s，Go1从四足站立开始，躯干接近水平并逐渐前倾，未处于头朝下、躯干近竖直的倒立姿态。"
+     },
+     {
+      "t": 1.12,
+      "log_step": null,
+      "text_zh": "观察：逐帧检查视频 0.68–1.52 s，Go1反复抬脚和俯仰，视频 1.12–1.20 s虽后腿收起、后端抬高，但躯干仍明显倾斜；视频 1.28–1.52 s前腿折低、机体下坠，始终未出现持续合格倒立。成功参考视频 0.80–2.60 s中近竖直的高位姿态仅用于对照，非本次进度证据。"
+     },
+     {
+      "t": 1.52,
+      "log_step": 34,
+      "text_zh": "观察：log step 34在最后一次重置后执行第7个episode，服务返回control_steps=76、done=True、reason=fell；评分器同样记录n_actions=76、fell=1、live_matches_replay=1，终态IMU高度0.1678 m、前向轴与竖直向下夹角49.14°，分别未达0.40 m和30°要求；与视频末帧低位折倒相符。log step 35只修改脚本，未再执行机器人动作。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–0.64 s，Go1四脚接触地面支撑；视频 0.68–1.20 s转起过程中前后脚交替抬起和触地，前脚曾短暂承重，后脚未能持续悬空；视频 1.28–1.52 s腿部折低，未形成稳定的前腿倒立支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better",
+   "value": 49.14
+  },
+  {
+   "name": "imu_height",
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better",
+   "value": 0.1678
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "front_leg_handstand",
+   "t": 1.28,
+   "log_step": 34,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 1.12–1.20 s，Go1尝试收后腿并以前脚支撑抬高后端；视频 1.28 s开始明显回落，视频 1.36–1.52 s前腿折低、头端降到地面附近，未重新建立倒立。log step 34的最后episode随后以fell结束，评分器fell=1。"
+  },
+  "unmet": {
+   "front_leg_handstand": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.28,
+    "log_step": 34,
+    "evidence_zh": "观察：视频 0.68–1.20 s，Go1支撑脚不断变化、躯干前后俯仰；视频 1.28–1.52 s支撑高度塌落、前腿折叠，机体向地面下坠，未稳住。log step 34服务终止原因为fell。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "t": 1.28,
+    "log_step": 34,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 34的受评控制脚本用IMU俯仰和角速度反馈，每5个控制步更新前脚位置，并依据后脚触地读数切换后腿目标；同一日志显示角速度w由1.04变为-3.54、再至-6.54 rad/s，前脚位置修正dx达到-0.120 m限幅，随后fell。结合视频 0.80–1.52 s的俯仰反复与支撑塌落，表明这套转起反馈未能稳定姿态和角速度；不能据此确定某个增益或逆解公式是唯一原因。"
+   }
+  ],
+  "gap_zh": "无状态日志，无法逐控制步核验IMU角度、高度与全部接触，也无法区分最终fell具体由前腿非脚部触地还是躯干过度翻倒触发。控制根因来自log step 34的脚本、稀疏遥测与视频联合推断，缺少逐步目标与实际关节轨迹，不能进一步确定失稳参数或模型误差。"
+ },
+ "summary_zh": "受评轨迹是log step 34的最后一次episode；Go1启动了转起尝试，但反复俯仰、换脚支撑，前腿倒立节点始终未完成，终态为0/1。视频 1.28–1.52 s显示前腿折低、机体下坠，服务以fell结束；脚本与遥测支持转起反馈未能稳住姿态和角速度。无状态日志，具体触发终止的接触或姿态边界及失稳参数无法确定。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "前腿倒立节点未达成、终态0/1、断点与平衡/反馈控制标签与视频帧和评分器字段（fell=1、高度0.1678、角度49.14°）一致，未发现需修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:05:48Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--fpg6bcp/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--fpg6bcp/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "95dd6b61c839aabcaf46d085a94f737047ef6874449ff1a1c806a5609ff07f67",
+   "video_sha256": "2e5a6c3ca32f839671685a0f9158520399bdb957036efc49d005ce5f0edaec8e",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--hwfgywj.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--hwfgywj.json
new file mode 100644
index 0000000000000000000000000000000000000000..9706979f3ec728fb714146b6a7d216a381235c6c
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--hwfgywj.json
@@ -0,0 +1,190 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-standard--hwfgywj",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "graded_by_live": 1,
+   "replays": 0,
+   "finalized": 1,
+   "missing_log": 0,
+   "sim_error": 0,
+   "grader_error": 0,
+   "infra_error": 0,
+   "n_actions": 25,
+   "sim_steps": 25,
+   "physics_steps": 125,
+   "steps": 25,
+   "elapsed_s": 1834.5,
+   "video_rendered": 1,
+   "fell": 0,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.255,
+   "end_pose_angle_deg": 91.88,
+   "instance_mismatch": 0,
+   "no_verdict": 0,
+   "gave_up": 0,
+   "sim_crashed_on_call": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 0,
+   "reset_budget": 10,
+   "episodes": 1,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.246,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.239,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-standard--hwfgywj/replay.mp4",
+  "duration_s": 0.52,
+  "fps": 25.0,
+  "frames": 13,
+  "sha256": "3568125d9f16ec6ba8ce7dcbeb20ecafaac6eb224c205547fad5db86080f3637"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：已逐帧查看本次视频0.00–0.48 s的全部13帧（间隔0.04 s）；track与side两视图中Go1躯干始终近水平，前后脚处于地面支撑位置，仅有轻微起伏，没有头朝下、躯干近竖直和后腿抬至身体上方的倒立动作。"
+     },
+     {
+      "t": 0.48,
+      "log_step": 6,
+      "text_zh": "观察：log step 6中agent脚本只向实时机器人服务提交25行相同的初始站立关节目标，dt=0.02 s，用于延迟测试；服务返回executed=25、control_steps=25，末尾四脚接触读数均为1。评分器记录单一episode、无reset/resume、总计25个控制步，终态前向夹角91.88度、IMU高度0.255 m；视频0.48 s末帧仍为四足站姿。"
+     },
+     {
+      "t": null,
+      "log_step": null,
+      "text_zh": "观察：成功参考视频0.40–2.60 s中可见Go1头朝下、躯干近竖直、后腿处于上方的明确倒立外观；本次视频所有帧均没有这一外观。推断：本次前腿倒立节点从未成立，并非仅由success=0推定。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "value": 91.88,
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better"
+  },
+  {
+   "name": "imu_height",
+   "value": 0.255,
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "front_leg_handstand",
+   "t": 0.0,
+   "log_step": 6,
+   "where": "首步前",
+   "evidence_zh": "观察：视频0.00–0.48 s中Go1一直保持近水平四足站姿；log step 6唯一执行的控制是重复初始站姿25步的延迟测试，log step 7只查询正向/逆向运动学，未提交倒立动作。推断：评分轨迹停留在初始站姿保持，前腿倒立步骤未启动；视频0.00 s仅作首步前的导航锚点。"
+  },
+  "unmet": {
+   "front_leg_handstand": false
+  },
+  "phenomena": [
+   {
+    "cat": "没尝试",
+    "item": "没尝试",
+    "t": 0.0,
+    "log_step": 6,
+    "evidence_zh": "观察：视频0.00–0.48 s全部13帧均未见Go1转起或抬后腿进入倒立；log step 6提交的25行目标均为qhome初始站姿，评分器n_actions=25且episodes=1、resets=0，与这段站姿保持相符。推断：有评分轨迹，但其中没有开始倒立步骤。"
+   }
+  ],
+  "root_causes": [],
+  "gap_zh": "观察：inputs内没有状态日志；log step 6和log step 7的候选倒立脚位逆解查询返回ok=false，但均未执行相应动作。log step 8与log step 10没有调用、输出或说明，完整日志也无最终自述。推断：这些查询不足以区分候选目标或种子不当、求解限制与后续未执行的具体原因，故根因留空；不能由未尝试或预算耗尽反推规划或控制根因。倒立未启动，未将初始四足站立的背景支撑接触记作该步骤的contact_actual。"
+ },
+ "summary_zh": "评分轨迹仅包含初始四足站姿保持，前腿倒立从未达到，最终完成0/1节点；视频0.00–0.48 s全部帧均无转起或后腿上举，评分器终态夹角91.88度、IMU高度0.255 m也未达标。断点为前腿倒立的首步前，现象为没尝试；日志中的逆解查询失败未形成实际倒立动作，缺少足够证据确定未执行的根因。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标注与证据一致：视频全部帧为四足站姿，log step 6 仅提交25步初始站姿延迟测试，评分器25步、91.88度/0.255 m；节点未达成、断点首步前、没尝试、根因留空并有gap_zh，均成立。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:21:32Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--hwfgywj/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--hwfgywj/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "619e33df09ae19e5c906f6e668679a62cdebde61820206dcd61f804646811dfc",
+   "video_sha256": "3568125d9f16ec6ba8ce7dcbeb20ecafaac6eb224c205547fad5db86080f3637",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--mexor4y.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--mexor4y.json
new file mode 100644
index 0000000000000000000000000000000000000000..08fca6283e2259a7a4a97e18370158390b2f58fb
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--mexor4y.json
@@ -0,0 +1,190 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-standard--mexor4y",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "live_success": 1,
+   "replay_success": 1,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 129,
+   "sim_steps": 129,
+   "physics_steps": 645,
+   "steps": 129,
+   "elapsed_s": 1823.9,
+   "video_rendered": 1,
+   "fell": 0,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.5525,
+   "end_pose_angle_deg": 9.15,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 0,
+   "reset_budget": 10,
+   "episodes": 1,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.85,
+   "peak_processes": 60,
+   "peak_disk_gb": 0.384,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-standard--mexor4y/replay.mp4",
+  "duration_s": 2.6,
+  "fps": 25.0,
+  "frames": 65,
+  "sha256": "17cbf22a45075d67f7a088787f881499a2398569c0578d9838feaeaeba2a5e11"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [
+     [
+      2.56,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.00 s的track与side并排画面中，Go1躯干近水平、四脚在地面支撑，尚未形成前腿倒立。"
+     },
+     {
+      "t": 0.6,
+      "log_step": null,
+      "text_zh": "观察：逐帧检查视频0.00–0.96 s，Go1先屈伸腿部，随后头端向下、后躯抬起；视频约0.60 s躯干已近竖直，前脚处于下方支撑位置，双后脚位于身体上方并离地。此时只是开始维持姿态，尚不能计为完成连续保持窗口。"
+     },
+     {
+      "t": 2.56,
+      "log_step": null,
+      "text_zh": "观察：逐帧检查视频1.00–2.56 s，尤其视频1.60–2.56 s末段，Go1保持头端向下、躯干近竖直，左右前脚接地支撑，后脚悬空；视频2.56 s末帧仍为该姿态。成功参考视频末段具有相同的头朝下、后脚离地特征，仅用来识别姿态，不作为本次完成证据。"
+     },
+     {
+      "t": 2.56,
+      "log_step": 79,
+      "text_zh": "观察：log step 79中agent脚本向机器人实时服务提交joint_position动作，dt=0.02；该步未记录返回结果。inputs/run.json评分字段另记录n_actions=129、sim_steps=129、live_success=1、replay_success=1、deterministic=1、live_matches_replay=1、fell=0，终态end_pose_angle_deg=9.15、end_imu_height=0.5525。"
+     },
+     {
+      "t": 2.56,
+      "log_step": null,
+      "text_zh": "推断：本次评分器的成功结果确认末尾100个控制步满足倒立验收，且视频末段姿态与此一致，因此final为true；节点含1.98 s保持窗口，不在初次转起时计为完成，而把窗口完成时刻近似锚定在视频末帧2.56 s。无状态日志，无法从25 fps画面精确重建50 Hz控制步的首次合格时刻或逐步传感器值，时间存在约一帧至数帧的不确定性。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00 s起Go1以脚接触地面；转起后左右前脚接地支撑机体，前腿屈伸调整躯干姿态，双后脚抬起并在倒立保持阶段悬空，末帧仍由前脚支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "value": 9.15,
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better"
+  },
+  {
+   "name": "imu_height",
+   "value": 0.5525,
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better"
+  }
+ ],
+ "failure": null,
+ "summary_zh": "观察：Go1从四足站立转起，视频约0.60 s形成前脚支撑、头朝下且后脚悬空的倒立姿态，并保持至视频2.56 s末帧。推断：结合评分器live_success=1、replay_success=1及终态角度9.15°、IMU高度0.5525米，唯一节点“前腿倒立”在末尾连续保持窗口完成后达成，终态为1/1，无失败断点。观察：本次没有状态日志，无法独立逐控制步核验100步窗口及精确接触阈值，窗口完成时间近似标在视频末端；log step 78的本地测试后续翻倒不属于本次评分视频。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "视频帧显示Go1由四足站立转起，0.6 s起头朝下、躯干近竖直、后脚离地并保持至2.56 s末帧；评分器live_success=1、fell=0，节点final=true、无失败断点与标注一致，时钟均为视频时间。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:11:56Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--mexor4y/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--mexor4y/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "64ea4a9224396eb4dc5f426c63bda80e1c9a7bf00473d398ee90de3916fe3914",
+   "video_sha256": "17cbf22a45075d67f7a088787f881499a2398569c0578d9838feaeaeba2a5e11",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--pjmrq6k.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--pjmrq6k.json
new file mode 100644
index 0000000000000000000000000000000000000000..7865b8e3190c1f7dadd99fb393e3ff74f6fb6729
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--pjmrq6k.json
@@ -0,0 +1,209 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-standard--pjmrq6k",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 188,
+   "sim_steps": 188,
+   "physics_steps": 940,
+   "steps": 891,
+   "elapsed_s": 1317.7,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.2298,
+   "end_pose_angle_deg": 22.17,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.57,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.833,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-standard--pjmrq6k/replay.mp4",
+  "duration_s": 3.8,
+  "fps": 25.0,
+  "frames": 95,
+  "sha256": "f66f7e0cf552ef4e2e03cfe8c22a8c973528cf3d4e7eafad5d0bffe6090e835e"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00 s 的两个视角均显示Go1四足站立，躯干近水平，初始并未倒立。"
+     },
+     {
+      "t": 3.2,
+      "log_step": 79,
+      "text_zh": "观察：视频 0.20–3.40 s 中Go1逐渐调整腿部、前移机体，躯干一直近水平，视频 3.20 s 头端还略高于后端；没有出现躯干近竖直、后腿位于身体上方的持续倒立。"
+     },
+     {
+      "t": 3.76,
+      "log_step": 83,
+      "text_zh": "观察：逐帧检查视频 2.60–3.76 s；视频 3.56 s 起Go1头端快速下沉、后腿离地，视频 3.68–3.76 s 前腿折低，头端与前腿段接近或落到地面，没有抬高身体并稳定保持倒立。评分器reward报告end_pose_angle_deg=22.17、end_imu_height=0.2298 m、fell=1；末态角度虽在30度内，高度低于0.40 m且已判倒地，不能完成该节点。"
+     },
+     {
+      "t": 3.6,
+      "log_step": 85,
+      "text_zh": "推断：任务文本规定只评最后一局；log step 71–83记录最后一次重置后的动作，log step 85的机器人服务状态确认episode=11、control_steps=188、done=true、reason=fell。log step 7所读服务规格为50 Hz，故log step 81的control step 180对应视频 3.60 s，已与画面头端下倾核对；log step 83的control step 188对应视频 3.76 s，吻合最后一帧。前面重置掉的试验不计入节点进度。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–3.40 s，Go1前后脚与地面支撑并调整机体；视频 3.56–3.76 s 后脚抬离地面、前脚继续触地，但前腿折低，末帧前腿段及头端贴近地面。实际尝试包含足地支撑，未形成成功的前腿倒立支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "value": 22.17,
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better"
+  },
+  {
+   "name": "imu_height",
+   "value": 0.2298,
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "front_leg_handstand",
+   "t": 3.6,
+   "log_step": 81,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 3.56–3.76 s，Go1在抬后腿、向头端转起时前部快速下沉，前腿折低，未能建立高位倒立平衡；log step 81–83记录补救关节指令后仍快速俯转，最终机器人服务以fell终止。失败发生在前腿倒立的执行中，之后无恢复。"
+  },
+  "unmet": {
+   "front_leg_handstand": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 3.6,
+    "log_step": 83,
+    "evidence_zh": "观察：视频 3.56 s 起Go1后腿抬起而头端下坠，视频 3.68–3.76 s 前腿折低至地面，未能用伸展的前腿支撑机体；log step 83的服务反馈为done=true、reason=fell，评分器fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：agent的控制没有使前腿转起与机体俯仰稳定下来。log step 77的脚本按IMU和关节反馈生成前腿位置指令；log step 78–79显示运行该控制时俯仰从接近水平偏至-13.7度，随后log step 80发出20步恢复轨迹，log step 81在control step 180读到俯仰27.7度、俯仰角速度4.07 rad/s。log step 82–83继续两次补救指令后角速度仍增至4.79、5.61 rad/s并判fell，与视频 3.60–3.76 s 前部下沉、前腿折低一致。证据支持运动控制失稳，尚不足以确定某个增益、模型参数或接触设计是具体根因。",
+    "source_zh": "日志 / 视频"
+   }
+  ],
+  "gap_zh": "没有逐步状态日志或前腿各段触地传感器记录，无法独立定位最先触地的腿段及精确触发控制步；agent在log step 84自述前腿段触地，视频末段支持前腿折低，但不据此细分具体传感器。日志中的ref,F是agent脚本计算的参考量，不是实测接触力；缺少实测力及控制器验证，根因仅标控制，不进一步细分。"
+ },
+ "summary_zh": "评分的第11局完成0/1节点：Go1先以足地支撑调整机体，随后抬后腿、向头端转起，但视频 3.60–3.76 s 前部下沉、前腿折低，未连续保持倒立并被服务判fell。末态IMU高度0.2298 m低于0.40 m；补救指令未抑制俯仰转动，根因归为控制，缺少逐步触地和力数据以确定具体失稳机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "节点未达成、断点在front_leg_handstand、平衡+控制的标注与视频和log一致：视频 3.48–3.76 s 机体快速前俯、头端/前腿触地，log step 81–83 俯仰27.7→46.1→67.8度并判fell；时钟换算（188步×0.02=3.76 s）与末帧吻合。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:05:47Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--pjmrq6k/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--pjmrq6k/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "44332ce1047d6d9687410599a17b8ea8997f37aa2ca136499157d9779b6a9f6d",
+   "video_sha256": "f66f7e0cf552ef4e2e03cfe8c22a8c973528cf3d4e7eafad5d0bffe6090e835e",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--vxa5wag.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--vxa5wag.json
new file mode 100644
index 0000000000000000000000000000000000000000..ce5519aa25f0cf9eb2063f2525ea20f1dfd67ddd
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--vxa5wag.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-standard--vxa5wag",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 135,
+   "sim_steps": 135,
+   "physics_steps": 675,
+   "steps": 838,
+   "elapsed_s": 1338.4,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.279,
+   "end_pose_angle_deg": 68.2,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-standard--vxa5wag/replay.mp4",
+  "duration_s": 2.72,
+  "fps": 25.0,
+  "frames": 68,
+  "sha256": "faf6c58cd3c713df53a026e11a78cd5da1d5702b51ddcdc058108da9bd625ab4"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00–0.72 s，Go1四脚位于地面，躯干近水平并逐渐下蹲，尚未倒立。"
+     },
+     {
+      "t": 0.88,
+      "log_step": null,
+      "text_zh": "观察：已逐帧检查视频 0.00–2.68 s；视频约 0.80–0.88 s，Go1后躯和双后腿抬高、头端向下，之后以前脚落地与短暂腾空交替支撑；视频约 2.32–2.68 s，躯干逐渐向后倾回并下降，末帧前腿折低，未保持倒立。"
+     },
+     {
+      "t": 2.4,
+      "log_step": 69,
+      "text_zh": "推断：视频中近竖直倒立仅约从视频 0.88 s持续到视频 2.32 s，短于100个控制步所需的1.98秒采样跨度，故不记为完成区间。log step 69的服务状态明确为最后一次episode=11、control_steps=135、reason=fell，与评分器n_actions=135、fell=1及末段下落一致；评分器终态前向夹角68.2度、高度0.279米也均不达标。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–0.72 s，Go1四脚接地支撑并下蹲；视频约 0.76–0.88 s，后脚蹬地后离地，随后左右前脚反复落地支撑机体，双后脚悬空；视频约 2.40–2.68 s，前腿弯折、机体下降。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "value": 68.2,
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better"
+  },
+  {
+   "name": "imu_height",
+   "value": 0.279,
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "front_leg_handstand",
+   "t": 2.4,
+   "log_step": 69,
+   "where": "操作中",
+   "evidence_zh": "观察：视频约 2.40–2.68 s，正在以前脚支撑倒立的Go1后躯向后落下，躯干由近竖直转为明显倾斜，前腿折低；此后没有恢复。log step 69的服务记录最后一次episode=11在control_steps=135以reason=fell结束；评分器也报告fell=1、live_matches_replay=1。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 2.4,
+    "log_step": 69,
+    "evidence_zh": "观察：视频约 0.88–2.32 s，Go1以前脚落地和腾空交替支撑，后脚在空中；视频约 2.40–2.68 s，后躯持续向后落下、前腿弯折，倒立支撑失稳，未恢复。log step 69的服务终止原因为fell。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "evidence_zh": "推断：log step 42中agent编写的控制脚本在下蹲和蹬腿后，仅以IMU俯仰角及俯仰角速度的线性反馈修正前腿等关节目标；log step 57给该脚本加入初始站立段，log step 69以俯仰增益0.15、角速度增益0.12运行最后一次episode=11。视频约 0.88–2.32 s显示机器人已抬起后躯并能短时以前脚支撑，视频约 2.40–2.68 s却持续向后倾倒，说明这套反馈未能稳定维持倒立；这里不把某个具体增益或模型误差认定为已证实原因。",
+    "source_zh": "日志 / 视频"
+   }
+  ],
+  "gap_zh": "没有逐控制步状态日志，无法精确核验连续合格计数、IMU角度和高度的越界时刻，也无法在躯干过度翻倒与前腿非脚部触地两种终止条件间确定实际触发项。近竖直阶段的起止时间由画面近似判断；控制根因可定位到反馈失稳，具体动力学或增益缺陷未被独立验证。",
+  "unmet": {
+   "front_leg_handstand": false
+  }
+ },
+ "summary_zh": "评分采用最后一次episode=11：Go1下蹲蹬地后抬起双后腿，并短时以前脚支撑，但没有完成连续100个控制步的倒立，节点最终为0/1。视频约 2.40–2.68 s，机器人在倒立操作中向后倾倒、前腿折低，现象为平衡失稳，日志中的俯仰反馈控制未能维持姿态。缺少逐步状态数据，无法精确确认合格窗口边界及具体终止传感器。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "节点未达成、断点 front_leg_handstand（操作中）、现象平衡、根因反馈控制均与视频和日志一致；近竖直阶段约 0.9–2.3 s 短于 1.98 s，末段 2.5 s 后塌倒，final=0 正确。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:05:07Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--vxa5wag/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--vxa5wag/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "9de7a5d7f92401b22bf21d3f588bd778fce6e254b33aaf42e05e9f9aab97c1bf",
+   "video_sha256": "faf6c58cd3c713df53a026e11a78cd5da1d5702b51ddcdc058108da9bd625ab4",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--wydcxp2.json b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--wydcxp2.json
new file mode 100644
index 0000000000000000000000000000000000000000..2955a3c4a4ac86f35b0b2d91d0d7831ddbd39c7f
--- /dev/null
+++ b/phase2/mujoco-playground/go1_handstand/trajectories/mujoco_playground_go1_handstand-standard--wydcxp2.json
@@ -0,0 +1,206 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_handstand-standard--wydcxp2",
+ "taskKey": "mujoco-playground/go1_handstand",
+ "dag_sha256": "e5a5a01397ef8139c61c86d38bf73d3ac237709765240ddc30438b6a648fb70c",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 78,
+   "sim_steps": 78,
+   "physics_steps": 390,
+   "steps": 1576,
+   "elapsed_s": 1571.9,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": 0.1723,
+   "end_pose_angle_deg": 106.57,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_handstand-standard--wydcxp2/replay.mp4",
+  "duration_s": 1.6,
+  "fps": 25.0,
+  "frames": 40,
+  "sha256": "f6040baa4aca2b02bdeaa18d3e1e8686750a72b492176a5d736981f81e7dfa0e"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "front_leg_handstand": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.00–0.76 s的并排视图显示Go1从四足站立开始，躯干近水平，前后脚位于地面，未处于前腿倒立姿态。"
+     },
+     {
+      "t": 1.04,
+      "log_step": 69,
+      "text_zh": "观察：视频0.80–1.28 s中Go1前腿向前伸展、后腿弯折，头端仍高于后部，躯干未转到头朝下的近竖直姿态；视频1.32–1.56 s后部继续下沉，最后前腿也折低，未出现合格倒立或保持窗口。已逐帧检查视频0.00–1.56 s的全部40帧。"
+     },
+     {
+      "t": 1.56,
+      "log_step": 71,
+      "text_zh": "观察：log step 71的机器人实时服务status返回episode=11、control_steps=78、done=true、reason=fell、resets_left=0；评分器字段live_matches_replay=1、fell=1，末态end_pose_angle_deg=106.57、end_imu_height=0.1723。推断：该最后一局即受评轨迹；78个控制步×0.02秒=仿真1.56 s，与视频1.56 s的低伏终止姿态核对一致，节点未达成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00 s起Go1前后脚接地支撑身体；视频0.80–1.28 s伸展前腿、弯折后腿时，脚部仍与地面交互支撑，末段后部下沉、前腿折低。实际没有形成双前脚承重且双后脚持续悬空的倒立支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "pose_angle_deg",
+   "value": 106.57,
+   "unit": "deg",
+   "threshold": 30,
+   "direction": "lower_better"
+  },
+  {
+   "name": "imu_height",
+   "value": 0.1723,
+   "unit": "m",
+   "threshold": 0.4,
+   "direction": "higher_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "front_leg_handstand",
+   "t": 1.32,
+   "log_step": 69,
+   "where": "操作中",
+   "evidence_zh": "观察：视频1.28–1.56 s，Go1在伸前腿、收后腿的姿态转换中后部明显下沉，身体倾斜并低伏，前腿末段也折低，始终未起成倒立；log step 69返回ENDED: fell、control_steps=78，log step 71确认episode 11已经结束，之后没有恢复或新一局。"
+  },
+  "unmet": {
+   "front_leg_handstand": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.32,
+    "log_step": 69,
+    "evidence_zh": "观察：视频1.32–1.56 s中Go1后腿收折后身体后部向地面塌下，躯干倾斜，前腿随后折低，未能维持稳定的脚部支撑；同一受评局的机器人实时服务在log step 69报告fell，评分器字段fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": 1.32,
+    "log_step": 69,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 69实际执行的控制脚本按阶段伸前腿、收后腿，并用IMU重力方向及速度调节前腿路径和髋外展；同一步输出的rol由-7.0度增至-16.4度，随后服务报告fell。结合视频0.96–1.56 s的身体倾斜、后部下沉及前腿折低，控制器未能在姿态转换时维持稳定支撑，支持控制层面的归因；不能据此确定某个增益或接触几何就是具体根因。"
+   }
+  ],
+  "gap_zh": "无逐控制步的状态日志或非脚部接触传感记录，无法确定触发fell的具体前腿部位，也无法精确重建支撑力和重心轨迹；因此控制根因不细分。log step 69的knee与h是agent脚本的运动学估算，未当作真实接触或IMU高度。agent在log step 77对初始冲量、触地机制和后脚离地的解释仅为其自述，未据此补记物理事实。"
+ },
+ "summary_zh": "受评的最后一局（episode 11）中，Go1从四足站立伸展前腿并收折后腿，但始终未形成头朝下的近竖直倒立，节点最终0/1。视频约1.32–1.56 s身体后部下沉、前腿折低，实时服务在78个控制步后以fell结束；控制器未能稳定完成姿态转换。缺少逐步状态与接触力数据，控制根因不细分，未将此前重置测试或agent的机制解释当作受评局事实。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "受评局为最后一局（episode 11，log step 71 的服务status显示fell、control_steps=78，78×0.02=1.56 s与视频时长吻合）；视频逐帧显示始终是四足站立后伸腿、后部下沉，从未倒立，节点false、断点、平衡现象与控制归因均有证据支持，未发现需修改之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:10:01Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--wydcxp2/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_handstand-standard--wydcxp2/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "55c2f81a997292b8b89e4687601eb102ece3ff6050bb934215cb7669f0ed8d28",
+   "video_sha256": "f6040baa4aca2b02bdeaa18d3e1e8686750a72b492176a5d736981f81e7dfa0e",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--6xynkat.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--6xynkat.json
new file mode 100644
index 0000000000000000000000000000000000000000..f64b820564a94de05b1d2133a6279902c8892739
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--6xynkat.json
@@ -0,0 +1,198 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-privileged--6xynkat",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.7456,
+   "worst_yaw_err": 0.4176,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 1823,
+   "agent_resets": 1829,
+   "agent_sim_steps": 501825,
+   "agent_replays": 6,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 1,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 36.874,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.255,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0,
+   "over_memory_gb": 1
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--6xynkat/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "f76b69e177f04b54ffe5f46291200e3dbc347ab51e992ae27bb827534f08b26e"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.714,
+      "log_step": null,
+      "text_zh": "观察：视频0.000–2.000 s逐帧检查可见Go1四条腿交替抬落，足端接触方格地面并推动机体移动和转向，已经开始按指令行走的尝试；起始站立不代表完成整个时间表。"
+     },
+     {
+      "t": 6.0,
+      "log_step": null,
+      "text_zh": "观察：视频4.571–7.000 s密集帧及视频8.0–19.0 s抽帧中，Go1持续屈伸腿、交替落足，躯干保持朝上，但地面格线相对机体仅缓慢移动；视频19.286–20.286 s末段逐帧仍如此，未见翻倒。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 180,
+      "text_zh": "观察：log step 179重新生成最终轨迹，log step 180对该交付文件的官方回放报告steps=horizon=1000、fell_at=null、tracking_ok=false、success=false；四段(lin_err m/s, yaw_err rad/s)依次为(0.2680,0.3013)、(0.6557,0.4176)、(0.6247,0.3604)、(0.7456,0.3330)，每段均超过0.2 m/s和0.3 rad/s阈值；最差误差与本次reward字段完全一致。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 180,
+      "text_zh": "推断：完整步数和未翻倒已经满足，但各段速度误差不达标，故唯一的完整时间表节点从未成立，最终为false；此结论来自最终交付的回放结果和本次视频，不将会话内其他试验的结果计入进度。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.000–2.000 s、4.571–7.000 s及19.286–20.286 s，Go1四条腿的足端交替着地、承重和推进，实际支撑对象为方格平地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.7456
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.4176
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.714,
+   "log_step": 180,
+   "where": "操作中",
+   "evidence_zh": "观察：视频0.571–0.857 s机器人已在交替踏步移动；log step 180最终交付文件的官方回放表明，首段平均平面速度误差0.2680 m/s及偏航误差0.3013 rad/s已经超过阈值，后续三段也全部超限。推断：失败发生在按指令行走的执行过程中，首段的整段验收失败未被后续运动修复；视频0.714 s仅为首段运动附近的粗略导航锚点，不是由画面测得的瞬时越阈时刻。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 6.0,
+    "log_step": 180,
+    "evidence_zh": "观察：视频4.571–7.000 s以及视频8.0–20.286 s中，Go1不断交替抬腿落足，机体相对地面格线的平移很慢，后段没有形成成功参考视频6.0–20.0 s所示的持续斜后行位移；末秒逐帧未见翻倒。log step 180确认全程无翻倒但速度跟踪不合格。推断：这是踏步产生的行进速度和方向未满足指令的步态失败。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "观察：log step 146的最终生成器用固定周期N=10的关节正弦目标驱动对角腿，并仅以pitch、roll修正关节目标；代码未将实测平面速度或偏航角速度与指令之差用于反馈。后两段沿用前段参数并将转向偏置置零，最后一段直接复制上一段参数；log step 179重新运行该生成器，log step 180官方重放得到与评分器相同的超限误差。推断：交付的关节运动控制没有实现所需的速度跟踪，解释了视频中的缓慢行进；证据不足以进一步确定为单一参数、逆解或接触几何问题，停在控制大项。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有状态日志或最终交付回放的逐控制步速度序列，视频也不显示速度指令或IMU读数，无法独立重算误差均值或精确定位首个瞬时偏差；分段误差与完整1000步以log step 180最终官方回放及reward为依据。控制根因仅能确定到大项，不能从现有证据锁定某个错误参数或具体接触力问题。",
+  "unmet": {
+   "command_schedule_completed": false
+  }
+ },
+ "summary_zh": "Go1实际执行了全部1000步且未翻倒，但四段平面速度与偏航速度误差均超限，唯一节点“按指令行走”从未达到，最终0/1。失败在行走操作中，视频后段持续踏步而平移缓慢，最终生成器和官方回放支持控制能力不足的归因。没有逐控制步状态日志，精确误差依赖最终官方回放，断点视频时间仅为首段的粗略定位。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "节点false、断点ref、步态现象与控制根因均有log step 179/180官方回放和视频支持，未发现需要修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:08:19Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--6xynkat/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--6xynkat/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "fba5cbafe58be084d4adc55778482bf20a642bab9e909ca47dd924450e536d19",
+   "video_sha256": "f76b69e177f04b54ffe5f46291200e3dbc347ab51e992ae27bb827534f08b26e",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--hkzlm86.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--hkzlm86.json
new file mode 100644
index 0000000000000000000000000000000000000000..f54186194182a86de266d32bb1a37b66d37c432a
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--hkzlm86.json
@@ -0,0 +1,215 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-privileged--hkzlm86",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.2216,
+   "worst_yaw_err": 0.2997,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 46,
+   "agent_resets": 202,
+   "agent_sim_steps": 133025,
+   "agent_replays": 5,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--hkzlm86/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "d7f1cad565055280134004f75bc6c89f7c295aca4cee7cbb0a3e7586318401b8"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.714,
+      "log_step": null,
+      "text_zh": "观察：本次视频 0.00–0.86 s 中，Go1 已抬腿、落足并相对地面格线移动；视频 1.00–5.86 s 中机身转向，视频 6.00–20.29 s 中继续迈步、朝向大体保持，末秒逐帧仍为足端支撑，未见翻倒。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：本次评分器 reward 报告 n_actions=horizon=1000、fell=0、deterministic=1、replay_success=0；worst_lin_err=0.2216 m/s 超过 0.2 m/s，worst_yaw_err=0.2997 rad/s 未超过 0.3 rad/s。"
+     },
+     {
+      "t": 0.714,
+      "log_step": 57,
+      "text_zh": "推断：log step 57 的脚本从已保存的交付文件 trajectory.npz 读取前42个控制动作，复位后测试得到 lin_err=0.2216、yaw_err=0.2997，与最终评分器两个最大误差相同，提示首段已有未消除的跟踪偏差。节点要求整个时间表每段均达标，因此从未成立；不能把视频中行走且未倒视为节点完成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：本次视频 0.00–20.29 s，Go1 四条腿的足端交替接触方格平地、承托机身并推进，前段伴随转向，后段保持朝向移动。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.2216,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.2997,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.714,
+   "log_step": 57,
+   "where": "操作中",
+   "evidence_zh": "推断：唯一节点“按指令行走”在实际行进中未满足速度跟踪验收。log step 57 对交付文件前42个动作的测试已有 0.2216 m/s 平面均值误差，与本次评分器 worst_lin_err 相同且超限；视频 0.57–0.86 s 可见机器人正在迈步、机身低伏但仍有足端支撑，以视频 0.71 s 作为首段动作的粗略导航锚点，不代表能从单帧定位误差越限瞬间。最终评分器仍报告 replay_success=0。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 0.714,
+    "log_step": 57,
+    "evidence_zh": "观察：本次视频 0.00–20.29 s 中，Go1 持续以足端交替着地行走，前段转向，末秒仍保持足端支撑，未见翻倒；inputs/run.json 的评分器 reward 报告 worst_lin_err=0.2216 m/s，超过 0.2 m/s，fell=0、n_actions=horizon=1000。推断：机器人虽完成行走时长且未翻倒，行进速度跟踪仍未达到指令要求，符合“全身·步态”；log step 57 对交付文件前42个动作的测试得到相同误差，支持偏差已存在于首段。视频 0.71 s 仅为首段导航锚点，不能据此认定具体超限瞬间或偏差方向。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 57,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 48 的脚本按四段正确指令生成关节位置动作，调用完整 replay 并保存候选；log step 57 明确读取该交付文件的前42个动作重新测试，初始平面速度均值误差为 0.2216 m/s，50代样条扰动搜索后仍未改善，也未在该步保存改进文件。结合最终评分器相同的超限误差和完整1000步、无翻倒，支持交付的运动控制序列未能达到所需跟踪精度；证据不足以细分为某个反馈增益、接触布局或逆解问题。"
+   }
+  ],
+  "gap_zh": "没有状态日志，最终评分记录只给最大分段误差，没有逐控制步速度或完整分段明细；首段定位由 log step 57 对交付文件前42个动作的测试与最终评分数值吻合而推断。已检查本次全部143帧及成功参考视频；两者均能行走、转向并保持直立，跟随视角与7 fps录像不足以可靠分辨这项小幅速度超限究竟来自过快、过慢还是横向偏差，但评分器字段结合视频足以支持“行进速度跟踪未达标且未翻倒”的步态现象。根因仅支持到“控制”，不能确定更细机制；未拍摄的末尾控制动作由评分器1000步及fell字段补充。"
+ },
+ "summary_zh": "Go1 在本次视频中持续以足端支撑行走，前段转向、后段保持朝向移动；评分器确认完成1000步且未翻倒，但最大分段平面速度误差为0.2216 m/s，超过0.2 m/s，唯一节点“按指令行走”最终为0/1。log step 57 对交付文件前42个动作的测试得到相同超限误差，支持首段运动控制精度不足；现象记为“全身·步态”：行进速度跟踪未达标且未翻倒；无逐步状态记录，录像不能可靠区分偏差来自过快、过慢还是横向偏移。",
+ "conf": 7,
+ "review": {
+  "status": "agreed",
+  "rounds": 2,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "fix",
+    "conf": 7,
+    "summary_zh": "节点值、断点和根因基本成立，但现象层留空不当：评分器 worst_lin_err=0.2216 m/s 超过 0.2 且 fell=0，这是已被准入证据支持的“走得没达到要求但没摔倒”，应记步态。",
+    "n_issues": 1,
+    "n_major": 1
+   },
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 2,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标签成立：评分器 worst_lin_err=0.2216>0.2、fell=0、1000 步，节点 false、断点在唯一节点、步态现象与“控制”根因均有准入证据支持；log step 57 对交付文件前42动作的测试得到 0.2216/0.2997，与评分器一致，首段定位作为推断标注正确。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:25:29Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--hkzlm86/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--hkzlm86/04_review_r1"
+   },
+   {
+    "stage": "reconcile",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--hkzlm86/05_reconcile_r1"
+   },
+   {
+    "stage": "traj_review",
+    "round": 2,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--hkzlm86/04_review_r2"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "ce475c6eecca97d99d3db7e70ad87ef98db09c5e9952a90ee917a0a3983ec1f9",
+   "video_sha256": "d7f1cad565055280134004f75bc6c89f7c295aca4cee7cbb0a3e7586318401b8",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--hyoy23r.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--hyoy23r.json
new file mode 100644
index 0000000000000000000000000000000000000000..23d56d3944bcd72074544472309bcea38513ad90
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--hyoy23r.json
@@ -0,0 +1,183 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-privileged--hyoy23r",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.1619,
+   "worst_yaw_err": 0.1857,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 221,
+   "agent_resets": 1679,
+   "agent_sim_steps": 50140,
+   "agent_replays": 2,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 15.334,
+   "peak_processes": 11,
+   "peak_disk_gb": 0.825,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--hyoy23r/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "97c1a800f4dc7786413f0e3bc720f304b6fedfa2be7ada65b8f82f3daeaededd"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [
+     [
+      20.286,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.000 s的双画面中，黑色Go1以屈腿姿态支撑在方格平地上；视频 0.143–2.000 s可见足端交替抬起、落地，机器人开始移动并改变朝向。初始站姿不等于完成整段指令行走。"
+     },
+     {
+      "t": 4.857,
+      "log_step": null,
+      "text_zh": "观察：视频 0.000、1.000、2.000、3.000、4.000 s的机器人朝向逐渐改变；视频 4.286–6.571 s逐帧可见Go1继续交替踏步，转向逐渐停止，躯干保持朝上。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：视频 7.000–20.000 s每秒取帧及18.857–20.286 s逐帧检查中，Go1保持近似朝向，地面格线相对足端持续位移；末帧视频 20.286 s仍在踏步，未见躯干翻倒或趴地。"
+     },
+     {
+      "t": null,
+      "log_step": 22,
+      "text_zh": "观察：log step 20、21调用同一最终交付轨迹的独立回放，log step 22返回两份一致的回放结果：steps=horizon=n_actions=1000，success=true，tracking_ok=true，fell_at=null，state_sha均为321566f3ce2c；四段平面速度均误差分别为0.1619、0.1246、0.1079、0.1186 m/s，偏航角速度均误差分别为0.1857、0.0953、0.0899、0.0725 rad/s，均低于任务阈值。"
+     },
+     {
+      "t": null,
+      "log_step": 23,
+      "text_zh": "观察：log step 23显示生成器重生成1000步成功轨迹，并通过动作数组哈希一致性断言；评分器reward同时报告success=1、replay_success=1、deterministic=1、n_actions=horizon=1000、fell=0。最终交付的验收结果与录像所见持续行走相符。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 22,
+      "text_zh": "推断：本节点必须完成整段1000控制步并通过各段平均误差验收，因此只在片尾记为达成，不在起始站立或前段行走时提前计入。以已检查的末帧视频 20.286 s作为约半秒精度的完成锚点；该时间不是由agent工作时间换算，完整步数与数值达标依据log step 22的最终回放及评分器字段。没有状态日志，且DAG注明录像最后采样止于第994控制步，未拍到的末尾6步不能从画面独立复核。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.000–2.000 s、4.286–6.571 s及18.857–20.286 s的连续帧中，Go1的足端轮流落在方格平地上，另一些足端抬起摆动；四条腿通过足端着地支撑躯干并使机器人移动。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.1619
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.1857
+  }
+ ],
+ "failure": null,
+ "summary_zh": "Go1完成唯一的“按指令行走”节点：视频前段持续踏步转向，后段保持朝向移动，直到视频 20.286 s末帧仍未见翻倒，节点仅在片尾记为完成。最终交付轨迹的两次独立回放均完成1000步、无翻倒且末态哈希相同（log step 22），最差分段平面速度均误差0.1619 m/s、偏航角速度均误差0.1857 rad/s均达标，评分器判定成功，无失败断点。没有状态日志，精确误差和录像未采样的最后6控制步依赖最终回放与评分器字段，完成时刻以视频末帧作近似锚点；本次oracle与被标注运行相同，未作独立参考比较。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "单节点成功运行：视频末帧Go1仍在踏步未翻倒，评分器字段（success=1、fell=0、worst_lin_err 0.1619、worst_yaw_err 0.1857）支持节点达成，无失败断点；时钟与log step引用均有效，未发现需改动的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:17:30Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--hyoy23r/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--hyoy23r/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "7cc85270f4bc22f674687d974d09fe0aa5569619ffcb1fc6edc6d060088c7f22",
+   "video_sha256": "97c1a800f4dc7786413f0e3bc720f304b6fedfa2be7ada65b8f82f3daeaededd",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--jjb2xye.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--jjb2xye.json
new file mode 100644
index 0000000000000000000000000000000000000000..fb8101e384acc6c229887900eec69eb1055cea8f
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--jjb2xye.json
@@ -0,0 +1,198 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-privileged--jjb2xye",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 1.5101,
+   "worst_yaw_err": 0.604,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 58,
+   "agent_resets": 1002,
+   "agent_sim_steps": 627910,
+   "agent_replays": 2,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 1.444,
+   "peak_processes": 18,
+   "peak_disk_gb": 0.981,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--jjb2xye/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "1ebcd98a1f6653c9409c3f8cd9b0d4d887183e8d0d30b0cae3d8f809e42a72e9"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.714,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00–0.86 s，Go1由初始站姿开始屈伸踏步并转向，地面格线相对机身的平移较小；视频 0.86–5.57 s 继续轮换足端支撑并转动，没有出现躯干翻倒。"
+     },
+     {
+      "t": 8.571,
+      "log_step": null,
+      "text_zh": "观察：视频 5.71–11.29 s，Go1持续踏步，右侧视图中机身从斜侧向转到正对相机、再转到另一斜侧向；后段要求保持朝向的行走没有稳定实现。视频 11.43–20.29 s 仍在小范围踏步移动，末尾视频 19.43–20.29 s 各帧均保持躯干上表面朝上。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 110,
+      "text_zh": "观察：log step 110 将最终参数生成的轨迹保存，输出 fell_at=None、worst_lin_err=1.5101；log step 112 确认 actions 形状为 (1000,12)。本次评分器 reward 记录 n_actions=horizon=1000、deterministic=1、fell=0，但 worst_lin_err=1.5101 m/s、worst_yaw_err=0.604 rad/s，分别超过 0.2 m/s、0.3 rad/s 阈值。"
+     },
+     {
+      "t": null,
+      "log_step": 108,
+      "text_zh": "推断：log step 108 中唯一未翻倒候选的参数由 log step 110 原样用于最终保存，且最坏误差与本次评分器一致；该候选各段平面误差约为 1.51、0.65、0.61、0.55 m/s，均超限。因此完整时间表验收从未成立，不能因一直站立或跑满步数给节点记完成；log step 101 的早期翻倒轨迹已被覆盖，不属于本次评分录像。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.57–20.29 s，Go1四条腿反复屈伸，足端轮流落在方格平地上支撑躯干，并伴随机身转向和小范围平移。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 1.5101
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.604
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.714,
+   "log_step": 110,
+   "where": "操作中",
+   "evidence_zh": "推断：视频 0.57–0.86 s 已开始踏步，但格线显示前进位移很小；log step 108 的首段平面误差约 1.51 m/s，该候选由 log step 110 保存，且最终评分器 worst_lin_err=1.5101。失败发生在按指令行走的执行中，后续完成全部控制步仍未修复整段验收；视频 0.71 s 仅为早段跟踪不足的粗略导航锚点，并非逐步误差首次越限的精确时刻。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 0.714,
+    "log_step": null,
+    "evidence_zh": "观察：视频 0.57–0.86 s 及视频 2.86–5.57 s，Go1足端反复抬落，机身平移相对格线较小；视频 8.57–11.29 s 机身还从面向相机转为侧向，未在后段稳定保持朝向。视频 19.43–20.29 s 仍直立踏步；评分器 fell=0，故可见问题是行走速度和方向跟踪不足。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 110,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 19 的脚本按速度指令生成运动学步态关节目标；log step 108 搜索所得候选被 log step 110 生成并保存，保存时已测得 worst_lin_err=1.5101，最终评分器复现同值且 deterministic=1。结合视频 0.57–20.29 s 有持续踏步却未实现要求的速度与朝向，说明所生成的步态控制未能把指令转成达标的机身运动。证据不足以进一步区分接触布局、关节目标跟踪或反馈律问题。"
+   }
+  ],
+  "gap_zh": "没有状态日志，视频也无逐步IMU速度、接触力或关节目标与实测值，无法独立重算每段误差或定位首个偏差的精确时刻；完整步数及精确误差采用本次评分器字段。DAG注明录像最后采样为控制步994，末尾6个控制步未入镜。agent 在 log step 113 将问题归因于软接触和侧向支撑，但现有证据未排除其他控制原因，故不采纳其具体物理归因。",
+  "unmet": {
+   "command_schedule_completed": false
+  }
+ },
+ "summary_zh": "Go1完成全部1000个控制步，录像中持续踏步且未翻倒，但“按指令行走”节点未达成，最终为0/1。早段前进不足、后段朝向仍变化，评分器最坏平面速度误差1.5101 m/s、偏航角速度误差0.604 rad/s均超限，归为步态现象及控制根因。无状态日志，精确误差依赖评分器，不能确认更细的控制机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成、final=0、断点、步态现象与控制根因均有视频、log step 108/110 及评分器字段支持，log step 引用均为 steps[].id 且存在，未发现需改动的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:08:25Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--jjb2xye/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--jjb2xye/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "a4c4ef67a8ef37d8ca4ae25571de9d0686c58d76603f0c65df6726e7c0715a12",
+   "video_sha256": "1ebcd98a1f6653c9409c3f8cd9b0d4d887183e8d0d30b0cae3d8f809e42a72e9",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--k4kjm9y.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--k4kjm9y.json
new file mode 100644
index 0000000000000000000000000000000000000000..076f3c51d2e3fda0ce3610ae9f873f0e24a932d9
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--k4kjm9y.json
@@ -0,0 +1,198 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-privileged--k4kjm9y",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 1.5329,
+   "worst_yaw_err": 0.8488,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 66,
+   "agent_resets": 223,
+   "agent_sim_steps": 236440,
+   "agent_replays": 3,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.878,
+   "peak_processes": 10,
+   "peak_disk_gb": 0.832,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--k4kjm9y/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "7befa3986965c9e145efe9d6f6832924b752d15fd6973b2839bbe943b7ed6646"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.714,
+      "log_step": null,
+      "text_zh": "观察：视频0.00–2.86 s双画面中，Go1先屈腿下沉，随后小幅周期性摆腿，躯干相对附近地面格线移动很少，未见前段指令要求的持续快速平移和明显转向。"
+     },
+     {
+      "t": 6.0,
+      "log_step": null,
+      "text_zh": "观察：视频4.00–6.86 s逐帧及视频7.00–20.00 s每秒抽帧中，Go1继续小幅摆腿，整体仅缓慢移位、转动；视频19.29–20.29 s逐帧仍由足端支撑，未见翻倒。"
+     },
+     {
+      "t": null,
+      "log_step": 75,
+      "text_zh": "观察：log step 75中，agent脚本保存的1000步候选经独立进程回放，报告steps=1000、fell_at=null、tracking_ok=false；首段lin_err=1.5329 m/s、yaw_err=0.5467 rad/s，最差yaw_err=0.8488 rad/s。最终评分器reward同样记录n_actions=horizon=1000、fell=0、deterministic=1及最差误差1.5329 m/s和0.8488 rad/s，均超过DAG阈值0.2 m/s和0.3 rad/s。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "推断：结合视频中推进与转向不足、最终评分器的两项误差超限，唯一节点“按指令行走”从未满足完整时间表验收条件，final=false；保持站立和执行足够步数不能单独使该节点成立。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00–20.29 s，Go1四条腿的足端接触方格平地承重，并在小幅周期性屈伸、摆腿时维持躯干支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 1.5329,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.8488,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.714,
+   "log_step": 75,
+   "where": "操作中",
+   "evidence_zh": "观察：视频0.57–0.86 s，Go1已经开始摆腿，但相对地面格线移动很少，未形成首段指令要求的前进与右移；log step 75已保存候选的首段平面和偏航误差均超限，最差平面误差与最终评分器一致。推断：断点为正在执行的“按指令行走”，视频0.71 s仅作早期推进不足的导航锚点；后续持续摆腿没有完成该步骤的验收。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 0.714,
+    "log_step": null,
+    "evidence_zh": "观察：视频0.57–2.86 s，Go1腿部虽有周期运动，躯干相对地面格线却几乎停留原处；视频4.00–6.86 s仍没有明显完成前段转向，后段视频7.00–20.29 s仅缓慢移位和转动。推断：行进的推进和转向不足；机器人仍保持足端支撑，故对应步态未达要求。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 75,
+    "source_zh": "日志",
+    "evidence_zh": "推断：agent在log step 27–75反复调试足端周期轨迹、逆解、步幅和相位；log step 75将平滑后的速度指令转换为周期足端目标并生成关节轨迹，保存后独立回放仍报告tracking_ok=false，其最差误差与最终评分器一致。视频0.57–6.86 s及后段显示关节摆动没有产生所需的推进和转向，支持控制实现不足；现有记录不能单独证明某个增益、接触布局或逆解公式是决定性原因。"
+   }
+  ],
+  "gap_zh": "没有状态日志，无法独立复核逐控制步的IMU速度、足地接触力和精确误差来源；视频末尾未覆盖DAG注明的最后6个控制步，完整步数和无翻倒依评分器字段。日志包含多个候选，log step 83的另一回放最差偏航误差为0.8744 rad/s，不能替代最终评分器的0.8488 rad/s；控制根因因此只标到大类。",
+  "unmet": {
+   "command_schedule_completed": false
+  }
+ },
+ "summary_zh": "Go1完成1000个控制步且评分器未记录翻倒，但唯一节点“按指令行走”未达成，最终0/1。视频显示机器人小幅摆腿、推进及转向不足，首段已出现未修复的行进失败；最差平面误差1.5329 m/s、偏航误差0.8488 rad/s均超限，根因支持控制实现不足。没有状态日志，精确误差来源和控制子类型无法独立核实。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "唯一节点未达成、final=0 与评分器误差（1.5329/0.8488 超阈值）一致；视频显示小幅摆腿、推进不足，步态现象与控制根因有证据支持，未发现需修改之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:14:10Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--k4kjm9y/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--k4kjm9y/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "41505100ad3815754a663bd41507ad77063cbff37bc5f123823d5235929951a8",
+   "video_sha256": "7befa3986965c9e145efe9d6f6832924b752d15fd6973b2839bbe943b7ed6646",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--vlbaxpe.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--vlbaxpe.json
new file mode 100644
index 0000000000000000000000000000000000000000..8b307fb4aeae46a7840556261d84463f86d1106c
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--vlbaxpe.json
@@ -0,0 +1,194 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-privileged--vlbaxpe",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.8245,
+   "worst_yaw_err": 0.2775,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 0,
+   "agent_resets": 0,
+   "agent_sim_steps": 0,
+   "agent_replays": 0,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 1.578,
+   "peak_processes": 12,
+   "peak_disk_gb": 0.241,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--vlbaxpe/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "275e2b8184dd5c9adf88bc1609ec1d86aa8d952c4d199c6739f2e5674b156b7a"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.429,
+      "log_step": null,
+      "text_zh": "观察：视频0.000 s机器人由四足支撑；视频0.286–0.429 s四腿屈曲外展、躯干迅速降至地面高度，视频0.857–4.857 s仍低伏摆腿转动，未恢复正常站立行走。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：视频5.143–20.286 s机器人保持低伏姿态，地面格线与机身相对位置几乎不变；已逐帧核对视频19.286–20.286 s，末尾仍未起身或完成持续斜向行走。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：inputs/run.json评分器字段n_actions=horizon=1000、deterministic=1、fell=0、replay_success=0；worst_lin_err=0.8245 m/s超过0.2 m/s，worst_yaw_err=0.2775 rad/s低于0.3 rad/s。完整动作数已满足，但平面速度跟踪验收未通过。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.000–0.286 s Go1足端接触方格平地，支撑躯干并开始动作；此后腿部仍在地面附近屈伸，但躯干降至贴地高度，未维持正常站立支撑。不能仅凭渲染断言躯干具有物理碰撞接触。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.8245
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.2775
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.429,
+   "log_step": null,
+   "where": "操作中",
+   "evidence_zh": "观察：视频0.286–0.429 s机器人在开始按速度指令运动时，腿部屈曲、躯干降到地面高度；视频0.857–4.857 s继续低伏摆腿，视频5.143–20.286 s几乎停在原处，未恢复支撑行走。评分器报告1000步已执行、平面速度误差0.8245 m/s超限。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 0.429,
+    "log_step": null,
+    "evidence_zh": "观察：视频0.000 s机身由足端支撑离地；视频0.286–0.429 s腿部折下、躯干降到地面高度，此后虽继续摆腿却始终未恢复站立支撑，视频20.286 s仍低伏。这里是支撑塌陷，不声称躯干翻转超过90°；评分器fell=0仅说明未触发该翻倒阈值。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": 0.429,
+    "log_step": 39,
+    "evidence_zh": "推断：agent在log step 23编写优化脚本，将参数化周期关节位置目标直接保存为交付轨迹，log step 36运行该优化；log step 38–39的已保存最佳结果为1000步、无阈值翻倒、末态IMU高度约-0.00084 m，最差平面及偏航误差分别为0.8244956 m/s与0.2774777 rad/s，四舍五入后均与最终评分字段吻合。结合视频0.286–0.429 s支撑塌陷及视频5.143–20.286 s持续低伏，说明交付的关节目标序列未形成能支撑机身并推进的运动，归为控制；现有证据不足以区分接触设计、轨迹生成或具体参数问题。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有状态日志或最终交付轨迹的逐步IMU、关节及接触力记录，无法独立定位具体失效关节或控制参数，根因保留在控制大类。log step 39是会话中已保存最佳候选的诊断，凭其误差与评分器吻合及录像姿态对应来关联交付；log step 43–46是后续另一方案的测试和优化脚本语法错误，不作为正式回放的翻倒或脚本中断现象。视频末次采样未覆盖最后少量控制步，完整1000步依据最终评分字段。"
+ },
+ "summary_zh": "机器人交付并执行了1000步轨迹，但“按指令行走”节点从未达成，最终进度0/1；最差平面速度误差0.8245 m/s超限，偏航误差0.2775 rad/s达标。视频约0.429 s四足支撑塌陷、躯干贴地，随后一直未恢复行走，归因于交付关节目标序列的控制失败；评分器fell=0仅表示没有超过90°的翻倒。缺少逐步状态与接触力数据，无法确定具体控制子因。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "节点未达成、final=0、断点与平衡现象、控制根因均有视频与评分器/日志支撑；log step 39 的误差0.8244956/0.2774777 与评分字段吻合，推断标注正确。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:13:18Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--vlbaxpe/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--vlbaxpe/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "e42fe3413420acefa929b3fe946661ee3426b5412016b60f2b26c41cf769087e",
+   "video_sha256": "275e2b8184dd5c9adf88bc1609ec1d86aa8d952c4d199c6739f2e5674b156b7a",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--y745osd.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--y745osd.json
new file mode 100644
index 0000000000000000000000000000000000000000..e63bac59608173c36b5f5db9711fad892ba2f889
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-privileged--y745osd.json
@@ -0,0 +1,188 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-privileged--y745osd",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 1.59,
+   "worst_yaw_err": 0.7047,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 25,
+   "agent_resets": 37,
+   "agent_sim_steps": 85895,
+   "agent_replays": 8,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.638,
+   "peak_processes": 12,
+   "peak_disk_gb": 0.241,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--y745osd/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "d3ea079f808bc018522ffbbbc7c0299d932ea36dcf7a56222212a9b1b6d1b38c"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 40,
+      "text_zh": "观察：视频 0.0–2.0 s 的逐帧检查显示，Go1仅在起始时轻微屈腿和调整躯干，没有启动交替迈步；log step 40 中 agent 的脚本将同一 home 关节目标重复1000次并保存为最终轨迹。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 41,
+      "text_zh": "观察：视频 0.0–20.286 s 的逐秒双画面及视频 19.0–20.286 s 的逐帧检查显示，Go1持续原地站立，足端附近格线与机身朝向在起始收敛后基本不变，未出现指令要求的转向和斜后行。log step 41 明确对已交付轨迹调用最终回放：steps=1000、fell_at=null、tracking_ok=false，四段线速度平均误差依次为1.59、0.6678、0.6635、0.824 m/s，前两段偏航误差为0.6193、0.7047 rad/s，超出0.2 m/s和0.3 rad/s阈值；评分器reward同样报告replay_success=0、fell=0。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–20.286 s，Go1的四条腿和足端接触方格平地承重，实际保持原地站姿；未见交替迈步产生持续推进。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 1.59
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.7047
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.0,
+   "log_step": 40,
+   "where": "首步前",
+   "evidence_zh": "观察：视频 0.0–2.0 s 只有初始姿态调整，随后直至视频 20.286 s 均维持原地站姿，按指令行走始终未启动。log step 40 保存的是1000行固定home关节目标，log step 41 的最终回放确认完整执行但速度跟踪失败；会话中其他控制测试的翻倒没有出现在这份交付轨迹中。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "没尝试",
+    "item": "没尝试",
+    "t": 0.0,
+    "log_step": 41,
+    "evidence_zh": "观察：视频 0.0–20.286 s，Go1在双画面中维持同一地面区域和基本固定朝向，没有启动指令行走。log step 40 的保存代码和log step 41 的FALLBACK分支均重复固定home姿态，最终回放tracking_ok=false；交付的是从未启动行走的静态基线。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 39,
+    "source_zh": "日志",
+    "evidence_zh": "推断：agent未能生成稳定的步态控制，因而最终改交只能站立的轨迹。log step 38 在逆解往返校验通过后，原地小跑测试仍报告fell_at=33；log step 39 的零指令、不同抬脚及高度控制设置测试分别报告fell_at=16、23、33；log step 40 先保存固定home轨迹，随后进行的步态诊断仍显示生成的关节目标明显偏离home并发生失稳，log step 41 将复现脚本切到FALLBACK=True。这些会话测试用于解释交付选择，不作为评分视频中的翻倒现象；无法据此唯一确定控制子类。"
+   }
+  ],
+  "gap_zh": "没有逐控制步的状态日志，无法独立重算IMU速度及各段均值，精确验收量依据log step 41 对交付文件的最终回放和评分器字段；视频末尾也不能独立证明所有1000步。会话诊断未定位修正逆解后的剩余失稳机制，因此根因停留在“控制”，不细分或把测试翻倒算入评分轨迹。"
+ },
+ "summary_zh": "观察：评分轨迹在视频 0.0–20.286 s 仅有起始轻微姿态调整，随后原地站立；最终回放完成1000步且无翻倒，但最大线速度误差1.59 m/s、最大偏航误差0.7047 rad/s均超限，唯一节点“按指令行走”未完成（0/1）。推断：断点为该节点的首步前，交付轨迹没有启动行走；会话控制测试反复失稳后，agent改交固定站姿，根因归为控制。缺少逐步状态日志，精确误差依赖最终回放与评分器，剩余控制缺陷无法细分。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "视频 0–20.3 s（双画面）Go1 始终原地站立，节点未达、final=0 与评分器字段（success=0、fell=0、worst_lin_err 1.59）一致；首步前 + 没尝试符合 03 对“交付静态基线”的规定，根因停在控制并标为推断，且未把会话翻倒当作评分视频现象，未发现需改之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:09:20Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--y745osd/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-privileged--y745osd/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "726748aee6abc691bf216c4620a29db1a2e109acdf1a91775e059646e07da659",
+   "video_sha256": "d3ea079f808bc018522ffbbbc7c0299d932ea36dcf7a56222212a9b1b6d1b38c",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--4h4nbt3.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--4h4nbt3.json
new file mode 100644
index 0000000000000000000000000000000000000000..858c421c7fe88259d1479b0224148eb5078816bf
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--4h4nbt3.json
@@ -0,0 +1,215 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-standard--4h4nbt3",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 1000,
+   "sim_steps": 1000,
+   "physics_steps": 5000,
+   "steps": 1842,
+   "elapsed_s": 1585.0,
+   "video_rendered": 1,
+   "fell": 0,
+   "worst_lin_err": 0.4798,
+   "worst_yaw_err": 0.5662,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.566,
+   "peak_processes": 10,
+   "peak_disk_gb": 0.831,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-standard--4h4nbt3/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "a03895c334b4eaff1be08bafc81ef271676d83432d7ef07ec825e78d1a8a0a99"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.714,
+      "log_step": 74,
+      "text_zh": "观察：log step 74使用最后一次重置启动名为final的1000步控制程序；视频0.000–2.143 s中Go1的足端交替抬起和落地，躯干保持正立，地面格线相对机身移动，说明最终评分回合确实开始行走。此前第1–10回合的测试和翻倒不计入该轨迹。"
+     },
+     {
+      "t": 0.714,
+      "log_step": 77,
+      "text_zh": "观察：视频0.429、0.571、0.714、0.857 s中Go1继续迈步而未翻倒；log step 77对最终回合记录的自算分析显示，仿真0.50–0.84 s内平均IMU速度约为[1.118,-0.293,-0.016]米/秒，而开场指令为[1.413,-0.611]米/秒，平面及偏航平均误差自算为0.4417米/秒、0.5469弧度/秒。该分析是agent脚本对最终回合传感读数的计算，非评分器原生值。"
+     },
+     {
+      "t": 5.143,
+      "log_step": 83,
+      "text_zh": "观察：视频4.143–4.857 s机身仍在转向，视频5.143–6.714 s逐渐保持朝向并继续斜向迈步；log step 83打印最终第11回合control_steps为250、275时的IMU偏航角速度为0.084、0.119弧度/秒。推断：按任务DAG注明的每7控制步一帧和视频7 fps，控制步s对应视频s/49秒；第238步撤销偏航指令约对应视频4.857 s，与此后画面中转向减弱相符，仅用于核对两种时钟。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 95,
+      "text_zh": "观察：视频0–20.286 s的全程抽帧，以及视频19.000–20.286 s逐帧检查，均显示Go1以足端支撑躯干并持续迈步，末帧未翻倒。log step 95的机器人服务状态为episode=11、control_steps=1000、done=True、reason=horizon，证明最终回合完成全部控制步。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 96,
+      "text_zh": "观察：inputs/run.json的评分器字段为fell=0、n_actions=1000、live_matches_replay=1，但worst_lin_err=0.4798米/秒、worst_yaw_err=0.5662弧度/秒，分别高于0.2和0.3的阈值。log step 96对final记录的自算结果也只有开场段两项超限，后面三段均在阈值内。推断：本节点要求完整时间表的每段均达标，故即使全程行走且未翻倒，节点从未达到，最终为false。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.000–20.286 s的跟随及侧面画面可见Go1四足轮流着地和抬起，足端接触方格平地以承重和推进，躯干离地；视频19.000–20.286 s仍在交替迈步。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.4798,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.5662,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.714,
+   "log_step": 77,
+   "where": "操作中",
+   "evidence_zh": "观察：视频0.429–0.857 s中Go1正在足端交替支撑的行走动作中，尚未翻倒；log step 74明确启动最终第11回合，log step 77对该回合开场仿真0.50–0.84 s的自算结果已显示前进及侧向速度不足且偏航跟踪误差超限。评分器最终worst_lin_err=0.4798、worst_yaw_err=0.5662均超标，log step 95显示仍完成1000步。推断：最早未修复的失败是按指令行走的开场段跟踪不足，视频0.714 s作为该段的粗略导航锚点，后续走稳不能改写开场段的历史误差。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 0.714,
+    "log_step": 77,
+    "evidence_zh": "观察：视频0.429–0.857 s显示Go1持续低姿态迈步，视频1.143–5.143 s随后转向，未发生摔倒；log step 77记录最终回合开场仿真0.50–0.84 s平均前进速度约1.118米/秒、侧向速度约-0.293米/秒，未跟上指令的1.413、-0.611米/秒；自算偏航误差也超限，评分器两项最坏段误差同样超限。推断：所需的前进、侧移和转向跟踪未实现，属于无翻倒的步态不足。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "log_step": 74,
+    "evidence_zh": "推断：agent的速度反馈步态控制未能在短开场段内维持所需平移与偏航速度。log step 11的控制代码直接读取IMU速度和角速度，以速度误差生成期望加速度和偏航力矩并转换为关节目标；log step 74最终版本在仿真0.32–0.84 s提高速度反馈增益后启动第11回合，log step 75在第25控制步打印速度[1.372,-0.651,-0.018]、偏航角速度0.269，而log step 77的最终回合整段自算平均速度退到[1.118,-0.293,-0.016]，两项误差超限。视频0.429–0.857 s仍在行走、评分器fell=0，支持跟踪不足而非翻倒；这些证据不足以锁定某一增益、摩擦参数或具体足端几何为根因。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有状态日志，画面不能独立复算IMU误差或确定控制器内部哪项设计造成跟踪不足。log step 77、96的agent脚本对最终回合自算开场误差为0.4417米/秒和0.5469弧度/秒，与评分器的0.4798和0.5662略有差异；两者均超限，原生指标以评分器为准，采样或统计边界差异无法复核。按DAG记录的渲染采样方式，视频20.286 s末帧为第994控制步，最后6步未拍摄，完成1000步及无翻倒的最终结论结合log step 95服务状态和评分器fell=0。"
+ },
+ "summary_zh": "最终评分的是第11回合：Go1完成1000个控制步，视频全程及末秒密集帧均显示持续迈步而未翻倒，但“按指令行走”节点未达到，最终0/1。开场段的平移与偏航跟踪不足构成全身·步态现象，结合最终控制程序及传感分析归因于反馈控制；评分器最坏段误差为0.4798米/秒、0.5662弧度/秒，均超阈值。无状态日志，agent自算误差与评分器数值略有不同，且末6控制步没有视频，精确统计及最终步数需依赖评分器与服务状态。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "节点未达到、final=false、断点在开场段（步0–42，视频约0–0.86 s）与评分器worst误差及log step 77/96自算（仅开场段超限）一致；步态+反馈控制归因有证据支持，时钟换算有说明，未发现需修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:11:39Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--4h4nbt3/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--4h4nbt3/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "08c360739fddfad2f3185f7c03d6150aa9ca73b28d1a2d08b07edb58aab0382e",
+   "video_sha256": "a03895c334b4eaff1be08bafc81ef271676d83432d7ef07ec825e78d1a8a0a99",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--7rd5swv.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--7rd5swv.json
new file mode 100644
index 0000000000000000000000000000000000000000..57c61d07f285f86c968b6fdde16c3ac04f1bf303
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--7rd5swv.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-standard--7rd5swv",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 99,
+   "sim_steps": 99,
+   "physics_steps": 495,
+   "steps": 502,
+   "elapsed_s": 471.3,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 1.59,
+   "worst_yaw_err": 0.8458,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-standard--7rd5swv/replay.mp4",
+  "duration_s": 2.0,
+  "fps": 25.0,
+  "frames": 50,
+  "sha256": "09cee26f523ac2d487abdf9613784b284bf0b139a926b1826b6ddebc84c09e45"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.8,
+      "log_step": 20,
+      "text_zh": "观察：视频 0.00–0.80 s 中 Go1 主要保持屈腿站姿，足端接触平地，未呈现成功参考中连续转向行走的运动；log step 20 的最终控制脚本先保持初始姿态 0.84 仿真秒，再启动对角腿交替步态。"
+     },
+     {
+      "t": 1.52,
+      "log_step": 23,
+      "text_zh": "观察：视频 0.88–1.44 s 中 Go1 开始交替抬腿并移动，视频 1.52–1.96 s 中躯干侧倾持续加剧，末帧明显翻转；机器人没有恢复直立或继续后续指令段。足端着地支承与抬腿交替在双视角中可见。"
+     },
+     {
+      "t": 1.96,
+      "log_step": 24,
+      "text_zh": "观察：log step 23 明确启动最后评分回合并输出 DONE fell 99；log step 24 的机器人服务状态（仿真服务返回的回合计数及终止读数）确认为 episode=11、control_steps=99、done=True、reason=fell、resets_left=0。评分器字段 sim_steps=99、fell=1、live_matches_replay=1，且 worst_lin_err=1.59 m/s、worst_yaw_err=0.8458 rad/s，均高于节点阈值；完整 1000 控制步从未完成，故节点没有成立区间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–1.44 s 中 Go1 的四足接触方格平地支撑躯干，迈步时足端交替抬起和着地；视频 1.52–1.96 s 中原有足端支撑未能阻止侧翻。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 1.59
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.8458
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 1.6,
+   "log_step": 23,
+   "where": "操作中",
+   "evidence_zh": "观察：Go1 正在尝试按速度指令迈步，视频 1.52–1.96 s 中躯干逐渐侧翻，未恢复；log step 23 的最后评分回合输出 DONE fell 99，log step 24 的服务状态确认第 11 回合因 fell 结束。失败步骤是按指令行走。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.8,
+    "log_step": 23,
+    "evidence_zh": "观察：视频 1.52 s 后 Go1 躯干侧倾加大、部分腿悬空，视频 1.84–1.96 s 中躯干进一步滚转至明显翻倒姿态；log step 24 的服务返回 reason=fell，评分器也记录 fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：log step 20–21 的最终脚本使用本体速度和重力方向反馈生成足端目标，每 3 个控制步更新，并在倾斜过大时切换四足落地恢复姿态；log step 23 的评分回合输出中，重力方向由 g=[0.01,-0.05,-1.00] 恶化为 g=[-0.11,-0.97,-0.20]，且出现恢复标记 E，随后仍输出 DONE fell 99。结合视频 1.52–1.96 s 的持续侧翻，控制器未能稳定行走中的躯干姿态；证据不足以进一步确定接触布局、逆解或具体反馈参数中哪一项导致失稳。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有逐控制步状态日志或足端受力记录，日志仅有最终脚本的稀疏传感输出，不能独立重算各段误差，也不能细分侧翻的具体控制机制。评分器的误差值、翻倒标记及服务回合终止读数与视频相符；此前重置掉的回合仅属会话测试，不计入本次节点进度或失败现象。"
+ },
+ "summary_zh": "评分对象为第 11 回合：Go1 先保持站姿再尝试迈步，在视频约 1.5–1.96 s 持续侧翻，服务记录仅完成 99/1000 控制步即因 fell 结束，唯一节点“按指令行走”未达成。失败现象为平衡，最终脚本的恢复控制未能阻止翻倒，且评分器报告的平面速度和偏航误差均超阈值。没有逐步状态及受力日志，无法独立重算误差或确定具体失稳机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标签与视频、日志和评分器字段一致：评分回合（第11回合）99/1000 控制步因 fell 结束，唯一节点未达成；断点、平衡现象和控制根因均有证据支持，时间都在视频 2.0 s 以内，未把 agent 时间换算成视频时间。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:24:15Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--7rd5swv/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--7rd5swv/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "9cc1dd1531fa39fbeb8602964bb6f1e8f28c32efeebc09d93c470c6d7ced723c",
+   "video_sha256": "09cee26f523ac2d487abdf9613784b284bf0b139a926b1826b6ddebc84c09e45",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--fskgfca.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--fskgfca.json
new file mode 100644
index 0000000000000000000000000000000000000000..51e684c9d05692481ec3489ebea876527b16e5a1
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--fskgfca.json
@@ -0,0 +1,206 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-standard--fskgfca",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 32,
+   "sim_steps": 32,
+   "physics_steps": 160,
+   "steps": 674,
+   "elapsed_s": 1002.6,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 2.3425,
+   "worst_yaw_err": 1.6899,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-standard--fskgfca/replay.mp4",
+  "duration_s": 0.68,
+  "fps": 25.0,
+  "frames": 17,
+  "sha256": "6772dbe178b9a2b1c53f88778cda320e9a8a961d3845e873c5ae504143260fc6"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.28,
+      "log_step": 77,
+      "text_zh": "观察：log step 77 的服务输出为 FINAL_EPISODE 11、resets_left 0；随后 agent 启动 gen.py。视频 0.00–0.28 s 中 Go1 足端接触方格地面，腿部开始摆动，因此本次确有行进尝试；此前重置掉的回合不计入节点进度。"
+     },
+     {
+      "t": 0.64,
+      "log_step": null,
+      "text_zh": "观察：已逐帧检查视频 0.00–0.64 s 的全部17帧；视频 0.44–0.56 s 躯干逐渐倾斜，视频 0.60–0.64 s Go1 向侧方翻倒，结束时未恢复支撑。run.json 评分器字段 sim_steps=32、fell=1、live_matches_replay=1；完整1000控制步时间表未执行完，节点始终未达成。"
+     },
+     {
+      "t": 0.64,
+      "log_step": null,
+      "text_zh": "观察：run.json 评分器报告 worst_lin_err=2.3425 m/s、worst_yaw_err=1.6899 rad/s，均超过 DAG 的0.2 m/s和0.3 rad/s阈值。这些是评分器给出的指标，未从视频估测速度。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–0.40 s 中 Go1 足端在方格平地上着地承重，部分腿抬起、回落；视频 0.44–0.64 s 支撑姿态逐渐崩溃并侧翻。实际接触为腿和足端对地面的支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 2.3425,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 1.6899,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.44,
+   "log_step": null,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 0.28–0.40 s Go1 已在摆腿，视频 0.44 s 躯干开始明显倾斜，随后在视频 0.60–0.64 s 侧翻，未能恢复。评分器记录 sim_steps=32、fell=1，故失败发生在按指令行走的执行中，未完成完整时间表。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 0.44,
+    "log_step": null,
+    "evidence_zh": "观察：视频 0.44–0.56 s 中 Go1 躯干倾斜加剧、腿部收拢，视频 0.60–0.64 s 躯干滚向侧方且足端不能维持正常支撑；评分器字段 fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": 0.44,
+    "log_step": 77,
+    "evidence_zh": "推断：agent 在 log step 77 声称最终采用四拍爬行、渐增足端偏移和较小抬脚高度，并启动最后一回合控制脚本；视频 0.28–0.64 s 显示摆腿后姿态逐渐失稳并侧翻。会话中的固定 home 关节目标测试在 log step 28 服务输出中执行50控制步仍 done=False，而最终步行控制在评分器记录中32步即翻倒，支持将失败归于步行控制未能维持支撑和躯干稳定；不足以定位某个控制子类。",
+    "source_zh": "日志 / 视频 / 评分器字段"
+   }
+  ],
+  "gap_zh": "没有逐控制步状态日志；最终回合的详细传感器序列未出现在日志输出中，多次控制器编辑仅记录为 patch));，无法复核最终足端轨迹、反馈律或参数。因此可确认翻倒和总体控制失效，但不能细分接触设计、轨迹/逆解或反馈控制；误差数值采用评分器报告。"
+ },
+ "summary_zh": "最后评分回合的 Go1 已用足端支撑并开始摆腿，但仅执行32/1000控制步，“按指令行走”未达成，最终进度0/1。视频约0.44 s 起失衡，视频0.60–0.64 s 侧翻，评分器也记录 fell=1；平面速度和偏航误差指标均超限，根因归于控制未能维持稳定。缺少最终控制器完整实现和逐步状态，无法进一步确认具体控制子类。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标签与视频、日志一致：评分回合仅32控制步（视频0.64 s，17帧）即侧翻，节点未达成，断点、平衡现象和控制根因均有证据支持，时钟标注正确。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:23:35Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--fskgfca/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--fskgfca/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "7f9c7f45ea3e216ea02707b9d3fde5c72dba16dbe67dd49f57a4e3086d4f2998",
+   "video_sha256": "6772dbe178b9a2b1c53f88778cda320e9a8a961d3845e873c5ae504143260fc6",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--gkeihss.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--gkeihss.json
new file mode 100644
index 0000000000000000000000000000000000000000..b4b4602127cc88b05880670ea5a3648f2ed1b04d
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--gkeihss.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-standard--gkeihss",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 179,
+   "sim_steps": 179,
+   "physics_steps": 895,
+   "steps": 1358,
+   "elapsed_s": 996.0,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 1.59,
+   "worst_yaw_err": 0.7502,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.307,
+   "peak_processes": 8,
+   "peak_disk_gb": 0.275,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-standard--gkeihss/replay.mp4",
+  "duration_s": 3.6,
+  "fps": 25.0,
+  "frames": 90,
+  "sha256": "24542e034d7fc309b74af57e970db04d75e35e824bc45911186f8bdaccb50b77"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 49,
+      "text_zh": "观察：视频 0.00–2.40 s 中Go1主要保持屈腿站姿，地面格线仅有少量相对位移；视频 2.60–3.56 s 中腿部调整后躯干侧倾并翻倒，没有完成整段行走。log step 49 的最终控制代码先保持初始关节目标，再进行四足扫动和步态控制；初始站立即不满足本节点的完整时间表判据。"
+     },
+     {
+      "t": 3.56,
+      "log_step": 50,
+      "text_zh": "观察：视频 3.40–3.56 s 中Go1连续侧翻，末帧躯干落地、腿朝侧上方；log step 50 的服务返回明确为最后的 episode=11、control_steps=179、done=True、reason=fell、resets_left=0。评分器记录 sim_steps=179、fell=1、live_matches_replay=1，且 worst_lin_err=1.59 m/s、worst_yaw_err=0.7502 rad/s，均超过DAG阈值。"
+     },
+     {
+      "t": 3.56,
+      "log_step": 50,
+      "text_zh": "推断：log step 50 的最终回合控制步数179乘以0.02 s为仿真 3.58 s，与视频 3.56 s 的末帧翻倒和视频 3.60 s 的容器时长相符；该回合只执行179/1000步，完整20秒时间表从未达成，因此本节点全程不成立。此前被重置的回合不计入本次进度。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–3.00 s 中Go1足端接触方格平地并承重，腿部屈伸、移脚；视频 3.20–3.56 s 中足端支撑失效，躯干侧面最终接触地面。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 1.59
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.7502
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 3.2,
+   "log_step": 50,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 2.60–3.00 s 中Go1开始较明显地调整足位，视频 3.20 s 起躯干持续侧倾，视频 3.40–3.56 s 侧翻落地，未恢复。log step 50 的最终回合服务状态以 reason=fell 在179个控制步后终止；故失败发生在“按指令行走”的执行中。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 3.2,
+    "log_step": 50,
+    "evidence_zh": "观察：视频 3.00、3.20、3.40、3.48、3.56 s 连续画面中，Go1从足端支撑的低姿态逐渐向一侧滚转，躯干最终侧翻着地、腿朝侧上方。log step 50 服务终止原因为 fell，评分器 fell=1；这是失去支撑和翻倒。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：log step 49 的最终脚本通过关节位置目标实现先保持、再四足扫动与步态控制，并依据倾斜阈值切回初始关节目标；但视频 2.60–3.56 s 中移动足位后侧倾持续扩大，未被控制器抑制，log step 50 记录该回合翻倒。log step 51 对该最终回合服务本体感觉数据的打印中，足端接触由1111变为0101，横向IMU速度读数从约+0.29增至+1.30 m/s，重力方向横向分量从约+0.19增至+0.99，支持运动过程中的控制失稳。log step 53 的四腿正运动学对照误差均约0.00025 m，不能据此把翻倒归为已经证实的运动学模型错误；具体控制子类未能隔离。",
+    "source_zh": "日志 / 视频"
+   }
+  ],
+  "gap_zh": "没有逐控制步状态日志；精确段均值误差使用评分器字段，无法从画面独立重算。log step 51 的本体感觉打印没有每行对应的控制步编号，故不把这些行号换算为视频时刻。agent 在 log step 58 自述怀疑摆腿/支撑切换或摆腿外展，但没有隔离实验，无法确定具体失稳机制或不当参数。"
+ },
+ "summary_zh": "Go1在最终第11回合进行了站姿保持和短暂移脚，但在视频约 3.20 s 开始持续侧倾、至视频 3.56 s 已侧翻，唯一节点“按指令行走”未达成，最终进度0/1。服务记录179/1000个控制步后因翻倒终止，两个原生速度误差指标也均超限；视频与最终回合日志支持控制失稳，缺少逐步状态及隔离实验以进一步定位具体机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "视频仅3.6 s、末段侧翻，与179控制步/fell=1一致；节点未达成、断点、平衡现象与控制根因均有证据支持，未发现需改动之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:05:47Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--gkeihss/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--gkeihss/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "911001594f895a59ec492d3d62746093b7c6ff2ca778cfe67e835413de4e088b",
+   "video_sha256": "24542e034d7fc309b74af57e970db04d75e35e824bc45911186f8bdaccb50b77",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--mtqanz6.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--mtqanz6.json
new file mode 100644
index 0000000000000000000000000000000000000000..c3010a6bd74ceeb276749c277316ec0500eed907
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--mtqanz6.json
@@ -0,0 +1,205 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-standard--mtqanz6",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 123,
+   "sim_steps": 123,
+   "physics_steps": 615,
+   "steps": 123,
+   "elapsed_s": 774.1,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 1.59,
+   "worst_yaw_err": 0.9037,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 0,
+   "reset_budget": 10,
+   "episodes": 1,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-standard--mtqanz6/replay.mp4",
+  "duration_s": 2.48,
+  "fps": 25.0,
+  "frames": 62,
+  "sha256": "18c97feb0a44e6871cfdfd6c3b41c6adfb5e7eeb53183b40f2d32f530463c17d"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 1.12,
+      "log_step": 21,
+      "text_zh": "观察：视频 0.00–1.08 s，Go1主要保持屈腿站姿；视频 1.12–2.16 s，机器人足端交替抬落并在方格平地上移动、转向。log step 13、20、21显示agent脚本在同一实时回合先提交home_pose，累计到control_steps=56；log step 29随后从该计数开始提交行走动作。"
+     },
+     {
+      "t": 2.44,
+      "log_step": 29,
+      "text_zh": "观察：视频 2.24–2.44 s，Go1躯干向一侧持续翻转，末帧腿部伸向侧方，已失去正常足端支撑。log step 29的实际服务返回control_steps=123、done=True；评分器字段fell=1、n_actions=123、worst_lin_err=1.59 m/s、worst_yaw_err=0.9037 rad/s，未完成1000步且两项误差超限，唯一节点全程未达到。"
+     },
+     {
+      "t": 2.4,
+      "log_step": 29,
+      "text_zh": "推断：本次只有一个实时回合且无重置或恢复；log step 29的control_steps按50 Hz映射，起步56/50=视频约1.12 s、末段120/50=视频约2.40 s，分别与所见迈步开始、明显侧翻核对一致。此映射使用控制步计数，未使用agent工作时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 1.12–2.24 s，Go1四条腿的足端交替接触平坦地面，承重并推进、转向；视频 2.24–2.44 s躯干侧翻，正常足端支撑随之丧失。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 1.59,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.9037,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 2.36,
+   "log_step": 29,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 2.24–2.44 s，正在迈步转向的Go1躯干连续侧倾、翻转，未恢复站立；log step 29实际服务在control_steps=123返回done=True，评分器记录fell=1、仅执行123步。因此失败发生在“按指令行走”的操作中。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 2.36,
+    "log_step": 29,
+    "evidence_zh": "观察：视频 2.24 s机器人躯干开始明显向侧面翻转，视频 2.36–2.44 s足端无法维持躯干直立，末帧呈侧翻姿态；log step 29的重力方向读数gz由control_steps=112时的-0.878变为control_steps=120时的-0.372，最后control_steps=123时为+0.030且done=True，评分器fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 29,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 29中agent脚本每8个控制步用一组平面速度及偏航误差生成足端轨迹，再通过解析逆解提交关节位置；脚本读取的重力方向只用于打印，未参与姿态恢复。该实际执行中偏航角速度跟踪在control_steps=112、120持续明显偏离指令，随后重力方向翻转并结束；结合视频 2.24–2.44 s的失衡，支持行走控制未维持稳定。具体是反馈更新、足端轨迹还是模型参数的问题尚不能区分，归于控制总类。"
+   }
+  ],
+  "gap_zh": "没有逐控制步状态日志，无法独立复算各指令段IMU平均误差或精确定位首次翻倒采样；误差采用评分器字段。log step 30之后是agent另建本地MuJoCo模型的调试，未重置或重跑受评回合，不能计为本次进展；现有证据不足以把控制根因细分到某个增益或接触参数。"
+ },
+ "summary_zh": "Go1在本次受评回合先维持站姿，随后迈步转向，但在视频约2.36 s失去平衡并侧翻；只执行123/1000个控制步，唯一“按指令行走”节点未达到，最终0/1。实际行走控制未维持躯干稳定，评分器同时报告平面速度及偏航误差超限；无逐步状态日志，具体控制失稳机制及各段误差不能独立复核，后续本地调试不属于受评轨迹。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "唯一节点未达到、控制步映射(123步/50Hz≈2.46 s,视频2.48 s)与可见侧翻一致,断点、平衡现象与控制根因均有日志(service control_steps、gz读数)和评分器字段支持,未发现需改动之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:16:43Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--mtqanz6/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--mtqanz6/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "98fe7c106a62081fcc374f7a02fe0a675d3115759392199acff088e132d827c4",
+   "video_sha256": "18c97feb0a44e6871cfdfd6c3b41c6adfb5e7eeb53183b40f2d32f530463c17d",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--v9ckhkj.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--v9ckhkj.json
new file mode 100644
index 0000000000000000000000000000000000000000..a08007494ecd20170f468128e0642f834781f7d0
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--v9ckhkj.json
@@ -0,0 +1,211 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-standard--v9ckhkj",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 151,
+   "sim_steps": 151,
+   "physics_steps": 755,
+   "steps": 1395,
+   "elapsed_s": 1076.6,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 1.6452,
+   "worst_yaw_err": 1.1202,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-standard--v9ckhkj/replay.mp4",
+  "duration_s": 3.04,
+  "fps": 25.0,
+  "frames": 76,
+  "sha256": "518a039356f4dcc249922930a18119a0e6610a2bf771a57b9f89cf796719b9cd"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.64,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00–2.00 s 的跟随与侧面画面中，Go1从站姿开始抬落四足并移动，躯干反复起伏和倾斜；已实际尝试行走，但尚未完成整段指令时间表。"
+     },
+     {
+      "t": 2.8,
+      "log_step": 68,
+      "text_zh": "观察：视频 2.60–3.00 s 中Go1躯干越来越陡、足端无法维持正常站立支撑；log step 68 中服务状态报告 episode=11、control_steps=151、done=true、reason=fell。最终评分器同样记录 sim_steps=151、fell=1、live_matches_replay=1，未完成所需1000控制步。"
+     },
+     {
+      "t": 3.0,
+      "log_step": 82,
+      "text_zh": "观察：log step 82 中服务仍报告第11回合已因fell结束并拒绝继续提交；最终评分器字段 worst_lin_err=1.6452 m/s、worst_yaw_err=1.1202 rad/s，分别超过0.2 m/s和0.3 rad/s阈值。"
+     },
+     {
+      "t": 3.0,
+      "log_step": 68,
+      "text_zh": "推断：log step 68 为最终第11回合，log step 11 给出50 Hz控制率，151×0.02=3.02 s，与视频 3.00 s 的近竖直失衡末帧及3.04 s容器时长相符；仅用该回合control_steps核对录像，未将agent工作时间或1395个跨回合累计步数换算为视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–2.60 s 中Go1四条腿的足端先后接触方格平地，支撑躯干并伴随机体移动；视频末段足端支撑未能维持躯干姿态。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 1.6452
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 1.1202
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 2.8,
+   "log_step": 68,
+   "where": "操作中",
+   "evidence_zh": "观察：Go1在视频 2.60–3.00 s 行走过程中持续倾倒，躯干从明显倾斜变至近竖直，未恢复；log step 68 的服务状态确认最终第11回合在第151/1000控制步因fell终止，正在执行的“按指令行走”步骤未完成。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 2.8,
+    "log_step": 68,
+    "evidence_zh": "观察：视频 2.60、2.80、3.00 s 中Go1躯干倾斜逐步增大、前后足高度关系显著改变，末帧已无法保持四足站立姿态；log step 68 服务报告reason=fell，最终评分器记录fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "t": 2.8,
+    "log_step": 72,
+    "source_zh": "日志",
+    "evidence_zh": "推断：agent在log step 66写入基于IMU倾角的降速与冻结步态反馈，log step 68实际运行该控制器；同回合log step 72记录第140、145、150控制步倾角依次为0.87、1.24、1.55 rad，均已safe=1、scale=0.15，倾角仍继续增大，随后服务在第151控制步报告fell。结合视频 2.80–3.00 s 的持续倾倒，说明该姿态反馈未能抑制失稳；不能由此指定某一个增益或横滚符号错误。"
+   }
+  ],
+  "gap_zh": "没有独立的逐帧状态日志，无法复算各指令段IMU误差均值或定位具体失稳参数；原生误差采用最终评分器字段。log step 72是agent脚本打印的本回合传感反馈，支持控制归因，但不等同于独立状态日志；更早测试回合不计入节点进度。",
+  "unmet": {
+   "command_schedule_completed": false
+  }
+ },
+ "summary_zh": "Go1已以足端交替接地尝试行走，但“按指令行走”节点始终未达成，最终为0/1。视频 2.60–3.00 s 显示躯干持续倾倒，最终第11回合在第151/1000控制步因翻倒终止，评分器两项速度误差也均超限；日志支持姿态反馈未能抑制失稳。没有独立逐帧状态日志，无法进一步确认具体失稳参数或复算各段误差。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成、断点、平衡现象与反馈控制根因均与视频及log一致；时钟仅用第11回合control_steps（151×0.02=3.02 s）并说明核对，无换算问题。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:09:52Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--v9ckhkj/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--v9ckhkj/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "ef8f039f8d05819b0a9fc7bd1a4912459cb831bf2ec82358cfbf66db1cdb911e",
+   "video_sha256": "518a039356f4dcc249922930a18119a0e6610a2bf771a57b9f89cf796719b9cd",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--yualwlo.json b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--yualwlo.json
new file mode 100644
index 0000000000000000000000000000000000000000..a3af3a5a359631ce6de2be0bdbcf29398c3bd666
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_flat_terrain/trajectories/mujoco_playground_go1_joystick_flat_terrain-standard--yualwlo.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_flat_terrain-standard--yualwlo",
+ "taskKey": "mujoco-playground/go1_joystick_flat_terrain",
+ "dag_sha256": "869ae70c21c400f0d0532c63bf5be4417cd9459d397a35e1ff6684ee2b3521cc",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "graded_by_live": 1,
+   "replays": 0,
+   "finalized": 1,
+   "missing_log": 0,
+   "sim_error": 0,
+   "grader_error": 0,
+   "infra_error": 0,
+   "n_actions": 43,
+   "sim_steps": 43,
+   "physics_steps": 215,
+   "steps": 559,
+   "elapsed_s": 1763.7,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 2.2745,
+   "worst_yaw_err": 0.7495,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "instance_mismatch": 0,
+   "no_verdict": 0,
+   "gave_up": 0,
+   "sim_crashed_on_call": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.292,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.25,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_flat_terrain-standard--yualwlo/replay.mp4",
+  "duration_s": 0.88,
+  "fps": 25.0,
+  "frames": 22,
+  "sha256": "a26facf53c4a3b286b926788e074895dc95f26804945e552aabcb9e5020cfd84"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：已逐帧检查本次视频 0.00–0.84 s 的全部22帧。Go1先以足端着地、腿部屈伸支撑身体，视频 0.60–0.84 s 躯干持续后仰、前足抬离，末帧机身接近竖直并继续翻转；没有完成持续行走。"
+     },
+     {
+      "t": 0.84,
+      "log_step": 31,
+      "text_zh": "观察：log step 31 是最后一次重置及 final5 回合，输出 resets_left=0；此后 log step 32 只有最终报告。评分器 reward 报告 episodes=11、resets=10、graded_by_live=1、finalized=1、sim_steps=43、fell=1；该短回合与本次视频的后仰失衡及结束相符。"
+     },
+     {
+      "t": 0.84,
+      "log_step": null,
+      "text_zh": "推断：结合本次视频的翻倒过程及评分器记录的43个控制步，要求完成1000控制步且不翻倒的“按指令行走”从未达成，故区间为空、最终为假；这并非仅由 success=0 推得。评分器 worst_lin_err=2.2745 m/s、worst_yaw_err=0.7495 rad/s 也分别超过0.2 m/s、0.3 rad/s阈值。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–0.52 s，Go1的足端先后接触方格平地承重，腿部屈伸改变机身高度与支撑；视频 0.60–0.84 s，前足抬起、躯干后仰，支撑未能维持身体平衡。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 2.2745,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.7495,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.64,
+   "log_step": 31,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 0.60–0.84 s，正在执行启动支撑和腿部运动的Go1前足抬离、躯干越来越后仰，至末帧接近竖直并继续翻转，未恢复；评分器记录 fell=1、sim_steps=43。log step 31 的最后一次重置运行是本次评分回合，断点取“按指令行走”，视频 0.64 s 为明显失衡的导航锚点。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 0.64,
+    "log_step": 31,
+    "evidence_zh": "观察：视频 0.64、0.76、0.84 s，Go1前足悬空、躯干由倾斜转为近竖直并后翻；最后一帧仍未恢复足端稳定支撑。评分器 fell=1 确认该回合发生翻倒。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：log step 25 的脚本用足端目标、解析逆解、支撑/摆动切换及足高修正控制Go1；log step 30 将启动稳定阶段延长为仿真前1.00 s，log step 31 又在该阶段关闭姿态修正门控，并执行最后回合。该控制程序运行后的本次视频 0.60–0.84 s 仍出现持续后仰翻转，评分器记录 fell=1，说明启动阶段的腿部运动和支撑调节未能维持平衡，归入控制；现有证据不足以细分为某一反馈律、接触布局或增益错误。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "观察：inputs/有本次及参考视频和完整会话记录，但无逐帧模拟器状态日志。本地脚本在 log step 31 输出 steps=44、fell=False、首段误差2.190和1.218，与评分器 sim_steps=43、fell=1、worst_lin_err=2.2745、worst_yaw_err=0.7495 不同；log step 21 显示脚本把采样帧重构到自建步数网格，因此这些本地统计不替代评分器字段，也不用于映射视频时间。视频能确认失衡过程，不能独立复核精确IMU翻倒采样、速度误差或具体控制失稳机制；log step 32 对支撑面收窄、参数及任务不可行的解释属于agent自述，未据此添加细分根因或任务设计问题。",
+  "unmet": {
+   "command_schedule_completed": false
+  }
+ },
+ "summary_zh": "本次评分为最后一次重置后的短回合：Go1以足端支撑并屈伸腿部，随后在视频约0.60–0.84 s持续后仰翻倒，“按指令行走”未达成，最终0/1节点。断点在该步骤操作中，现象为全身平衡丧失，结合脚本及执行结果将根因保守归为控制；评分器记录仅43控制步、fell=1，且两项速度误差均超限。无逐帧状态日志，本地脚本的计数和判定与评分器不一致，精确翻倒采样和控制子原因无法独立复核。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "视频全部帧显示Go1先站立屈伸、随后后仰翻倒，节点未达成、断点、平衡现象与控制根因均有证据支持，未发现需改动之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:24:39Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--yualwlo/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_flat_terrain-standard--yualwlo/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "b841088ee24929fa3a97354e8a5963aacbd35abd370f63d0aeb09ec63fc8d139",
+   "video_sha256": "a26facf53c4a3b286b926788e074895dc95f26804945e552aabcb9e5020cfd84",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--arwtd3f.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--arwtd3f.json
new file mode 100644
index 0000000000000000000000000000000000000000..54da0cfc023fbed46f05081e1f9cbf7fe684543a
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--arwtd3f.json
@@ -0,0 +1,192 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-privileged--arwtd3f",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 1.3528,
+   "worst_yaw_err": 1.8017,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 45,
+   "agent_resets": 2757,
+   "agent_sim_steps": 75450,
+   "agent_replays": 2,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 1.883,
+   "peak_processes": 16,
+   "peak_disk_gb": 0.505,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--arwtd3f/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "6e10fb96f7063efcc72f07f2615cf22ef5f8c6a052f2835f49e572e3bb984988"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 1.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0–2.7 s，Go1 四足反复抬落并接触棕色起伏地面，躯干朝向来回摆动；相邻地面纹理显示机身推进很少。视频 2.9–13.4 s 仍以小幅踏步和反复摆转为主，未形成连续有效的指令行进。"
+     },
+     {
+      "t": null,
+      "log_step": 35,
+      "text_zh": "观察：log step 35，agent 的生成脚本将所选参数生成的关节目标保存为提交轨迹 trajectory.npz，返回 success=false、fell_at=null；五个非空评分段的平移误差依次为 1.1028、1.3528、1.0107、0.6104、0.1950 m/s，偏航误差依次为 1.0873、1.8017、1.7707、1.5234、0.5365 rad/s。log step 40 对该提交文件的新进程重放再次返回 steps=1000、horizon=1000、tracking_ok=false，前述可见误差数值一致。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：视频 13.7–20.286 s，Go1 机身逐渐降低，末秒仅上部机身清楚可见；评分器 reward 记录 n_actions=horizon=1000、fell=0、deterministic=1、worst_lin_err=1.3528 m/s、worst_yaw_err=1.8017 rad/s。推断：完整回合虽已执行且没有评分器翻倒记录，两个误差指标分别超过 0.2 m/s 和 0.3 rad/s，故“按指令行进”的整回合验收条件从未成立，终值为假。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–13.4 s，Go1 的四条腿反复抬落，可见足端落在粗糙地面并支承机身，机器人据此踏步和摆转；末段足部被地形遮挡，不据此另判接触方式。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 1.3528
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 1.8017
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 1.0,
+   "log_step": 40,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 0.9–2.7 s，Go1 已在踏步，机身反复摆转而地面相对位移很小；视频 2.9–13.4 s 仍未形成有效的持续行进。log step 35 保存的提交轨迹在 log step 40 完整重放 1000 控制步，仍为 tracking_ok=false；评分器平移与偏航误差均超标。推断：失败发生在“按指令行进”的执行中，视频 1.0 s 是早期运动缺陷的导航锚点，并非精确超阈值时刻。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 1.0,
+    "log_step": 40,
+    "evidence_zh": "观察：视频 0.9–2.7 s 及视频 4.3–10.9 s，Go1 四足持续踏动，但机身主要反复左右摆转，围绕相近地面纹理活动，平移推进很少；视频 13.7–20.286 s 机身进一步降低。提交文件重放在 log step 40 报告 tracking_ok=false，评分器未记录翻倒。推断：这是行进速度和方向跟踪不足的步态失败。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志",
+    "evidence_zh": "推断：agent 编写的足端轨迹与关节目标控制器未能产生足够的机身平移和稳定的偏航响应。log step 7 的脚本按指令和传感器速度生成对角小跑足端轨迹，再经解析逆解记录关节目标；log step 18 加入速度和转向增益并搜索参数。log step 35 用所选参数保存的实际提交文件、log step 40 对它的新进程重放及评分器 reward 均给出相同的最差平移误差 1.3528 m/s 和偏航误差 1.8017 rad/s，与视频 0.9–13.4 s 的小幅踏步及摆转一致。现有证据支持控制能力不足，但不能证实某个增益、逆解或足地接触设计是单独根因。"
+   }
+  ],
+  "gap_zh": "没有状态日志或提交回合的逐步速度、姿态和接触力记录；相机随动，精确误差依赖评分器及已核对为提交文件的重放结果。视频 13.7–20.286 s 机身降低、腿部被地形遮挡，无法确定足端支承或地形相交的具体机制；不据此认定仿真故障，也无法将控制根因细分。断点时间仅为可见缺陷的粗略锚点。"
+ },
+ "summary_zh": "Go1 执行了完整 1000 控制步，评分器未记录翻倒，但平移和偏航误差分别为 1.3528 m/s、1.8017 rad/s，均超阈值，唯一节点“按指令行进”未达到（终态 0/1）。视频显示机器人反复小幅踏步和摆转、有效推进不足，断在该步操作中，根因归为控制。缺少逐步状态与接触力日志，末段机身降低且腿部遮挡，不能进一步确定控制子类或足地作用机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "唯一节点未达成（评分器 worst_lin_err 1.3528、worst_yaw_err 1.8017 均超阈值，完整 1000 步、fell=0），视频抽帧显示机身反复摆转、末段机身降低，与标注一致；断点、现象（步态）和根因（控制）有证据且表述克制，未发现需修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:07:27Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--arwtd3f/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--arwtd3f/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "83863475721c56e980be52913781b96bc70dff8852d5e253ded8ecc45f34ffca",
+   "video_sha256": "6e10fb96f7063efcc72f07f2615cf22ef5f8c6a052f2835f49e572e3bb984988",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--dhzezf3.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--dhzezf3.json
new file mode 100644
index 0000000000000000000000000000000000000000..33ccb3686be9d3462aedb3fd7d154b31d55fd1f8
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--dhzezf3.json
@@ -0,0 +1,188 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-privileged--dhzezf3",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 1.2858,
+   "worst_yaw_err": 0.7827,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 71,
+   "agent_resets": 75,
+   "agent_sim_steps": 56090,
+   "agent_replays": 4,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.691,
+   "peak_processes": 8,
+   "peak_disk_gb": 0.255,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--dhzezf3/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "76b27ef0d5266bf11d5f3d2db6114c0d9417198e8b1e698b7feb308144625049"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：本次视频 0.00–2.86 s 的逐帧画面中，Go1 落地后只小幅调整腿部和躯干，四足在机身周围支承，没有启动连续迈步；视频全程逐秒检查至 19.00 s，以及视频 19.00–20.29 s 的逐帧末段，机器人均基本留在同一处地形上，未出现指令要求的持续平移和转向。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 155,
+      "text_zh": "观察：log step 153 执行最终生成脚本，log step 154 重放该交付文件；log step 155 的基准重放检查报告 steps=1000、horizon=1000、fell_at=null，但 success=false、tracking_ok=false，worst_lin_err=1.2858 m/s、worst_yaw_err=0.7827 rad/s，与 run.json 的评分器字段相符，分别超过 0.2 m/s、0.3 rad/s 的阈值。视频 20.29 s 机器人仍站立，因此未完成的是指令跟踪验收，不能将跑满步数和未翻倒计为节点达成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–2.86 s 及视频 19.00–20.29 s 的双视图可见 Go1 足端落在粗糙棕色地面上承重，腿部调整使躯干维持站姿；本次交付基本原地支撑，未见持续交替迈步行进。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 1.2858,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.7827,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.0,
+   "log_step": 153,
+   "where": "首步前",
+   "evidence_zh": "观察：视频 0.00–2.86 s 机器人只落地稳姿，此后逐秒画面至视频 19.00 s 及逐帧末段至视频 20.29 s 都没有连续迈步行进。log step 152 展示的最终脚本固定各足端的水平目标，仅根据重力和角速度调节足端高度；log step 153 执行该脚本保存交付轨迹，行进控制从未在该交付回合启动。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "没尝试",
+    "item": "没尝试",
+    "t": 0.0,
+    "log_step": 153,
+    "evidence_zh": "观察：视频 0.00–2.86 s 只有落地和站姿调整，视频逐秒画面及视频 19.00–20.29 s 的逐帧末段中机器人仍基本原地站立；最终生成脚本（log step 152–153）没有读取速度指令或生成水平摆腿轨迹。本次交付为站立基线，没有启动按指令行进；会话中其他测试的步行和摔倒不属于本次视频。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 138,
+    "evidence_zh": "推断：agent 未能形成可用的行进控制，最终回退到仅稳姿的交付方案。会话测试中，log step 90 测得首段平均前进速度约 0.341 m/s（目标 1.137 m/s），平均偏航角速度约 -1.263 rad/s（目标约 -0.703 rad/s）；log step 135 的步态测试在控制步 129 翻倒，log step 138 的完整步态测试在控制步 174 翻倒。随后最终生成脚本仍只固定水平足位、调节腿高（log step 152–153），且其重放明确跟踪失败（log step 155）。这些会话记录支持行进控制能力不足，但不足以归到某个控制子类。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "本次没有状态日志，视频也没有速度或指令叠字，无法独立重建每个控制步的速度误差，精确误差采用最终交付文件的重放检查和评分器字段。控制根因由多次会话步态测试失败与最终站立基线的选择推断，具体动力学或控制子类不能确定；此前测试的翻倒未计入本次轨迹。"
+ },
+ "summary_zh": "Go1 跑满了交付轨迹的 1000 控制步且没有翻倒，但本次视频中只在原地调整站姿，唯一节点“按指令行进”始终未达成（最终 0/1）。断点为该步首步前：交付的站立基线没有启动行进，平移和偏航最差段误差分别为 1.2858 m/s、0.7827 rad/s；会话中多次步态控制失败支持其回退到站立方案的控制根因。没有状态日志，精确误差依赖最终重放检查及评分器，不能进一步确定控制子类。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "交付的 gen.py 只做站姿平衡（固定足端水平位置，无速度指令），视频 1–20 s 机器人原地未动，节点 0/1、断点首步前/没尝试、根因控制（推断）均有视频与日志支撑；未发现需改动的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:06:36Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--dhzezf3/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--dhzezf3/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "40b28efb585d2206981f7599442293b95270b31f2cadb8d701b63accc48155de",
+   "video_sha256": "76b27ef0d5266bf11d5f3d2db6114c0d9417198e8b1e698b7feb308144625049",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--egdgct6.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--egdgct6.json
new file mode 100644
index 0000000000000000000000000000000000000000..9202ca5238f31de6d12fd6527077f39caf9c2fab
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--egdgct6.json
@@ -0,0 +1,173 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-privileged--egdgct6",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.1137,
+   "worst_yaw_err": 0.1985,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 400,
+   "agent_resets": 1417,
+   "agent_sim_steps": 58030,
+   "agent_replays": 4,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 18.546,
+   "peak_processes": 19,
+   "peak_disk_gb": 0.834,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--egdgct6/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "6c1ad8da7b435ef61c2656184845636dd988285764aa65b7076bcf5e88ca1486"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [
+     [
+      20.286,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 2.714,
+      "log_step": null,
+      "text_zh": "观察：视频 0.000–2.714 s 的并排跟随与侧视画面中，灰色 Go1 在棕色起伏地面上交替抬足、落足，机身朝向逐渐改变；视频 3.000–10.714 s 继续行进，视频 11.000–13.714 s 再次转向并接近地形边缘，躯干始终保持向上。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：视频 14.143–20.286 s 机器人相对地形的位移显著减小，四腿仍交替运动；逐帧检查视频 19.286–20.286 s，直到末帧躯干均由足端支撑，未见翻倒。"
+     },
+     {
+      "t": null,
+      "log_step": 44,
+      "text_zh": "观察：log step 44 中，agent 运行最终生成脚本所得检查结果为 steps=1000、horizon=1000、tracking_ok=true、fell_at=null、success=true，重新生成的 actions 为 float64 (1000,12)，与保存版本逐元素相同；log step 46 的交付核验输出确认两次新进程回放结果一致且成功。inputs/run.json 的评分器 reward 同时给出 deterministic=1、replay_success=1、n_actions=1000、fell=0、worst_lin_err=0.1137 m/s、worst_yaw_err=0.1985 rad/s。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "推断：评分器确认完整指令表及两项误差阈值均通过，视频中的连续行进、末段减速与直立终态相符，故唯一节点完成。此节点要求整回合完成，起始站立和中途行进均不算达成；以检查过的末帧视频 20.286 s 作为完成时刻的近似锚点，精度约半秒，未将 agent 工作时钟或仿真时钟直接换算成视频时钟。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 2.429–3.286 s、10.714–12.429 s 及 19.286–20.286 s 可见 Go1 的足端交替落在棕色粗糙地面上，支撑躯干并随步行与转向交替抬起、着地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.1137
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.1985
+  }
+ ],
+ "failure": null,
+ "summary_zh": "交付回放完成唯一节点“按指令行进”（终态 1/1）：Go1 在粗糙地面上连续行进与转向，末段减速，视频 20.286 s 的末帧仍保持直立；节点按整回合验收，在回放末尾计为完成。评分器确认完成 1000 控制步且无翻倒，最差分段平均平移误差 0.1137 m/s、偏航误差 0.1985 rad/s，均低于阈值；没有状态日志，精确误差依据评分器字段，视频完成时刻仅为末帧附近的近似定位。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "唯一节点完成、终态1/1、无失败层均与评分器字段（success=1、fell=0、误差低于阈值）及视频末帧（躯干直立、足端支撑）相符；未发现需修改之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:07:35Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--egdgct6/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--egdgct6/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "b724c83ecb70a5043bb3a27b1c2434766a84a83c8242627cd94f11dcdfed7e0f",
+   "video_sha256": "6c1ad8da7b435ef61c2656184845636dd988285764aa65b7076bcf5e88ca1486",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--j9fyoaj.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--j9fyoaj.json
new file mode 100644
index 0000000000000000000000000000000000000000..caaf75031bd7a33a1f6c5ccc8aff4907eb490296
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--j9fyoaj.json
@@ -0,0 +1,198 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-privileged--j9fyoaj",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 1,
+   "fell_at_step": 95.0,
+   "worst_lin_err": 1.541,
+   "worst_yaw_err": 1.4355,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 229,
+   "agent_resets": 234,
+   "agent_sim_steps": 74015,
+   "agent_replays": 3,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 1.209,
+   "peak_processes": 10,
+   "peak_disk_gb": 0.243,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--j9fyoaj/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "f4b88b07038dfb52a9b71d9bf800bd736f7198d6f2644cdf39a243537246d9fd"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.43,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00–1.43 s 的并排跟随与侧视画面中，Go1 已摆腿迈步，足端接触粗糙棕色地面，确实开始了行进尝试；初始站立不等于完成整段指令表。"
+     },
+     {
+      "t": 1.85,
+      "log_step": 94,
+      "text_zh": "观察：视频 1.57–2.00 s，Go1 躯干向侧面翻转，足端抬起，失去正常站立支撑；评分 reward 的 fell=1、fell_at_step=95，log step 94 对最终保存文件的重放也返回 fell_at=95、success=false。"
+     },
+     {
+      "t": 20.29,
+      "log_step": 94,
+      "text_zh": "观察：视频 2.14–20.29 s 抽查及视频 19.29–20.29 s 逐帧检查中，两个视图主要只见地面，未再见机器人正常行走。评分器虽执行 n_actions=horizon=1000，但 worst_lin_err=1.541 m/s、worst_yaw_err=1.4355 rad/s 均超阈值；log step 94 的最终文件重放字段与评分一致。"
+     },
+     {
+      "t": 20.29,
+      "log_step": 94,
+      "text_zh": "推断：节点要求完成全部指令、全程无翻倒且各段误差达标；本次已发生翻倒且误差超标，故该节点全程未达成，final=false，而非从零分单独反推未完成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.14–1.43 s，Go1 的足端交替着地承重并随摆腿尝试推进；视频 1.57–2.00 s 机身侧翻、足端离开正常支撑位置。未见机械臂或轮式底盘接触。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 1.541,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 1.4355,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 1.85,
+   "log_step": 94,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 1.43 s 机器人仍在迈步，视频 1.57–2.00 s 躯干明显侧翻、足端抬起，行进中的支撑丢失；评分器记录 fell_at_step=95，log step 94 对最终交付的重放同样记录 fell_at=95。断点属于正在尝试的“按指令行进”，此后未完成该节点。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.85,
+    "log_step": 94,
+    "evidence_zh": "观察：视频 1.57 s 起 Go1 躯干侧倾落低，视频 1.71–2.00 s 机身翻向一侧、腿足朝上，丢失正常四足支撑；最终文件重放（log step 94）及评分 reward 均报告翻倒。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 94,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 84、85 的代码以周期相位安排支撑与摆腿、用速度反馈计算落足目标；log step 93 的粗糙地形参数试验仍在控制步 55–95 翻倒，log step 94 选取该控制器生成的 95 步动作并把最后一行补齐至 1000 步，最终文件重放仍在控制步 95 翻倒，与视频 1.57–2.00 s 的失衡相符。这支持行进控制未能稳定机身；不足以证明某个具体增益、逆解或足端接触设计是唯一原因。"
+   }
+  ],
+  "gap_zh": "没有逐步状态、足端接触力或关节跟踪日志，不能进一步区分失衡的控制子类；视频 2.14 s 后机器人不再清晰可见，后续精确运动不能凭画面恢复。log step 96 自述的控制步 87 是较早交付版本，最终版本采用 log step 94 与评分 reward 一致的控制步 95；本文视频锚点直接取自所检查帧，未将 agent 工作时间换算为视频时间。",
+  "unmet": {
+   "command_schedule_completed": false
+  }
+ },
+ "summary_zh": "Go1 已迈步尝试粗糙地形指令跟踪，但视频约 1.7–2.0 s 侧翻，唯一节点“按指令行进”未达成，最终为 0/1。最终交付虽有 1000 步，评分器记录翻倒，且最差平移与偏航误差分别为 1.541 m/s、1.4355 rad/s；最终文件重放和画面支持控制未能维持平衡。缺少逐步状态和接触力证据，无法确定更细的控制原因。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "视频约 1.6–2.0 s 机身侧翻，与 fell_at_step=95（约 1.9 s）一致；单节点未达成、断点、平衡现象和控制根因均有证据支持，时钟与 log step 均无问题。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:10:38Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--j9fyoaj/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--j9fyoaj/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "8d5093470b40e998261c36550dc0dbf76473ac5048665fb55ded655b01ca4dc5",
+   "video_sha256": "f4b88b07038dfb52a9b71d9bf800bd736f7198d6f2644cdf39a243537246d9fd",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--luvpqfp.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--luvpqfp.json
new file mode 100644
index 0000000000000000000000000000000000000000..fc00e95a47c57fcaec8cb5d04c8459658bd12aef
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--luvpqfp.json
@@ -0,0 +1,192 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-privileged--luvpqfp",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 1.235,
+   "worst_yaw_err": 0.6662,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 83,
+   "agent_resets": 1631,
+   "agent_sim_steps": 484835,
+   "agent_replays": 2,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 1.374,
+   "peak_processes": 14,
+   "peak_disk_gb": 0.981,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--luvpqfp/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "91eac5bda64118f3c0816eedf2c2da966c0641682c0bb1b3bd89aefaed322c58"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 1.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.857–2.714 s 的并排跟随与侧视画面中，Go1 足腿反复屈伸、机身小幅起伏并改变朝向，但相对邻近地面纹理的平移很小；视频 3–13 s 仍主要在同一小片地面踏步、摇摆及转向。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 152,
+      "text_zh": "观察：视频 19.286–20.286 s 的连续末尾帧中，Go1 仍由足端支承，躯干未翻倒。log step 152 中 agent 重新生成最终交付文件并调用基准的独立重放检查器，检查器报告 steps=1000、horizon=1000、fell_at=null、tracking_ok=false；段 1/2/3/5 的平移误差分别为 1.235/0.9883/0.7499/0.377 m/s，偏航误差分别为 0.4077/0.5544/0.6662/0.3056 rad/s，段 6 为 0.1703 m/s 和 0.1741 rad/s。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 152,
+      "text_zh": "推断：最终交付文件的重放误差与 inputs/run.json 的 reward.worst_lin_err=1.235、worst_yaw_err=0.6662 完全一致，且 reward.n_actions=1000、fell=0、deterministic=1，与视频中的完整、未翻倒回合相符；本节点要求整个日程各非空评分段均通过，前述超标没有被末段的达标修复，因此从未达到，终值为假。此处未将 agent 会话时间换算为视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.857–3.286 s 可见 Go1 的足端落在起伏棕色地面上，足腿屈伸并支承机身；视频 19.286–20.286 s 仍保持足端承重。实际作用端为四足机器人的足端，对象为粗糙地面。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 1.235,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.6662,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 1.0,
+   "log_step": 152,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 0.857–2.714 s，Go1 已在驱动足腿尝试行进，但机身主要在小范围内起伏、踏步；视频 3–13 s 继续出现行进不足。log step 152 对最终交付轨迹的独立重放检查显示首个非空评分段的平移误差为 1.235 m/s、偏航误差为 0.4077 rad/s，均超阈值，后续多个评分段也超标。推断：失败发生在持续执行“按指令行进”期间，视频 1.0 s 仅为早期行进不足的粗略导航锚点，并非阈值首次越界的精确时刻。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 1.0,
+    "log_step": 152,
+    "evidence_zh": "观察：视频 0.857–2.714 s，Go1 足腿运动而机身相对地形的净平移很小；视频 3–13 s 主要为小范围踏步、摇摆和转向，视频 19.286–20.286 s 仍站立。log step 152 的最终文件重放报告多个平移、偏航评分段超标且 fell_at=null，评分器 reward.fell=0。推断：这是行进速度及方向跟踪不足的步态失败，未见本回合翻倒。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志",
+    "evidence_zh": "推断：agent 在 log step 128 将解析腿部逆解及相位步态控制器封装为生成脚本并录制关节目标；log step 152 重新生成并独立重放最终文件，1000 个控制步均执行、没有翻倒，却在多个指令段保持很大的速度跟踪误差，与视频 0.857–13 s 的小范围踏步及转向相符。控制器未把所要求的机身速度转换为足够的实际行进，故归于控制；现有证据不足以确定为接触设计、逆解、反馈或某个参数的单一问题。"
+   }
+  ],
+  "gap_zh": "没有逐步状态日志或最终交付回合的足端接触力、关节目标与实际值对照；跟随相机也不能独立给出精确机身速度，精确误差采用评分器字段及 log step 152 的最终文件重放结果。视频足以定位早期行进不足，但不能确定误差首次超阈值的控制步，亦不能区分足底滑移、步态相位和关节跟踪的具体贡献；agent 在 log step 153 将不足归咎于较软的 PD 增益，该机制尚未独立证实，未据此细分根因。"
+ },
+ "summary_zh": "Go1 完成了 1000 个控制步并在视频末尾保持站立，但主要在小范围踏步、摇摆和转向，“按指令行进”未达到，最终节点为 0/1。最终文件的独立重放与评分器一致：最差平移误差 1.235 m/s、最差偏航误差 0.6662 rad/s，均超阈值，断点为行进操作中的步态不足，根因归于控制。缺少逐步状态和接触力记录，无法进一步确认具体控制机制。",
+ "conf": 7,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "单节点未达成、断点在行进操作中、步态现象与控制根因均有视频与 log step 152 的重放结果及评分字段支持；视频中机器人原地踏步、地面纹理几乎不移动，与成功示范明显不同，未见需修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:16:30Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--luvpqfp/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--luvpqfp/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "b2299ab0090f7a1ea2ad2057ab33c92ea70c19405d43148196435cd270ad0122",
+   "video_sha256": "91eac5bda64118f3c0816eedf2c2da966c0641682c0bb1b3bd89aefaed322c58",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--xbavacu.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--xbavacu.json
new file mode 100644
index 0000000000000000000000000000000000000000..fd3a2f9502a7693f1a4375a7e623f116df9c39ff
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--xbavacu.json
@@ -0,0 +1,186 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-privileged--xbavacu",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 1.2825,
+   "worst_yaw_err": 0.7788,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 53,
+   "agent_resets": 315,
+   "agent_sim_steps": 174785,
+   "agent_replays": 1,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.717,
+   "peak_processes": 8,
+   "peak_disk_gb": 0.251,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--xbavacu/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "d42a5044e9458ac8a765a5eaba23487e022aeaea46d924a43f4e8424ba7deb45"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 60,
+      "text_zh": "观察：log step 60 中 agent 的交付脚本计算固定站姿 q_stand，并用 np.tile(q_stand, (1000, 1)) 保存整条轨迹；输出确认保存基线成功。视频 0–2.0 s 密集帧中 Go1 仅落足、轻微调整站姿，没有启动交替迈步行进。"
+     },
+     {
+      "t": 10.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0–20.286 s 每秒抽帧及视频 19.0–20.286 s 的逐帧检查中，Go1 始终在同一片地面站立，未见前进、后退、侧移或持续转向，躯干没有翻倒。足端存在原地承重接触，但未启动该行进步骤，故不为未尝试节点列 contact_actual。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：inputs/run.json 的评分器字段 n_actions=horizon=1000、fell=0、deterministic=1；worst_lin_err=1.2825 m/s、worst_yaw_err=0.7788 rad/s，分别超过 0.2 m/s、0.3 rad/s 阈值。推断：回合长度满足要求，但整段指令跟踪验收从未成立，节点终值为 false。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 1.2825
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.7788
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.0,
+   "log_step": 60,
+   "where": "首步前",
+   "evidence_zh": "观察：视频 0–2.0 s 仅见 Go1 原地落足并站稳，后续至视频 20.286 s 均未开始按指令迈步；log step 60 的交付脚本把同一组站姿关节目标重复 1000 次。推断：交付的评分轨迹从首步前就未启动按指令行进，完整播放后仍未修复。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "没尝试",
+    "item": "没尝试",
+    "t": 0.0,
+    "log_step": 60,
+    "evidence_zh": "观察：视频 0–20.286 s 中 Go1 只保持原地站姿，没有交替迈步和随指令的行进、侧移、转向；log step 60 保存固定 q_stand 的 1000 行重复值。推断：评分尝试没有启动行进步骤。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 58,
+    "source_zh": "日志",
+    "evidence_zh": "观察：agent 的会话步态测试在 log step 41 报告 fell_at=171、tracking_ok=false；log step 48 的未翻倒方案仍有平移/偏航误差 0.2927/0.6695，超过任务阈值；log step 58 的未翻倒方案仍有 0.5146/1.5508 的误差，其他参数还导致翻倒；随后 log step 60 只保存固定站姿。推断：会话未形成满足速度跟踪要求的行进控制，最终交付退回站立基线，是控制实现失败；这些测试只用于根因证据，不是评分视频中的运动或摔倒。"
+   }
+  ],
+  "gap_zh": "没有逐步状态或传感日志，评分器仅提供全局最大段均值误差，无法精确定位各段误差或进一步确认某个控制参数、接触设计或反馈环节的责任；根因保留在控制大类。"
+ },
+ "summary_zh": "评分回合执行了全部 1000 控制步且未翻倒，但 Go1 在视频 0–20.286 s 只原地站立，唯一节点“按指令行进”始终未达到，最终为 0/1。交付脚本重复固定站姿，断点在首步前，现象为没尝试；会话多次步态测试仍失稳或跟踪超标，支持控制大类根因。没有逐步状态日志，精确分段误差与控制子类无法进一步确认。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "标注与视频、log 一致：视频 0–20.4 s Go1 只在原地站立，交付脚本（log step 60）保存固定站姿 1000 行，节点未达成、断点首步前、现象没尝试、根因控制（来自会话测试证据）均有依据，未发现需要修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:15:34Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--xbavacu/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--xbavacu/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "81063f3d8d2614ae5c7324d432f95bbe5fdece23cef3412978e0b6ba799e624d",
+   "video_sha256": "d42a5044e9458ac8a765a5eaba23487e022aeaea46d924a43f4e8424ba7deb45",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--xcg2udc.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--xcg2udc.json
new file mode 100644
index 0000000000000000000000000000000000000000..44301a7ebc824c7016526d0a7252494fb6a97072
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-privileged--xcg2udc.json
@@ -0,0 +1,192 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-privileged--xcg2udc",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.9732,
+   "worst_yaw_err": 0.9846,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 66,
+   "agent_resets": 701,
+   "agent_sim_steps": 1736775,
+   "agent_replays": 17,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.9,
+   "peak_processes": 10,
+   "peak_disk_gb": 0.833,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--xcg2udc/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "a837a45defc97012132f03ee52bdfbbe6bad952f5d5d8149e624ad754dd731c5"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 1.429,
+      "log_step": null,
+      "text_zh": "观察：视频 0.000–2.857 s 的双视图中，Go1 从初始姿态降至低蹲，四足小幅交替迈动，机身有移动和转向；视频 3.000–13.429 s 仍低姿态缓慢挪行，视频 13.571–20.286 s 趋于停稳，末秒未见翻倒。已逐帧检查整段视频。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 162,
+      "text_zh": "观察：log step 161 对最后保存的 trajectory.npz 调用正式 replay；log step 162 的录制与开环重放均为 steps=1000、fell_at=null、tracking_ok=false，非空评分段 1/2/3/5/6 的平移误差依次为 0.7409/0.9732/0.8668/0.2891/0.0425 m/s，偏航误差为 0.5139/0.6438/0.5735/0.9846/0.2278 rad/s。run.json 的最终评分字段同样报告 n_actions=1000、fell=0、worst_lin_err=0.9732、worst_yaw_err=0.9846、deterministic=1，超出 0.2 m/s 与 0.3 rad/s 阈值；完整按指令行进节点未达到。"
+     },
+     {
+      "t": 10.0,
+      "log_step": null,
+      "text_zh": "观察：成功参考视频 3.000–10.000 s 的相同双视图可见四腿较充分伸展并连续行进，地面背景持续变化；本次视频相应区间主要为低蹲小幅迈腿、背景变化较少。参考只用于辨认运动表现；本次精确跟踪失败由本次评分器字段和交付文件重放确认。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.429–13.429 s，Go1 四条腿的足端交替落在棕色粗糙地面上，支撑低蹲的机身并随小幅迈步移动、转向；视频 14.000–20.286 s 足端留在地面，机身停稳。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.9732
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.9846
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 1.429,
+   "log_step": 162,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 1.429–2.857 s，机器人已开始执行行进，但四腿深屈、迈步短，机身仅缓慢移动；后续视频 3.000–13.429 s 继续低蹲挪行。log step 162 的交付文件重放显示首个非空评分段已超标，后续第 2、3、5 段也超标，最终评分无翻倒却未通过跟踪验收。推断：失败发生在正在执行的“按指令行进”步骤；视频 1.429 s 是早期行进不足的导航锚点，不代表精确误差首次越阈时刻。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 1.429,
+    "log_step": 162,
+    "evidence_zh": "观察：视频 1.429–13.429 s，Go1 多数时间以低蹲姿态小幅交替迈腿，身体缓慢挪动，未见翻倒；视频 19.429–20.286 s 仍停在低姿态。log step 162 对交付文件的正式重放确认多个行进段平移和偏航跟踪误差超标，最终评分 worst_lin_err=0.9732 m/s、worst_yaw_err=0.9846 rad/s。推断：可见行进幅度不足且转向跟踪未达标，归为步态。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志",
+    "evidence_zh": "观察：log step 46 记载 agent 为稳定性改用深蹲步态；log step 131–150 的姿态、步幅和转向增益试验仍反复出现显著跟踪误差。log step 157 明确自述最佳完整候选仍超阈；log step 161–162 将最终生成的 1000 行关节目标正式重放，录制与开环重放的各段误差完全一致，并与最终评分一致。推断：生成关节轨迹的步态控制未能把指定机身速度及偏航速度实现出来，造成所见低蹲缓行和跟踪失败；无法据此把问题进一步锁定到某个增益、逆解或接触设计。"
+   }
+  ],
+  "gap_zh": "未提供逐步状态日志或本次交付轨迹的速度、接触力时间序列；双视图随机器人移动且无速度叠字，因此无法独立还原每一评分样本的误差、精确首次失配时刻或具体失效控制环节。节点终值依据交付文件正式重放和一致的最终评分字段，断点时间仅作视频导航。"
+ },
+ "summary_zh": "Go1 执行了完整 1000 控制步且未翻倒，但低蹲小幅迈步的轨迹未达到指令速度跟踪要求，唯一节点“按指令行进”最终为 0/1。交付文件重放与最终评分一致：最差平移误差 0.9732 m/s、偏航误差 0.9846 rad/s，失败现象为步态，根因保留在控制层级。缺少逐步速度与接触力记录，不能确认具体控制参数或精确首次失配时刻。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "唯一节点 0/1 由交付文件正式重放（log step 162）和最终评分字段（worst_lin_err 0.9732、worst_yaw_err 0.9846）共同支持；断点、步态现象、控制根因均有证据，未发现需改内容的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:17:41Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--xcg2udc/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-privileged--xcg2udc/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "30c74b9204288214adef70c7ce2ca6e1273adf05a72cd34535295ac6b959e788",
+   "video_sha256": "a837a45defc97012132f03ee52bdfbbe6bad952f5d5d8149e624ad754dd731c5",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--gr6rnpm.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--gr6rnpm.json
new file mode 100644
index 0000000000000000000000000000000000000000..9868ca1b62eb5f17ca7009fb46e42acf3c37f7e8
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--gr6rnpm.json
@@ -0,0 +1,206 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-standard--gr6rnpm",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 65,
+   "sim_steps": 65,
+   "physics_steps": 325,
+   "steps": 1001,
+   "elapsed_s": 711.0,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.683,
+   "worst_yaw_err": 1.6221,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.158,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.239,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-standard--gr6rnpm/replay.mp4",
+  "duration_s": 1.32,
+  "fps": 25.0,
+  "frames": 33,
+  "sha256": "6d1e9fa21b7bc3d5c77254acb673705afa2a0617c3b503cf08c51e9b09b39390"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.64,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00–0.92 s，Go1 在棕色起伏地面上从站立进入抬腿、落足和短距离移动，已开始行进尝试；初始站立并不满足完成全部指令表的条件。"
+     },
+     {
+      "t": 1.28,
+      "log_step": 24,
+      "text_zh": "观察：视频 1.00–1.28 s，Go1 躯干迅速倾斜、抬起并翻转，末帧未恢复正常足端支撑。log step 24 的机器人服务状态为 episode=11、control_steps=65、done=true、reason=fell；评分器 reward 记录 sim_steps=65、fell=1、live_matches_replay=1。该回合提前终止，未完成要求的 1000 控制步。"
+     },
+     {
+      "t": null,
+      "log_step": null,
+      "text_zh": "观察：inputs/run.json 的评分器 reward 给出 worst_lin_err=0.683 m/s、worst_yaw_err=1.6221 rad/s，分别超过 DAG 阈值 0.2 m/s 和 0.3 rad/s；这些是该提前终止回合已采样评分段的结果，后续未执行段没有完成记录。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.12–0.92 s，Go1 的足端在粗糙地面上落足承重，其他腿交替抬起以尝试行进；视频 1.00–1.28 s 机身倾斜加剧并失去正常支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.683,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 1.6221,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 1.04,
+   "log_step": 24,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 0.64–0.92 s，Go1 正在迈步移动；视频约 1.04 s 起躯干明显倾斜抬起，至视频 1.28 s 继续翻转而未恢复。log step 24 的第 11 次回合由机器人服务以 reason=fell 在 control_steps=65 终止，因此断点是正在执行的“按指令行进”。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.04,
+    "log_step": 24,
+    "evidence_zh": "观察：视频 1.00–1.28 s，Go1 的躯干由接近水平迅速转为大幅倾斜，末段身体翻转、部分足端离地，未恢复稳定支撑；log step 24 的服务终止原因是 fell，评分器 reward 的 fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": 1.04,
+    "log_step": 25,
+    "evidence_zh": "推断：agent 在 log step 22 设置了足端轨迹和姿态修正控制器，在 log step 24 用该控制器执行最后回合；该回合的传感读数经 agent 脚本在 log step 25 打印，仿真 0.88 s 的重力方向约为 [-0.027,-0.051,-0.998]，仿真 1.24 s 变为 [0.879,-0.325,-0.350]，偏航角速度同时由 0.58 变为 -3.92 rad/s。视频 0.88–1.28 s 同样可见机身从较平稳转为大幅翻转，随后服务报 fell（log step 24），支持行进控制未能维持机身稳定；现有证据不足以确定具体控制子类或哪一参数错误。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志或足地接触力记录，无法独立复算评分器的段平均误差，也无法区分具体的落足几何、姿态反馈或增益问题。log step 25 是最后回合传感量的脚本抽样输出，不是完整状态路径；其时间为仿真时钟。未把此前被重置的十次回合当成本次进度或现象，亦未以 agent 自述的重置耗尽作为失败原因。"
+ },
+ "summary_zh": "评分采用最后的第 11 次回合：Go1 已落足并尝试行进，但视频约 1.04 s 开始明显失稳，服务在第 65 个控制步因翻倒终止，唯一节点“按指令行进”始终未达成，最终 0/1。视频中的翻转与最后回合服务状态、评分器 fell=1 一致，两个已报告跟踪误差也均超标；根因支持归入控制，但缺少完整状态和接触力记录，不能确定具体控制错误。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "单节点未达成、断点、平衡现象与控制根因均与视频(约1.0 s起翻倒)、log step 24/25 和评分器 fell=1 一致；log_step 为 steps[].id，无时钟转换问题。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:05:00Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--gr6rnpm/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--gr6rnpm/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "ca8098b8a61366f62604e65c2f22a5bd8fc334aa51a85a4e0e4526d31ed91214",
+   "video_sha256": "6d1e9fa21b7bc3d5c77254acb673705afa2a0617c3b503cf08c51e9b09b39390",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--guyypyw.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--guyypyw.json
new file mode 100644
index 0000000000000000000000000000000000000000..121c4de8736cd7f582ff7f62114c48ec08430048
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--guyypyw.json
@@ -0,0 +1,233 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-standard--guyypyw",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 48,
+   "sim_steps": 48,
+   "physics_steps": 240,
+   "steps": 48,
+   "elapsed_s": 223.0,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 1.349,
+   "worst_yaw_err": 0.8134,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 0,
+   "reset_budget": 10,
+   "episodes": 1,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.329,
+   "peak_processes": 7,
+   "peak_disk_gb": 0.259,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-standard--guyypyw/replay.mp4",
+  "duration_s": 1.0,
+  "fps": 25.0,
+  "frames": 25,
+  "sha256": "afa61acc7da18d58c743f23e05703d9bb463967516e7ce8449a6ba0e0fd0ce24"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 9,
+      "text_zh": "观察：视频 0.00–0.96 s 的全部 25 帧中，Go1 起初足端支撑棕色粗糙地面，随后机身逐渐侧倾，未见交替迈步行进；log step 9 下发 25 行相同 home_pose 关节目标，log step 12 再下发 25 行相同的 [0,0.9,-1.8]×4 目标。"
+     },
+     {
+      "t": 0.96,
+      "log_step": 15,
+      "text_zh": "观察：视频 0.68–0.96 s 中机器人持续向侧面翻转，末帧足端朝侧上方，机身已翻倒；log step 15 的机器人服务状态为 control_steps=48、done=true、reason=fell、episode=1、resumes=0。评分器记录 steps=48、fell=1、worst_lin_err=1.349 m/s、worst_yaw_err=0.8134 rad/s，既未完成 1000 控制步，两项误差也超过 DAG 阈值。"
+     },
+     {
+      "t": null,
+      "log_step": 37,
+      "text_zh": "观察：log step 16–37 是 agent 自建 MuJoCo 模型中的局部测试，最后的 log step 37 仍报告自建模型翻倒；这些测试没有再向评分服务提交动作。评分器记录 resets=0、episodes=1、n_actions=48、live_matches_replay=1，不能把后续本地步态测试计为本次节点进度。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–0.40 s 中 Go1 的足端直接接触粗糙地面承重；视频 0.60–0.96 s 支撑逐渐失稳并侧翻，未见交替迈步。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 1.349
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.8134
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.6,
+   "log_step": 15,
+   "where": "首步前",
+   "evidence_zh": "观察：视频 0.00–0.96 s 中机器人只有站姿支撑及随后的侧翻，没有开始按指令迈步；log step 9、12 提交的均为固定关节姿态，log step 15 服务报告评分回合在 control_steps=48 时因 fell 结束。推断：交付的是站姿诊断基线，实际行进尚未启动即失衡；依 CASES.md FAIL-13 标首步前，视频约 0.60 s 的明显侧倾作为断点导航锚。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "没尝试",
+    "item": "没尝试",
+    "t": 0.0,
+    "log_step": 9,
+    "evidence_zh": "观察：log step 9、12 只向评分服务下发重复的固定站姿关节目标；视频 0.00–0.96 s 未见机器人启动交替迈步，后续 log step 16–37 的步态实验均在 agent 自建模型中进行。推断：评分回合未启动按指令行进，属于基线未开始该步骤。"
+   },
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 0.6,
+    "log_step": 15,
+    "evidence_zh": "观察：视频约 0.40 s 开始可见机器人侧倾，视频 0.60–0.84 s 倾斜持续加重，视频 0.88–0.96 s 躯干翻向侧面、腿足朝侧上方；log step 15 的服务终止原因为 fell，评分器 fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "交付",
+    "item": "交付操作出错",
+    "evidence_zh": "观察：log step 9、12 向评分服务提交的均为重复固定关节姿态；agent 在 log step 10 自述为“short diagnostic run”，log step 12 的代码测试无相机调用耗时。log step 3 的服务规格说明仅最后回合受评，reset 可替换当前回合；log step 15 服务报告 control_steps=48、done=true、reason=fell、episode=1、resets_left=10。log step 16–37 只有自建 MuJoCo 模型的本地测试，没有再 reset 或向评分服务提交动作；inputs/run.json 的 reward 记录 resets=0、episodes=1、n_actions=48、live_matches_replay=1，视频 0.00–0.96 s 也只有站姿到侧翻。推断：agent 未将诊断回合重置并替换为任务行进回合，使已翻倒的站姿诊断成为最终受评交付，属于交付操作出错。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有状态日志；原始 log step 9、12 的完整执行返回未保留在可见输出中，部分规格及本地测试代码也已被输入日志截断。视频可核对未迈步及侧翻，评分器给出最终误差，但无法独立重算逐步速度、接触力和确切翻倒阈值越界时刻。评分回合未发现行进控制器被执行，后续本地测试不属于受评回合，因此行进控制层面不可判，也不能仅由固定站姿翻倒确定反馈控制、轨迹或参数缺陷。"
+ },
+ "summary_zh": "评分回合仅提交固定站姿目标，执行 48/1000 个控制步，按指令行进节点从未达成，最终进度为 0/1。机器人在视频约 0.60 s 明显侧倾，视频 0.96 s 已翻倒，实际行进未启动，断点为首步前；agent 在诊断翻倒后仍有 10 次 reset 额度却仅继续本地模型测试，最终保留诊断回合作为评分交付，根因为交付操作出错。无状态日志，精确速度误差采用评分器数值；未发现行进控制器在评分回合执行，控制层面不可判。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 2,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "fix",
+    "conf": 7,
+    "summary_zh": "WHERE 标签（节点未达、断点首步前、0.6 s 锚点、平衡 + 没尝试）与视频、服务读数一致。问题在根因：把评分回合当作站姿诊断测试用掉，诊断翻倒后 resets_left=10 却从未 reset、也没再提交，最终判分的就是这次诊断回合；该证据支持“交付·交付操作出错”，而单独标“控制”缺少反馈控制失败的证据（提交的是固定站姿，不是行进控制器）。",
+    "n_issues": 2,
+    "n_major": 2
+   },
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 2,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "节点未达、断点（首步前，0.6 s 锚点）、没尝试 + 平衡、交付·交付操作出错根因均与视频（0–0.96 s 站姿到侧翻）、评分器读数（fell=1、steps=48、resets=0、episodes=1）一致；第 1 轮的修改已落实，未发现新错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:25:06Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--guyypyw/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--guyypyw/04_review_r1"
+   },
+   {
+    "stage": "reconcile",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--guyypyw/05_reconcile_r1"
+   },
+   {
+    "stage": "traj_review",
+    "round": 2,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--guyypyw/04_review_r2"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "bb30522d6a613ec0bb946de91d1037a62c1cc2b097f8a430174690fbb5d9c991",
+   "video_sha256": "afa61acc7da18d58c743f23e05703d9bb463967516e7ce8449a6ba0e0fd0ce24",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--q3yffzk.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--q3yffzk.json
new file mode 100644
index 0000000000000000000000000000000000000000..ed3c244d5fde94d31e524a22167027a9c3677953
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--q3yffzk.json
@@ -0,0 +1,221 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-standard--q3yffzk",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "graded_by_live": 1,
+   "replays": 0,
+   "finalized": 1,
+   "missing_log": 0,
+   "sim_error": 0,
+   "grader_error": 0,
+   "infra_error": 0,
+   "n_actions": 80,
+   "sim_steps": 80,
+   "physics_steps": 400,
+   "steps": 146,
+   "elapsed_s": 1731.2,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 1.0683,
+   "worst_yaw_err": 1.4999,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "instance_mismatch": 0,
+   "no_verdict": 0,
+   "gave_up": 0,
+   "sim_crashed_on_call": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 1,
+   "reset_budget": 10,
+   "episodes": 2,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.293,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.239,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-standard--q3yffzk/replay.mp4",
+  "duration_s": 1.64,
+  "fps": 25.0,
+  "frames": 41,
+  "sha256": "b94f786fcf9ce1a8f328d98ff012386ffac8c6364d7d25216ac8aa06d987a112"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.4,
+      "log_step": 23,
+      "text_zh": "观察：本次视频 0.00–1.60 s 的跟随与侧视画面显示 Go1 在棕色起伏地面上屈伸四腿、迈步并摇晃；视频约 1.12–1.28 s 机身明显侧倾、下沉并趴倒，视频 1.28–1.60 s 保持侧倒姿态并继续翻转，末段一侧腿抬高，未恢复站立。log step 23 记录重置后的 episode=2，服务以 control_steps=80、done=true、reason=fell 结束；该回合对应评分器 episodes=2、resets=1、sim_steps=80、fell=1。"
+     },
+     {
+      "t": 1.6,
+      "log_step": 23,
+      "text_zh": "推断：评分回合只执行 80/1000 个控制步且已翻倒，未完成完整指令表，因此按指令行进在整段视频中均不成立。评分器另报告 worst_lin_err=1.0683 m/s、worst_yaw_err=1.4999 rad/s，分别超过 0.2 m/s 和 0.3 rad/s 阈值；这些精确误差来自评分器字段，而非画面测量。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.20–约 1.1 s 的双视图显示 Go1 的足端在粗糙地面落地承重，四腿屈伸并交替抬起；视频约 1.12–1.28 s 机身侧倾、下沉并趴倒，足端支撑失稳；视频 1.28–1.60 s 保持侧倒姿态并继续翻转。实际交互包括足端与地面的支撑接触。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 1.0683,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 1.4999,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 1.2,
+   "log_step": 23,
+   "where": "操作中",
+   "evidence_zh": "观察：视频约 1.12–1.28 s，正在迈步的 Go1 机身明显侧倾、下沉并趴倒；视频 1.28–1.60 s 保持侧倒姿态并继续翻转，末段腿抬向上方，直到视频结束均未恢复站立。log step 23 的第 2 回合现场服务以 fell 终止，只执行 80 个控制步。失败发生在按指令行进过程中，视频约 1.2 s 为失稳过程的粗略定位。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.2,
+    "log_step": 23,
+    "evidence_zh": "观察：视频 1.04 s Go1 仍由足端支撑迈步；视频 1.12、1.20、1.28 s 的连续帧显示机身明显侧倾、下沉并趴倒，不再由足端稳定支撑；视频 1.28–1.60 s 保持侧倒姿态并继续翻转，末段一侧腿朝上。log step 23 服务返回 done=true、reason=fell，评分器 fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：log step 14–15 编写的控制器用足端轨迹、姿态反馈及解析逆解生成关节位置目标；log step 22 修改落足与姿态反馈后，log step 23 用该控制器运行评分回合，仍在 80 个控制步内翻倒。视频 0.40–1.00 s 可见机身摇晃，视频约 1.12–1.28 s 侧倾并趴倒、视频 1.28–1.60 s 继续翻转，说明实际执行的步态与姿态控制未能稳定机身；证据不足以进一步归到某项参数、接触布局或反馈子类。",
+    "source_zh": "日志 / 视频"
+   }
+  ],
+  "gap_zh": "没有独立状态日志，无法逐步核对足端接触力、关节跟踪和机身姿态数值；精确跟踪误差使用评分器字段。log step 21 是已被重置的第 1 回合，不作为节点进度或本次翻倒现象的证据；log step 25 在第 2 回合结束后才写出的进一步修改没有运行记录，不能当作评分回合已执行的控制。现有视频与日志支持控制失稳，但不足以确定具体控制子类。"
+ },
+ "summary_zh": "评分的是 log step 23 重置后运行的第 2 回合：Go1 已在粗糙地面迈步并以足端支撑，但仅执行 80/1000 个控制步，唯一的按指令行进节点从未达成，最终为 0/1。视频约 1.12–1.28 s 机器人侧倾并趴倒，视频 1.28–1.60 s 继续翻转，现场服务以 fell 结束；实际步态与姿态控制未维持机身稳定，评分器报告的平移和偏航误差也均超阈值。没有独立状态日志，具体失稳参数或接触机制无法确定，回合结束后保存的脚本修改不计入本次执行。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 2,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "fix",
+    "conf": 7,
+    "summary_zh": "节点、断点（command_schedule_completed）、平衡现象和控制根因均有依据；唯一问题是翻倒起始时间的观察描述过晚：视频约 1.1–1.3 s 机身已侧倾并趴下，而文件写成 1.48 s 前仍由足端稳定支撑、1.52 s 才开始翻倒。",
+    "n_issues": 1,
+    "n_major": 0
+   },
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 2,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "唯一节点未达成、断点 command_schedule_completed（约 1.2 s）、平衡现象和控制根因均有帧与评分器字段支持；上一轮的时间问题已修正。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:26:53Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--q3yffzk/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--q3yffzk/04_review_r1"
+   },
+   {
+    "stage": "reconcile",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--q3yffzk/05_reconcile_r1"
+   },
+   {
+    "stage": "traj_review",
+    "round": 2,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--q3yffzk/04_review_r2"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "bc3d953ea9389d4c41fd482095f0e4a6b7cbde22434a990e4cd0432df35d24e9",
+   "video_sha256": "b94f786fcf9ce1a8f328d98ff012386ffac8c6364d7d25216ac8aa06d987a112",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--x9yf6dm.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--x9yf6dm.json
new file mode 100644
index 0000000000000000000000000000000000000000..f660d99b928f373eeb8b833bc902b21051078ead
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--x9yf6dm.json
@@ -0,0 +1,205 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-standard--x9yf6dm",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 71,
+   "sim_steps": 71,
+   "physics_steps": 355,
+   "steps": 1300,
+   "elapsed_s": 1382.0,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 1.0488,
+   "worst_yaw_err": 0.8984,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.312,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.83,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-standard--x9yf6dm/replay.mp4",
+  "duration_s": 1.44,
+  "fps": 25.0,
+  "frames": 36,
+  "sha256": "f956881981670bc26537324bac2d8f1f70b9773092462813f8c00331d26bacdf"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.12,
+      "log_step": 72,
+      "text_zh": "观察：视频 0.00–0.24 s 中，Go1 从初始姿态下沉，四足落到起伏棕色地面上；视频 0.28–1.00 s 中机器人摆腿并移动，尚未完成整个指令表。log step 72 明确重置到最终第 11 回合，服务随后记录该回合的运动与传感数据，先前回合不计入本次进度。"
+     },
+     {
+      "t": 1.4,
+      "log_step": 72,
+      "text_zh": "观察：视频 1.08–1.40 s 中，Go1 躯干持续侧倾并翻倒，片段随即结束；log step 72 的最终回合结束状态为 control_steps=71、done=true、reason=fell。评分器报告 sim_steps=71、fell=1、live_matches_replay=1、worst_lin_err=1.0488 m/s、worst_yaw_err=0.8984 rad/s。"
+     },
+     {
+      "t": 1.4,
+      "log_step": 72,
+      "text_zh": "推断：log step 72 的最终回合控制步按每步 0.02 s 换算，71×0.02=1.42 s；已用视频 1.36–1.40 s 的翻倒和 1.44 s 的片长核对。回合因翻倒在 1000 控制步之前结束，且已评分段误差超过 0.2 m/s、0.3 rad/s，故“按指令行进”从未达成；不是仅凭 success=0 判定。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.12–1.00 s 中，Go1 的足端先落在粗糙地面上承重，随后摆腿、交替着地以支撑并移动机身；视频 1.08–1.40 s 中支撑失稳并侧翻。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 1.0488,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.8984,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 1.08,
+   "log_step": 72,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 1.08–1.40 s 中，正在摆腿行进的 Go1 躯干侧倾逐渐增大并翻倒，未恢复；log step 72 确认这是最后一次重置后的第 11 回合，服务在该回合第 71 控制步以 reason=fell 结束。失败发生在“按指令行进”的执行中。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.08,
+    "log_step": 72,
+    "evidence_zh": "观察：视频 1.08–1.40 s 中，Go1 的机身从侧倾转为侧翻，足端未能恢复稳定支撑；log step 72 中服务记录最终回合仿真 1.00 s 的重力方向为 [0.08,-0.32,-0.94]，仿真 1.42 s 为 [0.53,-0.85,0.08]，同时 done=true、reason=fell，评分器 fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 72,
+    "evidence_zh": "推断：log step 70 的 agent 自述将以宽站姿、渐进对角迈步和姿态反馈执行指令，log step 72 显示其控制脚本实际启动最终回合，服务在仿真 1.00 s 已读到滚转角速度 4.95 rad/s，到仿真 1.42 s 增至 6.77 rad/s 并因翻倒结束；视频 1.08–1.40 s 同样显示倾斜持续扩大。该步已启动，但关节运动控制未能稳定机身和维持支撑，归为控制；现有证据不能证明某个具体增益或接触布局就是原因。",
+    "source_zh": "日志 / 视频"
+   }
+  ],
+  "gap_zh": "没有独立状态日志，精确速度误差采用评分器字段；日志仅给最终回合的稀疏传感采样，且控制器多次编辑的正文未保留（编辑调用多显示 patch));），不能进一步区分支撑布局、步态轨迹与反馈参数的贡献。视频全 36 帧已检查，包含最后一秒；这些缺口不影响节点未达与翻倒的判定。"
+ },
+ "summary_zh": "评分尝试为 log step 72 最后重置后的第 11 回合：Go1 足端落地并摆腿行进，但视频约 1.08 s 起持续侧倾，至视频 1.40 s 已翻倒，仅执行 71/1000 控制步，“按指令行进”未达成（终态 0/1）。平移及偏航误差分别为 1.0488 m/s 和 0.8984 rad/s，均超阈值；失衡可归为控制失败，但没有独立状态日志及完整控制器修改内容来定位更细根因。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "单节点未达、最终回合 71 控制步翻倒的判定与视频 36 帧、log step 72 的服务读数及评分器字段一致；时钟换算（71×0.02=1.42 s 对 1.44 s 片长）已声明并核对，断点、平衡现象与“控制”根因均有证据支持，未发现需要改内容的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:19:24Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--x9yf6dm/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--x9yf6dm/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "d516f427ead42211af79981c229a80f4b027253697188504b5a100c3ec8f27d0",
+   "video_sha256": "f956881981670bc26537324bac2d8f1f70b9773092462813f8c00331d26bacdf",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--xjqeink.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--xjqeink.json
new file mode 100644
index 0000000000000000000000000000000000000000..7058c91537fbeacc3b179bc0c5e384789af5fee1
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--xjqeink.json
@@ -0,0 +1,200 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-standard--xjqeink",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 41,
+   "sim_steps": 41,
+   "physics_steps": 205,
+   "steps": 479,
+   "elapsed_s": 1099.6,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": -1.0,
+   "worst_yaw_err": -1.0,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-standard--xjqeink/replay.mp4",
+  "duration_s": 0.84,
+  "fps": 25.0,
+  "frames": 21,
+  "sha256": "0d9cbb49ee65f96dc618dedae0117413dd6ca29835dd38ce20999142c0c9c1f7"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.4,
+      "log_step": 108,
+      "text_zh": "观察：逐帧检查视频 0.00–0.80 s 的全部 21 帧；Go1 起初以足端接触起伏地面，视频约 0.24 s 后躯干逐渐侧倾，视频 0.60–0.80 s 滚向侧面、腿伸向上方，始终没有完成整段指令行进。"
+     },
+     {
+      "t": 0.8,
+      "log_step": 112,
+      "text_zh": "观察：log step 112 的服务状态为 episode=11、control_steps=41、done=true、reason=\"fell\"；评分器同样记录 n_actions=41、fell=1、live_matches_replay=1。最后回合在视频结束前翻倒，远未执行完要求的 1000 控制步；log step 110 的后续重置请求被拒绝，log step 114 只修改脚本默认值，未产生新的评分回合。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.04–0.44 s 可见 Go1 足端与棕色粗糙地面接触、腿承托机身；随后机身侧倾，视频 0.60–0.80 s 足端支撑姿态崩溃。没有将成功参考的交替步态当成本回合接触事实。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": -1.0
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": -1.0
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.4,
+   "log_step": 108,
+   "where": "操作中",
+   "evidence_zh": "观察：视频约 0.24–0.44 s 中，Go1 在足端支撑与姿态调整过程中开始持续侧倾，视频 0.60–0.80 s 翻向侧面，未恢复；log step 112 的最后回合服务状态为 control_steps=41、reason=\"fell\"。失败发生在按指令行进的启动及稳定过程中，未完成 20 秒指令表。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 0.6,
+    "log_step": 112,
+    "evidence_zh": "观察：视频 0.48–0.80 s 的连续画面中，Go1 躯干由倾斜发展为横倒，原来向下支撑的腿转向侧上方，没有重新站稳；log step 112 服务报告该评分回合以 fell 结束，评分器 fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "log_step": 108,
+    "source_zh": "日志 / 视频",
+    "evidence_zh": "推断：log step 108 的脚本已把姿态补偿改为 rot_x(ca) @ rot_y(cb)，并以 chunk=2 的关节目标控制运行最后回合；该次诊断输出的重力横向分量先减小后再次增大，仍以 fell 结束。视频 0.24–0.80 s 同样显示侧倾不断扩大直至翻倒，表明本回合姿态反馈未能稳定足端承托下的机身；log step 112 确认这是最后的第 11 回合。不能把之前回合已修改的补偿符号错误直接当成本回合的原因，也没有证据锁定某个具体增益。"
+   }
+  ],
+  "gap_zh": "没有状态日志，无法逐控制步独立核对足端接触力及姿态误差，具体反馈增益或延迟的贡献仍不确定。评分器 worst_lin_err 与 worst_yaw_err 均为 -1.0，原样记录，但不能解释为有效跟踪误差或达标证据；最后回合在误差统计窗口之前即结束。log step 108 的脚本自计 steps=42 与 log step 112 的服务 control_steps=41 不同，节点与断点时间均取自已检查的视频帧，未将脚本自计时间映射成视频时间。",
+  "unmet": {
+   "command_schedule_completed": false
+  }
+ },
+ "summary_zh": "评分的最后第 11 回合只执行 41/1000 控制步，唯一节点“按指令行进”从未完成，最终为 0/1。视频约 0.24 s 后 Go1 持续侧倾，视频 0.60–0.80 s 明显翻倒；最后回合的控制日志支持姿态反馈未能稳住机身。无状态日志，无法进一步锁定具体控制参数，评分器的两项误差值 -1.0 也不能用于判断跟踪精度。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "标签与证据一致：视频 0.84 s 内 Go1 逐渐侧倾并翻倒，log step 112 的服务状态（episode 11、control_steps=41、reason=fell）和评分器 fell=1 支持节点未达成；断点、平衡现象、反馈控制根因（标为推断）均有依据，时钟未把智能体时间转换成视频时间。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:19:07Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--xjqeink/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--xjqeink/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "148517f8f4786eecf31eff50b576d59f06338b5676c00dbc0e79ae817f62de09",
+   "video_sha256": "0d9cbb49ee65f96dc618dedae0117413dd6ca29835dd38ce20999142c0c9c1f7",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--yueveke.json b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--yueveke.json
new file mode 100644
index 0000000000000000000000000000000000000000..ed4a9d8dcb43dd7752019244247caf026f69e0ca
--- /dev/null
+++ b/phase2/mujoco-playground/go1_joystick_rough_terrain/trajectories/mujoco_playground_go1_joystick_rough_terrain-standard--yueveke.json
@@ -0,0 +1,206 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_go1_joystick_rough_terrain-standard--yueveke",
+ "taskKey": "mujoco-playground/go1_joystick_rough_terrain",
+ "dag_sha256": "4c574e4f27ece4fa9570e6019e251e5bec22eaa107b4e280f165495fad702f80",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "graded_by_live": 1,
+   "replays": 0,
+   "finalized": 1,
+   "missing_log": 0,
+   "sim_error": 0,
+   "grader_error": 0,
+   "infra_error": 0,
+   "n_actions": 208,
+   "sim_steps": 208,
+   "physics_steps": 1040,
+   "steps": 1269,
+   "elapsed_s": 1049.7,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.8284,
+   "worst_yaw_err": 1.47,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "instance_mismatch": 0,
+   "no_verdict": 0,
+   "gave_up": 0,
+   "sim_crashed_on_call": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.36,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.251,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_go1_joystick_rough_terrain-standard--yueveke/replay.mp4",
+  "duration_s": 4.2,
+  "fps": 25.0,
+  "frames": 105,
+  "sha256": "4f890a227a5cc7c0a2bc9a2069dcd2879d2c49b80b56f174931f763c0c047fed"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 105,
+      "text_zh": "观察：log step 105 将现场服务重置到 Episode 11，log step 106 启动最终控制脚本；视频 0.0–3.6 s 的跟随与侧视画面显示 Go1 在棕色起伏地面上交替迈腿、移动和转向，确已开始行进。"
+     },
+     {
+      "t": 3.84,
+      "log_step": 109,
+      "text_zh": "观察：视频 3.76–3.84 s 躯干开始明显侧倾，视频 3.88–4.16 s 持续翻滚，最后一帧仍未恢复直立支撑；log step 109 的服务状态为 episode=11、control_steps=208、done=True、reason=fell，评分器同时记录 fell=1、success=0。"
+     },
+     {
+      "t": 4.16,
+      "log_step": 109,
+      "text_zh": "推断：最终回合的 208 控制步按日志提供的 50 Hz 对应仿真 4.16 s，与视频 4.16 s 的翻倒末帧核对一致；尚未达到任务要求的 1000 控制步，故完整指令跟踪节点从未成立。评分器 worst_lin_err=0.8284 m/s、worst_yaw_err=1.47 rad/s 也分别超过 0.2 m/s、0.3 rad/s 阈值；精确误差取自评分器，未从画面估算。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.2–3.6 s，Go1 的足端交替落在粗糙地面上，支承机身并迈步移动；视频 3.8 s 后侧倾翻倒，足端支撑未能维持机身直立。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.8284,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 1.47,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 3.84,
+   "log_step": 109,
+   "where": "操作中",
+   "evidence_zh": "观察：Go1 正在粗糙地面上持续迈步，视频 3.76–3.84 s 起机身明显侧倾，视频 3.88–4.16 s 继续翻倒且没有恢复；log step 109 确认最终现场回合 Episode 11 在 208 控制步以 fell 结束，完整指令表尚未走完。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 3.84,
+    "log_step": 109,
+    "evidence_zh": "观察：视频 3.76–4.16 s 的两幅视图中，Go1 躯干从侧倾发展为翻倒，四腿无法维持直立支撑；log step 109 的现场服务终止原因为 fell，评分器 fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": 3.84,
+    "log_step": 104,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 104 的最终脚本读取本体观测、生成足端运动并经逆解提交关节位置目标，log step 105–106 在重置后的 Episode 11 执行该控制器；视频 3.76–4.16 s 的连续侧倾翻倒以及 log step 109 的 fell 终止表明，这套行进控制未能在粗糙地面运动中维持机身稳定。评分器还记录平移和偏航跟踪误差超标。证据支持一般控制失效，不能确定为某个增益、逆解或接触布局错误。"
+   }
+  ],
+  "gap_zh": "没有状态日志；最终回合日志提供终止状态，但没有逐控制步的姿态、关节目标及接触力序列，且 log step 104 的控制器中段代码在输入记录中已省略，无法进一步定位具体控制环节。原生速度误差采用评分器汇总值，视频只能核对运动和翻倒。"
+ },
+ "summary_zh": "最终评分回合 Episode 11 中，Go1 已在粗糙地面上迈步移动和转向，但视频约 3.8 s 开始侧倾翻倒，服务在 208/1000 控制步因 fell 终止，唯一节点“按指令行进”未达成（0/1）。平移和偏航跟踪误差也均超标，现有脚本与画面支持一般控制失效；agent 在 log step 112 的正面总结不改变本回合的翻倒事实。缺少最终回合逐步状态及完整控制器中段代码，无法进一步细分控制根因。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标注与视频和日志一致：Episode 11 在约 3.8 s 侧倾翻倒，208 控制步以 fell 终止，节点未达成，断点、平衡现象和一般控制根因均有证据支持。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:19:35Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--yueveke/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_go1_joystick_rough_terrain-standard--yueveke/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "d24e1fa7fef6b39f9f1ca0c905f0ae7ec89bb865872177029dab94b5a670e24e",
+   "video_sha256": "4f890a227a5cc7c0a2bc9a2069dcd2879d2c49b80b56f174931f763c0c047fed",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--8jeujnz.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--8jeujnz.json
new file mode 100644
index 0000000000000000000000000000000000000000..10725b579e0819bb0f4ba73bd2bdffce92d63deb
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--8jeujnz.json
@@ -0,0 +1,197 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-privileged--8jeujnz",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.7412,
+   "worst_yaw_err": 0.6272,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 63,
+   "agent_resets": 70,
+   "agent_sim_steps": 217430,
+   "agent_replays": 7,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.97,
+   "peak_processes": 10,
+   "peak_disk_gb": 0.833,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--8jeujnz/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "dbb98b549128fe30109a534f99994dceea3caeb92b298f0be43e24fbd68d8844"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 3.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.000–4.429 s，Spot足端交替触地、四腿持续踏步，机身保持离地；相对地面格线的机身转向幅度很小。成功参考视频同一首段有清晰的持续转向，仅用于比较可见表现。"
+     },
+     {
+      "t": 13.0,
+      "log_step": null,
+      "text_zh": "观察：视频5.000–12.000 s机器人继续迈腿，视频13.000–16.714 s基本原地站立；视频17.429–20.286 s恢复小幅踏步，但平移及转向幅度仍很小，最后一帧机身仍由足端支撑。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：inputs/run.json的评分器reward记录n_actions=1000、horizon=1000、fell=0、deterministic=1、replay_success=0，worst_lin_err=0.7412 m/s、worst_yaw_err=0.6272 rad/s，分别超过0.2 m/s和0.3 rad/s阈值。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "推断：结合视频中持续踏步但转向不足的表现，以及评分器明确超标的分段速度误差，完整时域虽已执行，整段指令跟踪的验收条件仍未成立；该完成事件没有达到时刻。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.000–12.000 s及17.429–20.286 s，Spot四条腿的足端交替落地承重并尝试行进；视频13.000–16.714 s足端支撑机身近似原地站立。接触对象为平地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.7412,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.6272,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 3.0,
+   "log_step": null,
+   "where": "操作中",
+   "evidence_zh": "观察：视频0.571–4.000 s，Spot已持续踏步，但首段机身转向很小；视频17.429–20.286 s的末段也仅小幅挪动。评分器记录完整1000步、未跌倒，但最大分段线速度和偏航角速度误差均超标。推断：失败发生在按指令行进的执行中；视频3.000 s是首段跟踪不足的导航锚点，不代表某个瞬时误差越界的精确时刻。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 3.0,
+    "log_step": null,
+    "evidence_zh": "观察：视频0.571–4.000 s机器人不断迈腿，却只有小幅机身转向；视频17.429–20.286 s虽然又开始踏步，横移和转向仍很弱，足端始终继续支撑机身。成功参考视频首段及视频18.000–20.000 s末段可见明显机身转向。推断：本次步行的平移、转向未跟上指令，属于未跌倒情况下的步态表现不足；评分器fell=0与此一致。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：视频0.571–4.000 s及17.429–20.286 s显示机器人能够执行踏步并保持支撑，却没有产生足够的指令对应运动；评分器worst_lin_err=0.7412、worst_yaw_err=0.6272进一步确认控制生成的行进结果不达标。日志中agent自述其按足部几何生成关节轨迹（log step 10），反复调整步态以改善跟踪（log step 53、170），最终仍明确承认跟踪失败（log step 194），支持步态控制能力不足；现有证据不足以确定具体的参数、接触设计或反馈环节。",
+    "source_zh": "视频 / 评分器字段 / 日志"
+   }
+  ],
+  "gap_zh": "没有逐步状态日志；低帧率跟随视角不能直接复算机体坐标系分段平均速度误差或定位误差开始的精确控制步，数值结论采用评分器字段。部分会话测试与最终交付不同：log step 169的末段误差为0.7933/0.6674，不能替代本次评分器的0.7412/0.6272；log step 185所述试验跌倒也不属于本次未跌倒的交付轨迹。现有日志多次编辑与运行的输出为空，无法进一步确定控制根因的子类。",
+  "unmet": {
+   "command_schedule_completed": false
+  }
+ },
+ "summary_zh": "Spot以足端支撑完成1000个控制步且未触发跌倒，但首段转向不足，末段横移与转向也很弱，唯一节点“按指令行进”未完成（0/1）。断点在行进操作中，现象为步态不足，根因归于控制；评分器最大分段线速度误差0.7412 m/s、偏航角速度误差0.6272 rad/s均超标。没有逐步状态日志，具体控制子类与精确偏差起点无法确认。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "唯一节点未完成、断点（首段指令跟踪）、步态现象与停在“控制”的根因均与视频、评分器字段（worst_lin_err 0.7412、worst_yaw_err 0.6272、fell=0）及日志（step 169/185/194）一致；gap_zh 对会话测试数值（0.7933/0.6674）与最终交付的区分正确。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:20:17Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--8jeujnz/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--8jeujnz/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "d2c30b2936b54cee1a2c196fd296f499d59f1bd91bdee6949e45fbb1ceffe009",
+   "video_sha256": "dbb98b549128fe30109a534f99994dceea3caeb92b298f0be43e24fbd68d8844",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--bsgtzeo.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--bsgtzeo.json
new file mode 100644
index 0000000000000000000000000000000000000000..848a5785cde83315646870142344ab475f9b9cfa
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--bsgtzeo.json
@@ -0,0 +1,193 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-privileged--bsgtzeo",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.793,
+   "worst_yaw_err": 0.6673,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 14,
+   "agent_resets": 434,
+   "agent_sim_steps": 64585,
+   "agent_replays": 403,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 1.485,
+   "peak_processes": 13,
+   "peak_disk_gb": 0.241,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--bsgtzeo/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "ba71b08d9299eb55a51f14704ca17a0b423154acdc5040f47a25693fafda27ec"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 7,
+      "text_zh": "观察：视频 0.000、0.143、0.286、0.429、0.571、0.714、0.857、1.000、1.143 s 的跟随视角与侧视中，Spot四足着地，开头机身略下沉后维持站姿，没有开始交替迈步；log step 7 的脚本将同一组home关节目标重复1000行并保存为最终交付位置的备用轨迹。"
+     },
+     {
+      "t": 10.0,
+      "log_step": null,
+      "text_zh": "观察：视频 1.0–19.0 s 每隔约1秒检查的两个视角中，Spot足端、机身朝向及其相对地面格线的位置基本不变，没有连续平移或转向；视频 19.286–20.286 s 逐帧仍为四足站立。成功参考视频 1.0–11.0 s 及17.0–19.0 s 同样视角可见明显踏步和朝向变化，仅用来确认这些运动在该视角可辨。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 7,
+      "text_zh": "推断：本次视频的静止表现、评分器n_actions=1000、fell=0、worst_lin_err=0.793及worst_yaw_err=0.6673，与log step 7保存备用站姿轨迹时的回放结果吻合；该步还记录非零指令段的误差超限。整段虽运行满步数且未倒地，却未完成指令跟踪，节点从未达到。log step 12–27的步态调试是会话测试，不能计入该视频。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.000–20.286 s 所检帧中，Spot四条腿的足端接触平地并承重，主要保持静态站姿，未见行进所需的交替迈步。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.793
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.6673
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.0,
+   "log_step": 7,
+   "where": "首步前",
+   "evidence_zh": "观察：视频 0.000–1.143 s 密集帧中，Spot仅轻微下沉并四足原地支撑；随后视频 2.0–19.0 s 的逐秒帧及19.286–20.286 s 的逐帧检查均未见迈步、平移或转向。log step 7明确保存固定home目标的1000行备用轨迹，评分器的两项最差段误差与该步回放完全一致；所交付轨迹没有启动按指令行进。"
+  },
+  "phenomena": [
+   {
+    "cat": "没尝试",
+    "item": "没尝试",
+    "t": 0.0,
+    "log_step": 7,
+    "evidence_zh": "观察：视频 0.000–20.286 s 所检两视角帧中，Spot维持四足站立，足端与格线关系基本不变，没有启动行进；log step 7保存的脚本内容为重复home站姿目标1000行。此处指评分视频中的行进未尝试，会话中的步态测试不改变这一事实。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 25,
+    "source_zh": "日志",
+    "evidence_zh": "推断：agent在log step 7先保存固定站姿备用轨迹；后续log step 16的完整步态回放仍为success=False，最差线速度/偏航误差为0.7991/0.7743；log step 25优化后的最佳候选五段误差为(0.3575,0.4089)、(0.2176,0.4217)、(0.3196,0.3222)、(0.074,0.3969)、(0.3747,0.4744)，tracking_ok与success仍为false；log step 26–27继续诊断该候选的速度、角速度及关节目标偏差，未记录将新步态覆盖交付文件。结合视频 0.000–20.286 s 静态站立及评分误差与log step 7吻合，未解决的行进控制问题使最终交付停留在站姿备用轨迹；证据只支持宽泛的控制归因，不能指定某个已修复的IK错误为最终根因。"
+   }
+  ],
+  "gap_zh": "没有状态日志，不能从逐控制步数据核验速度、倾角或精确接触时序；本次节点未完成由静止视频及原生评分读数共同确定。日志止于log step 27的会话诊断，没有最终完成报告，也未说明为何未以新候选替换备用轨迹；具体控制子类及该交付选择的决策原因不确定。",
+  "unmet": {
+   "command_schedule_completed": false
+  }
+ },
+ "summary_zh": "评分轨迹运行了1000个控制步，Spot始终四足支撑且未倒地，但没有启动指令要求的平移或转向，“按指令行进”未达到，最终为0/1。最差段线速度误差0.793 m/s、偏航角速度误差0.6673 rad/s均超限；视频和评分读数对应log step 7保存的站姿备用轨迹，后续会话步态测试仍未通过验收，支持宽泛的控制根因。没有状态日志，具体控制缺陷及未替换备用轨迹的决策原因未能确定。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标签与证据一致：交付轨迹为log step 7保存的home站姿备用轨迹（回放worst_lin_err 0.793/yaw 0.6673与评分读数完全相同），视频全程静止站立，节点未达到、final=0；断点、没尝试、控制归因均有日志支持，无需修改。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:17:55Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--bsgtzeo/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--bsgtzeo/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "94c902ef8c41a08528abf1f2f4b254434ebbc504da629d16f9b7e661cb77c009",
+   "video_sha256": "ba71b08d9299eb55a51f14704ca17a0b423154acdc5040f47a25693fafda27ec",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--pl68j2d.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--pl68j2d.json
new file mode 100644
index 0000000000000000000000000000000000000000..e70386e0f9a1cd4bacaa2af3d2cbf95f7d4c2786
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--pl68j2d.json
@@ -0,0 +1,178 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-privileged--pl68j2d",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.1001,
+   "worst_yaw_err": 0.2461,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 453,
+   "agent_resets": 457,
+   "agent_sim_steps": 1361395,
+   "agent_replays": 4,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--pl68j2d/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "695995b3a4e9a3442e20fb9d036b74b7eea8eff23474e1e2a3fa9a42665b096c"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [
+     [
+      20.286,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 4.0,
+      "log_step": null,
+      "text_zh": "观察：已逐帧检查视频0.000–20.286 s的143帧；视频0.000–4.000 s，黄色Spot交替抬足、落足，机身相对地面格线移动并转向；视频4.143–8.000 s朝向逐渐转回，视频8.143–12.000 s继续踏步行进。足端与平地接触支撑，机身保持离地。"
+     },
+     {
+      "t": 14.0,
+      "log_step": null,
+      "text_zh": "观察：视频12.571–16.143 s，Spot仍有小幅踏步，但机身朝向和相对地面格线的位置变化很小，呈近似原地停留；视频16.571–20.286 s再次明显改变朝向并移动。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 30,
+      "text_zh": "观察：视频19.429–20.286 s直到末帧，Spot仍以足端支撑并转向，没有机身倒地。log step 30中agent脚本重新生成最终交付轨迹后，评分工具对该文件的两次独立回放均报告steps=1000、horizon=1000、success=true、fell_at=null、tracking_ok=true，最终状态哈希同为19479f6ff5eb；inputs/run.json的reward同时给出success=1、replay_success=1、deterministic=1、fell=0。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 30,
+      "text_zh": "观察：评分器reward的worst_lin_err=0.1001 m/s、worst_yaw_err=0.2461 rad/s，分别低于DAG的0.2 m/s和0.3 rad/s阈值；log step 30的最终文件回放逐段读数与这两个最差值一致。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "推断：结合完整视频的行进、转向及近似原地段与评分器的完整时域成功结果，整段指令跟踪完成事件在视频结束附近锁存；以已检查的最后一帧视频20.286 s为约略完成锚点，时间精度约0.5 s，不将agent工作时间或仿真20秒直接换算为视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.000–20.286 s的双视角中，Spot四条腿的足端交替落在平地上承重并推进、转向；视频12.571–16.143 s近似原地踏步时仍由足端支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.1001
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.2461
+  }
+ ],
+ "failure": null,
+ "summary_zh": "Spot完成唯一的“按指令行进”节点，在视频末尾约20.3 s达到整段完成事件，最终进度1/1。视频全程显示足端支撑下的行进、转向及近似原地段；评分器确认1000步、未跌倒，最差分段平面速度误差0.1001 m/s和偏航角速度误差0.2461 rad/s均达标，无失败断点。没有状态日志或独立参考录像，精确误差验收依评分器读数，视频完成时刻仅作约略定位。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标签与视频、评分器读数一致：单一目标节点在末帧附近完成，contact_actual 为腿/躯干支撑，无失败断点。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:15:55Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--pl68j2d/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--pl68j2d/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "94fbbd97463638627a236241681bb4a26d2924e1c33ac921e1eb7e11784ebc4f",
+   "video_sha256": "695995b3a4e9a3442e20fb9d036b74b7eea8eff23474e1e2a3fa9a42665b096c",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--rxnt3yr.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--rxnt3yr.json
new file mode 100644
index 0000000000000000000000000000000000000000..b5e02c6fe80152232035d058fa69e3084edb528e
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--rxnt3yr.json
@@ -0,0 +1,192 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-privileged--rxnt3yr",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 1,
+   "fell_at_step": 19.0,
+   "worst_lin_err": 0.792,
+   "worst_yaw_err": 0.6673,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 151,
+   "agent_resets": 164,
+   "agent_sim_steps": 58365,
+   "agent_replays": 4,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 1.472,
+   "peak_processes": 17,
+   "peak_disk_gb": 0.241,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--rxnt3yr/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "db4ee8c25f712feb43b1a80512fca93ffdf60035bca838de8ef222bca0f60789"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.429,
+      "log_step": null,
+      "text_zh": "观察：视频 0.000–0.286 s，Spot从足端支撑的站姿开始摆腿；视频 0.429 s机身明显倾斜，视频 0.571 s接近竖直，随后翻倒，未形成持续平地行进。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：通看视频全程的逐秒帧，并密看视频 0.000–5.857 s和19.286–20.286 s；起步翻倒后机器人一直未恢复正常足端支撑，末帧仍机体倒置、腿朝上，没有完成后续指令段的正常行进。"
+     },
+     {
+      "t": null,
+      "log_step": 86,
+      "text_zh": "观察：log step 86对最终交付轨迹的评分器回放报告steps=1000、horizon=1000、fell_at=19、tracking_ok=false；worst_lin_err=0.792 m/s、worst_yaw_err=0.6673 rad/s，与inputs/run.json的reward一致。推断：虽然动作行数完整，但跌倒历史及两项超阈值误差使整段“按指令行进”事件从未达成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.000–0.286 s，Spot的足端接触平地承重，腿部抬落尝试启动行进；随后机体失去正常足端支撑并翻倒。这里只标可见的起步足地支撑，不把倒置后被画面遮挡的机身位置当作额外接触。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.792
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.6673
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 0.429,
+   "log_step": 86,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 0.286–0.714 s，正在摆腿起步的Spot迅速倾斜并翻倒，之后直至视频 20.286 s未恢复正常行进；log step 86对最终文件的回放报告首次跌倒控制步为19。断点锚取可见明显失稳的约视频 0.43 s，失败发生于“按指令行进”操作中。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 0.429,
+    "log_step": 86,
+    "evidence_zh": "观察：视频 0.429 s机身已大幅倾斜，视频 0.571 s接近竖直，视频 0.714–1.000 s继续翻转成腿朝上的姿态；视频 19.286–20.286 s仍未恢复站姿。最终文件的评分器回放在log step 86报告fell_at=19，reward亦报告fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志",
+    "evidence_zh": "推断：agent生成的起步步态控制未能稳定机身。log step 26记录生成并保存的轨迹已在控制步19跌倒，误差为0.792 m/s和0.6673 rad/s；log step 27直接读取该文件逐行诊断，起步已有负向前进速度、偏大的偏航和关节目标跟踪误差。log step 86最终回放复现相同跌倒步及误差，且视频 0.286–0.714 s呈现相应失稳翻倒，支持控制层归因；尚不能确定是落足几何、逆解还是反馈参数中的哪一项主导。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志；7 fps视频不能精确定位IMU上轴首次越过0.85阈值的瞬间，倾倒后部分跟随视角还被地面遮挡，但另一视角及末秒帧足以确认未恢复。精确误差与跌倒历史采用评分器字段；日志后续测试多次更换控制器，不能把其具体诊断直接归于交付轨迹，因此根因仅标“控制”。"
+ },
+ "summary_zh": "Spot尝试足端支撑起步，但约在视频 0.43 s明显失稳并翻倒，此后未恢复，唯一节点“按指令行进”始终未达成，最终为0/1。交付轨迹虽有1000个动作，评分器仍报告控制步19跌倒，最差平面速度误差0.792 m/s及偏航误差0.6673 rad/s均超阈值；根因可定位到起步控制失稳。没有逐帧状态日志，具体控制机制不能进一步确定；仅采用log step 26保存、log step 86复核的交付轨迹，不将后续会话测试计为进度。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标签与视频、log一致：Spot起步后约视频0.4–1.0 s翻倒且未恢复，节点未达成，断点、平衡现象和控制根因均有评分器读数（fell_at=19，误差0.792/0.6673）支持。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:13:44Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--rxnt3yr/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--rxnt3yr/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "db2c57536e87a1b76cadabff438724f968171f2ce343c5f0d3b7d3a67e77570a",
+   "video_sha256": "db4ee8c25f712feb43b1a80512fca93ffdf60035bca838de8ef222bca0f60789",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--ttzpdkw.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--ttzpdkw.json
new file mode 100644
index 0000000000000000000000000000000000000000..b3a98699c81368b018c38418855fe0e6cf190ce1
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--ttzpdkw.json
@@ -0,0 +1,173 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-privileged--ttzpdkw",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.1049,
+   "worst_yaw_err": 0.251,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 42,
+   "agent_resets": 438,
+   "agent_sim_steps": 429100,
+   "agent_replays": 7,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.49,
+   "peak_processes": 7,
+   "peak_disk_gb": 0.25,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--ttzpdkw/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "f9972b600c3412d13535b3d0a7872f1ac159ee3f5fae1e3639953353519b79fc"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [
+     [
+      20.286,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 4.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0–4.0 s 的双视图中，Spot足端交替落在格线平地上，机体发生平移并明显改变朝向；视频 4.1–8.9 s 继续踏步、平移并转回另一朝向。"
+     },
+     {
+      "t": 13.0,
+      "log_step": null,
+      "text_zh": "观察：视频 9.0–12.1 s 中Spot继续踏步行进；视频 12.6–16.1 s 中足端基本固定、机体近似原地站立，视频 16.4–20.286 s 再次踏步平移并转向。视频 19.0–20.286 s 的逐帧检查中机身均保持离地、足端承重，未见倾倒。"
+     },
+     {
+      "t": null,
+      "log_step": 58,
+      "text_zh": "观察：log step 58 中agent对最终交付trajectory.npz调用评分器回放两次，均报告steps=horizon=n_actions=1000、success=true、tracking_ok=true、fell_at=null，末态哈希均为5e852a160d36；五段线速度误差依次为0.1049、0.0724、0.0891、0.0103、0.0946 m/s，偏航误差依次为0.2510、0.1081、0.1558、0.0032、0.2063 rad/s。这些最终回放的最差误差与inputs/run.json的reward一致，区别于此前调试回合。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 58,
+      "text_zh": "推断：评分器reward确认交付轨迹完整执行1000步、未跌倒且最差段误差0.1049 m/s及0.251 rad/s均低于0.2 m/s及0.3 rad/s；结合视频末帧20.286 s仍由足端支撑并完成末段行进，将整段完成事件锚定在该末帧（约视频20.3 s）。此时刻是视频末端的近似完成锚点，未把agent工作时间换算成视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.0–20.286 s中，Spot四条腿的足端交替或同时接触格线平地，支承机身并完成平移、转向；视频约12.6–16.1 s近似站定时，足端继续支撑地面。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.1049
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.251
+  }
+ ],
+ "failure": null,
+ "summary_zh": "Spot完成唯一节点“按指令行进”，在视频约20.3 s的末端记为达成；视频可见平移、转向、近似原地停留及再次行进，足端持续承担地面支撑。最终评分器确认1000步全部完成、未跌倒，最差线速度误差0.1049 m/s、最差偏航误差0.251 rad/s均达标，最终两次回放结果和末态哈希一致。没有状态日志，精确分段误差及倾斜阈值依赖评分器读数，视频完成锚点约有半秒精度。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "单节点成功运行，节点达成、末帧锚点、contact_actual（腿/躯干 支撑）与评分器读数（worst_lin 0.1049、yaw 0.251、未跌倒）一致，未发现需修改之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:18:35Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--ttzpdkw/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--ttzpdkw/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "1f86ce37912f1b6f70aed505a548e0948aab8b956992e0ff92055a834d5f44fe",
+   "video_sha256": "f9972b600c3412d13535b3d0a7872f1ac159ee3f5fae1e3639953353519b79fc",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--xtguijl.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--xtguijl.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdde5df3e1501b8e9c30c4eb604af5f615e90181
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--xtguijl.json
@@ -0,0 +1,178 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-privileged--xtguijl",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.1287,
+   "worst_yaw_err": 0.179,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 8,
+   "agent_resets": 11,
+   "agent_sim_steps": 30320,
+   "agent_replays": 3,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.365,
+   "peak_processes": 10,
+   "peak_disk_gb": 0.828,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--xtguijl/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "f92acb5a4fe89173816c53cdfd04f2ff98a4dcc70e4e9fb230de39274338c8b6"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [
+     [
+      20.286,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.000–4.143 s 的跟随与侧视画面中，Spot从四足站立开始交替抬足、落足，机身相对地面格线平移且朝向持续改变；视频 4.286–8.000 s 中转向趋势改变，四足继续支撑行进。"
+     },
+     {
+      "t": 10.0,
+      "log_step": null,
+      "text_zh": "观察：视频 8.143–12.286 s 中，Spot继续交替踏步并相对地面格线平移，朝向变化较小；视频 12.857–16.286 s 中机身位置及朝向近似保持、足端仍有小幅踏步，未见翻倒。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：视频 16.714–20.286 s 中，Spot重新出现明显平移和转向；逐帧检查视频 19.429–20.286 s 的末段，机身始终由足端支撑且保持直立，末帧仍在踏步。"
+     },
+     {
+      "t": null,
+      "log_step": 14,
+      "text_zh": "观察：log step 14中，agent脚本重新生成交付轨迹，随后对同一交付文件执行两次独立CLI回放；两次工具结果均为steps=1000、horizon=1000、success=true、fell_at=null、tracking_ok=true，最终state_sha均为93e67989e5b3。本次run.json评分器字段另记success=1、replay_success=1、deterministic=1、n_actions=1000、fell=0、worst_lin_err=0.1287 m/s、worst_yaw_err=0.179 rad/s，两项误差分别低于0.2 m/s及0.3 rad/s阈值。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "推断：本次评分器确认完整时域的五段跟踪均达标且未跌倒，与视频中的行进、近似原地段和末段转向一致，因此整段完成事件在视频片尾约20.3 s达到，取最后可见帧视频 20.286 s作锚点；不是在某一瞬间速度吻合时提前完成。无逐步状态日志，完成时刻仅定位到片尾，未将agent工作时间换算成视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.000–20.286 s中，Spot四条腿的足端交替接触平地承重，抬足与落足配合产生平移和转向；视频 12.857–16.286 s近似原地段仍有小幅踏步支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.1287
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.179
+  }
+ ],
+ "failure": null,
+ "summary_zh": "Spot完成唯一节点“按指令行进”（终态1/1）：视频 0.000–20.286 s呈现连续行进、转向、零指令段近似原地踏步及末段再次转向，足端持续支撑，未见跌倒。本次评分器确认1000个控制步完整执行且回放成功，最差段平面速度误差0.1287 m/s、偏航角速度误差0.179 rad/s均达标，log step 14的两次独立交付回放最终哈希相同。证据缺口是没有逐步状态日志，视频不能精确测量逐段平均误差或倾倒阈值，数值验收依据评分器，完成时间仅定位为视频片尾约20.3 s。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "唯一节点按指令行进在视频片尾（约20.3 s）达到，与评分器success=1、1000步、未跌倒及log step 14的两次回放一致；时间、接触、证据与时钟均无错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:16:10Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--xtguijl/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--xtguijl/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "d46969488ad5776513fc7a6445dfca2d910be520559466dc6a4488cd0670fd8d",
+   "video_sha256": "f92acb5a4fe89173816c53cdfd04f2ff98a4dcc70e4e9fb230de39274338c8b6",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--zhwu8ov.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--zhwu8ov.json
new file mode 100644
index 0000000000000000000000000000000000000000..526dd8719e7bd0044a46b70456bf335b1a3363da
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-privileged--zhwu8ov.json
@@ -0,0 +1,183 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-privileged--zhwu8ov",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 1,
+  "reward": 1.0,
+  "source": "catalog",
+  "raw": {
+   "success": 1,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 1,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.163,
+   "worst_yaw_err": 0.2479,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 655,
+   "agent_resets": 2743,
+   "agent_sim_steps": 307565,
+   "agent_replays": 5,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--zhwu8ov/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "c3e006ba011eac8f44a9c6e0a4c7edfe9f5322ad6981f19e0ce6ab3349192f13"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 1,
+  "final": 1,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [
+     [
+      20.286,
+      null
+     ]
+    ],
+    "final": true,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.000–4.143 s 的跟随和侧面视图中，黄色Spot从站姿开始交替迈腿，足端落在平地，身体相对地面格线平移并持续改变朝向。"
+     },
+     {
+      "t": 4.143,
+      "log_step": null,
+      "text_zh": "观察：视频 4.143–8.286 s 中，Spot继续迈步，转向趋势改变，伴随横向位移；视频 8.286–12.286 s 中主要保持朝向并平移，未出现翻倒。"
+     },
+     {
+      "t": 13.0,
+      "log_step": null,
+      "text_zh": "观察：视频 12.571–16.286 s 中，Spot仍有小幅交替踏步及机身摆动，但相对附近地面格线近似原地停留；视频 16.571 s 后再次明显平移和转向。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：已检查视频全部143帧，视频 19.429–20.286 s 的末段各帧中，Spot继续迈腿转向，机身保持由足端支撑且未触地翻倒；最后可见帧为视频 20.286 s。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 148,
+      "text_zh": "观察：inputs/run.json 的本次评分器reward报告 success=1、replay_success=1、deterministic=1、n_actions=horizon=1000、fell=0、worst_lin_err=0.163 m/s、worst_yaw_err=0.2479 rad/s。log step 146记录agent脚本重新生成交付轨迹，log step 148记录基准回放CLI对该交付文件返回1000步、success=true、tracking_ok=true、fell_at=null及相同误差；这是交付文件的回放结果，与视频中的整段行进和未跌倒相符。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 148,
+      "text_zh": "推断：本次评分器确认整段1000步完成且各段误差达标，视频亦显示完整的行进、转向及近似原地阶段，因此将完成事件锚定在最后可见帧视频 20.286 s（时间精度约0.5秒）；不能从前面某一瞬间的迈步提前判完成。此时间取自视频片尾，不由agent工作时间换算。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.000–20.286 s 中，Spot四条腿的足端交替落在平地承重，支撑机身完成平移、转向和近似原地踏步；机身未见落地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.163
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.2479
+  }
+ ],
+ "failure": null,
+ "summary_zh": "本次交付轨迹完成“按指令行进”节点，片尾达到1/1：视频中Spot依次平移、转向、近似原地踏步后再次移动，全程未见跌倒。评分器确认1000步完整执行且未触发倾倒，最差段平均平面速度误差0.163 m/s、偏航角速度误差0.2479 rad/s均达标。没有逐帧状态日志，精确速度和倾斜边界依据本次评分器字段确认，完成时刻以视频 20.286 s 的最后可见帧作近似锚点。",
+ "conf": 9,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "成功运行，唯一目标节点 command_schedule_completed 在视频末帧 20.286 s 达成，与评分器 success=1、fell=0、1000 步及 log step 146/148 一致；contact_actual（腿/躯干·支撑）符合帧所见；无失败标签，无需修改。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:16:41Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--zhwu8ov/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-privileged--zhwu8ov/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "c5e88de099a924e0ed30e75c85fafda1caee68dd4a37c0ffd3b08c7e558642cb",
+   "video_sha256": "c3e006ba011eac8f44a9c6e0a4c7edfe9f5322ad6981f19e0ce6ab3349192f13",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--2vbjd6e.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--2vbjd6e.json
new file mode 100644
index 0000000000000000000000000000000000000000..781dba8005089638056f120b2dc53a46a69c3fa4
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--2vbjd6e.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-standard--2vbjd6e",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "graded_by_live": 1,
+   "replays": 0,
+   "finalized": 1,
+   "missing_log": 0,
+   "sim_error": 0,
+   "grader_error": 0,
+   "infra_error": 0,
+   "n_actions": 193,
+   "sim_steps": 193,
+   "physics_steps": 965,
+   "steps": 812,
+   "elapsed_s": 1757.2,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.3932,
+   "worst_yaw_err": 0.7872,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "instance_mismatch": 0,
+   "no_verdict": 0,
+   "gave_up": 0,
+   "sim_crashed_on_call": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 7,
+   "reset_budget": 10,
+   "episodes": 8,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.317,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.239,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-standard--2vbjd6e/replay.mp4",
+  "duration_s": 3.88,
+  "fps": 25.0,
+  "frames": 97,
+  "sha256": "e6aba5ad364435744167ae0a388aac33971c76f08b44306de94dea84742df84b"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00 s，Spot以四足站立于平地；视频 0.60–3.60 s，机器人抬腿踏步、平移并转向，已开始指令跟踪，但这不满足完整执行20秒指令的完成条件。"
+     },
+     {
+      "t": 3.84,
+      "log_step": 25,
+      "text_zh": "观察：视频 3.76–3.84 s，Spot前端下沉、后半身抬起，后足离地，机身明显前倾，录像随即结束；log step 25 的最后一次复位后运行报告服务终止原因为 fell、控制步为193，且其后log step 26只有诊断读取，没有再次复位或动作。评分器记录sim_steps=193、fell=1、success=0，完整1000步未执行。"
+     },
+     {
+      "t": 3.84,
+      "log_step": 25,
+      "text_zh": "推断：任务控制周期为0.02秒，最后回合193个控制步对应仿真 3.86 s，与视频 3.84 s 的末帧前倾及视频总长3.88秒吻合，因此录像是log step 25的最后回合，不包含此前复位掉的调试回合；唯一完成事件全程未达成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–3.84 s，Spot以四条腿的足端交替接触平地承重并行进、转向；末段前足支撑、后足抬离地面，机身前倾。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.3932
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.7872
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 3.76,
+   "log_step": 25,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 3.76–3.84 s，正在行进转向的Spot机身前端下倾、后半身上抬，未恢复就结束；log step 25报告服务在控制步193以fell终止，评分器亦记录fell=1。机器人仍处于第一段指令，未完成整个指令跟踪步骤。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 3.76,
+    "log_step": 25,
+    "evidence_zh": "观察：视频 3.76–3.84 s，两视角均显示Spot后半身迅速抬高、机身前倾，末帧前足承重而后足悬空；log step 25报告该回合在控制步193触发fell。录像结束前机身尚未整体落地，但已失衡，评分器fell=1证实倾倒终止判据被触发。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "推断：log step 19的脚本通过速度、陀螺仪和重力方向反馈生成足端目标及关节位置，log step 24加入速度与偏航误差积分，log step 25用该控制器复位后执行最后回合；视频 0.60–2.40 s机器人踏步转向时机身反复摆动，视频 3.76–3.84 s最终前倾失衡。该控制器未能在跟踪指令时稳定机身；评分器同时给出线速度误差0.3932米/秒及偏航误差0.7872弧度/秒。现有证据不足以进一步确定是落足布局、轨迹、反馈律还是具体增益导致。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志或接触力记录，无法独立复算逐段误差、精确倾斜阈值或定位具体失稳机制；保留控制大类。log step 25中agent脚本自算误差为0.3852、0.7472，与评分器0.3932、0.7872不同，原生指标采用评分器字段。最后回合身份由193控制步、末段可见前倾及后续无复位共同确认，未把此前调试回合计入节点进展。"
+ },
+ "summary_zh": "Spot在平地以足端支撑踏步、平移并转向，但唯一节点“按指令行进”未完成，最终为0/1。视频 3.76–3.84 s机器人前倾失衡，最后回合在193/1000控制步触发倾倒而终止，两项评分器跟踪误差也超标，根因保留为控制。无逐帧状态及接触力日志，不能进一步确定具体失稳机制；此前复位掉的调试回合不计入进展。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标签与视频、日志一致：唯一节点未达成，末段（视频 3.6–3.84 s）前倾失衡，log step 25 报告控制步193 fell，193×0.02=3.86 s 与视频时长 3.88 s 吻合；断点、平衡现象和控制根因证据有支撑且标明为推断。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:25:34Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--2vbjd6e/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--2vbjd6e/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "52411113e0b32611bbda6467465071ae8f87ff1c53548c524682d3e11908e0b6",
+   "video_sha256": "e6aba5ad364435744167ae0a388aac33971c76f08b44306de94dea84742df84b",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--ftqfu7c.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--ftqfu7c.json
new file mode 100644
index 0000000000000000000000000000000000000000..42ddb062bb382616a5d8030845e4cd59f9208b4c
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--ftqfu7c.json
@@ -0,0 +1,200 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-standard--ftqfu7c",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 57,
+   "sim_steps": 57,
+   "physics_steps": 285,
+   "steps": 966,
+   "elapsed_s": 1311.3,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.8551,
+   "worst_yaw_err": 0.7024,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-standard--ftqfu7c/replay.mp4",
+  "duration_s": 1.16,
+  "fps": 25.0,
+  "frames": 29,
+  "sha256": "3f246ab40bdae6b2c86f60436d30d1aaf1bbee280ef4faed9e6b32c052d3f3b2"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 77,
+      "text_zh": "观察：视频 0.00–0.84 s 的跟随与侧视画面显示黄色Spot足端接触平地，腿部活动并伴随机身起伏；log step 77 在最后一次复位后执行行进控制器，服务随后返回 fell。"
+     },
+     {
+      "t": 1.12,
+      "log_step": 83,
+      "text_zh": "观察：视频 0.92–1.12 s 中Spot机身先下俯再剧烈抬起、侧倾，录像随即结束；log step 83 的服务最终状态为 episode=11、control_steps=57、done=True、reason=fell。评分器 sim_steps=57、fell=1，而节点要求完整1000个控制步，故按指令行进事件从未完成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–0.84 s 中Spot四条腿的足端在平地承重，腿部伸屈并有足端抬起；视频 1.08–1.12 s 机身剧烈倾斜、足端离地，失去稳定支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.8551
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.7024
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 1.08,
+   "log_step": 77,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 0.92–1.12 s 中正在执行行进控制的Spot机身摆动扩大，最终明显抬起并侧倾，未恢复稳定站姿就结束；log step 77 的最后实时回合返回 fell，log step 83 确认第11回合仅执行57个控制步。失败发生在按指令行进过程中。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 1.08,
+    "log_step": 78,
+    "evidence_zh": "观察：视频 0.92、1.00、1.08、1.12 s 显示Spot机身由下俯转为剧烈上抬和侧倾，足端失去稳定支撑；log step 78 读取的该回合末条服务传感记录为四足接触标志全0、重力方向分量[-0.416,-0.375,-0.828]，评分器 fell=1。录像止于倾倒触发阶段，未显示机身砸地。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "log_step": 76,
+    "evidence_zh": "推断：姿态反馈控制未能稳定机身。log step 75–76 显示agent将IMU重力方向与角速度反馈转为各足高度修正，使用 k_att=0.5、k_d=0.10；log step 36 的主循环每批生成10个0.02秒动作后才更新传感反馈，log step 77 用该控制器执行最后回合，log step 78 的末段重力方向由接近竖直发展为[-0.416,-0.375,-0.828]，与视频 0.92–1.12 s 的机身摆动扩大、失衡相符。log step 79 降低增益后仅做离线计算，不能证明改动修复了评分回合；现有证据不把失稳进一步归为某个已证实错误的参数。",
+    "source_zh": "日志 / 视频"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志，无法精确分离姿态增益、反馈批次延迟与足端轨迹对失稳的贡献。agent在log step 79、84 将失败归因于增益过大，但修改后未再进行实时回合，故该具体解释未验证；log step 78 打印的时间由批次编号生成，未用于视频定位。倾倒时刻取自视频 1.08–1.12 s 画面，录像未继续显示落地结果。"
+ },
+ "summary_zh": "评分的第11回合中，Spot开始以足端支撑并尝试行进，但仅运行57个控制步，唯一节点“按指令行进”未完成，最终进度0/1。视频 0.92–1.12 s 显示机身摆动扩大并失衡，服务以fell终止；日志支持姿态反馈控制失稳，两个原生跟踪误差也均超过阈值。缺少逐帧状态与改参后的实时验证，不能确认某个具体增益是唯一原因，终止后的脚本修改不计为修复。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "单节点未达成、断点、平衡现象与反馈控制根因均有视频和日志支持，时间均在视频范围内，未发现需修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:14:51Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--ftqfu7c/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--ftqfu7c/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "0e161a1bcbb13c81dcb5825b8f97351ac7849995b1e042bc675f56d0ebe72d2b",
+   "video_sha256": "3f246ab40bdae6b2c86f60436d30d1aaf1bbee280ef4faed9e6b32c052d3f3b2",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--mnzgnf4.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--mnzgnf4.json
new file mode 100644
index 0000000000000000000000000000000000000000..d39f5a2fea16caa1a532e7afeac7304b1136faca
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--mnzgnf4.json
@@ -0,0 +1,205 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-standard--mnzgnf4",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 337,
+   "sim_steps": 337,
+   "physics_steps": 1685,
+   "steps": 337,
+   "elapsed_s": 547.1,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.7203,
+   "worst_yaw_err": 0.6003,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 0,
+   "reset_budget": 10,
+   "episodes": 1,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.186,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.261,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-standard--mnzgnf4/replay.mp4",
+  "duration_s": 6.647,
+  "fps": 17.0,
+  "frames": 113,
+  "sha256": "74710f9acdf7f53d1681c2be736405083e0fd7d6cbf70488949985a209e08962"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 9,
+      "text_zh": "观察：视频0.000 s中Spot以四足支撑平地；视频0.471–5.647 s中机器人交替抬脚、踏地并改变朝向，但未出现整段20秒跟踪完成。log step 9的服务足端接触读数为四足均接触。"
+     },
+     {
+      "t": 6.588,
+      "log_step": 20,
+      "text_zh": "观察：视频6.235–6.588 s中Spot机身越来越侧倾，末帧仍处于失稳运动；log step 20的实时服务数据经脚本汇总显示control step 337时up_z最低为0.841，随后回合停止。评分器记录steps=337、fell=1、success=0，未达到节点所需1000个控制步。"
+     },
+     {
+      "t": null,
+      "log_step": 54,
+      "text_zh": "观察：log step 54中机器人服务报告episode=1、control_steps=337、done=True、reason=fell；评分器记录resets=0、resumes=0、live_matches_replay=1。log step 26之后的本地模拟实验没有替换这一已结束的实时回合。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.000–6.588 s中，Spot四条腿的足端先共同支撑、随后交替触地承重与踏步；末段机身侧倾、足端支撑失稳。接触对象为平地，未见机身已经落地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.7203,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.6003,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 6.4,
+   "log_step": 20,
+   "where": "操作中",
+   "evidence_zh": "观察：视频6.235–6.588 s中，Spot在踏步行进时机身持续侧倾，未恢复稳定；log step 20中实时回合在control step 337停止，up_z最低0.841，log step 54的服务状态明确为done=True、reason=fell。机器人正在执行按指令行进，整段1000步未完成。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 6.4,
+    "log_step": 20,
+    "evidence_zh": "观察：视频6.235–6.588 s中Spot机身侧倾逐渐加剧，末帧两侧腿姿态明显不对称、机身向一侧翻倾；log step 20的实时服务重力方向数据经脚本汇总得到up_z从control step 316时的0.984降至control step 337时最低0.841，评分器fell=1。此处已触发倾倒判据，画面未显示躯干触地。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 20,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 20的agent脚本以周期足端轨迹、逆解和关节速度前馈生成动作，再按每批平均平面速度及偏航速度误差更新步幅输入；该脚本记录倾斜读数，但动作更新未使用机身姿态误差。实时服务数据在同一调用中显示速度跟踪误差持续存在、up_z最低从0.984降至0.841，并在control step 337触发倾倒；视频6.235–6.588 s也显示失稳加剧。因此行进控制未能维持支撑稳定，归为控制；现有证据不足以锁定具体控制子类或某个错误参数。"
+   }
+  ],
+  "gap_zh": "没有逐帧模拟状态日志，无法进一步区分足端滑移、支撑几何与姿态反馈等机制，也不能从画面复算逐段速度均值。视频时间直接取已检查帧，不把log中的control step或agent工作时间换算为视频时间；视频末帧为6.588 s，日志另记回合在control step 337结束。log step 55的模型过滤提示发生在log step 20已经倾倒结束、log step 54再次确认该终态之后，不能解释此前的物理失败。"
+ },
+ "summary_zh": "本次唯一实时回合中，Spot以足端支撑踏步并转向，但只执行337/1000个控制步，“按指令行进”始终未完成，节点最终为0/1。视频6.235–6.588 s可见机身侧倾加剧，服务及评分器确认倾倒；行进控制未能保持平衡，原生最差平移与偏航误差也均超阈值。没有逐帧状态日志，具体失稳机制无法细分；后续本地调参及回合结束后的模型过滤提示不改变本次结果。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "单节点未达成、断点与平衡/控制标注与视频、评分器读数（fell=1, steps=337）一致，未发现需修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:14:57Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--mnzgnf4/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--mnzgnf4/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "4a93670d828cd3690b6e99471b6f818a81428138b5cf5fe005797f51aebd0e06",
+   "video_sha256": "74710f9acdf7f53d1681c2be736405083e0fd7d6cbf70488949985a209e08962",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--nnhxg5r.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--nnhxg5r.json
new file mode 100644
index 0000000000000000000000000000000000000000..40a487f19671d35b3512e4c168f0a932e1f1a914
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--nnhxg5r.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-standard--nnhxg5r",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "graded_by_live": 1,
+   "replays": 0,
+   "finalized": 1,
+   "missing_log": 0,
+   "sim_error": 0,
+   "grader_error": 0,
+   "infra_error": 0,
+   "n_actions": 267,
+   "sim_steps": 267,
+   "physics_steps": 1335,
+   "steps": 2036,
+   "elapsed_s": 956.6,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.3537,
+   "worst_yaw_err": 1.4555,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "instance_mismatch": 0,
+   "no_verdict": 0,
+   "gave_up": 0,
+   "sim_crashed_on_call": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.379,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.251,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-standard--nnhxg5r/replay.mp4",
+  "duration_s": 5.36,
+  "fps": 25.0,
+  "frames": 134,
+  "sha256": "10a8d63ebcb255c9c506b24e601e8197f8ae5f48fd3aceaf2ee74244b8ffef80"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 120,
+      "text_zh": "观察：视频0.00–4.00 s的跟随和侧视画面显示黄色Spot足端交替支撑平地，机器人持续踏步、平移和转向；log step 120启动复位后的Episode 11，log step 123记录该回合从control_steps=0开始。此前十个回合是会话测试，不计入本次节点进度。"
+     },
+     {
+      "t": 5.32,
+      "log_step": 126,
+      "text_zh": "观察：已检查视频4.32–5.32 s的密集帧及末帧，Spot仍在踏步，机身姿态起伏，录像在完成整段指令前结束；log step 126的机器人服务状态为episode=11、control_steps=267、done=True、reason=fell，评分器同时报告sim_steps=267、fell=1。"
+     },
+     {
+      "t": 5.32,
+      "log_step": 126,
+      "text_zh": "推断：log step 8给出控制频率50 Hz，log step 118的脚本使用dt=0.02 s；最后回合267步对应仿真5.34 s，与视频5.32 s末帧和5.36 s时长吻合，已在末段终止处核对。这不是agent工作时间映射；该回合不足1000步且触发倾倒，因此整段行进事件从未完成。评分器worst_lin_err=0.3537 m/s、worst_yaw_err=1.4555 rad/s也分别超过0.2 m/s和0.3 rad/s阈值。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00–5.32 s，Spot四条腿的足端交替落在平地上承重，机器人一边踏步一边平移、转向；末段仍有足端支撑，未见躯干触地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.3537,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 1.4555,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 5.28,
+   "log_step": 126,
+   "where": "操作中",
+   "evidence_zh": "观察：视频5.04–5.32 s机器人仍在执行踏步行进，机身姿态起伏后录像结束，尚未完成后续指令；log step 126的服务状态明确报告Episode 11在267个控制步以fell结束，评分器fell=1。推断：按267步×0.02 s与视频末段核对，断点锚定视频约5.3 s的行进操作中。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 5.28,
+    "log_step": 126,
+    "evidence_zh": "观察：视频4.32–5.32 s的两路视角可见Spot持续踏步、机身起伏，最后一帧未见躯干倒地；log step 126的机器人服务终止原因为fell，评分器也报告fell=1。推断：结合DAG规定的机身上轴竖直分量低于0.85即判倾倒，属于行进时失衡，不能把尚未出现躯干触地理解为没有触发倾倒。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 118显示agent用周期性关节位置脚本驱动四足行进；log step 123中同一最终回合在control step 108的脚本传感汇总显示实际偏航角速度为-0.908 rad/s，而指令为+0.609 rad/s，已出现反向跟踪。该回合随后在log step 126被服务以fell终止，评分器还报告两项跟踪误差超阈值，支持行进控制未能兼顾指令跟踪与姿态稳定；现有记录不足以将失稳归到某个具体增益、足端接触设计或反馈环节。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志；视频末段可见姿态起伏，但不足以量出触发倾倒的精确角度，倾倒判定依据最终实时回合的服务reason=fell及评分器fell=1。最终脚本在log step 118的部分代码被日志截断，且缺少失稳前逐步姿态、足端接触力及关节跟踪记录，因此根因只标控制，不细分具体机制。",
+  "unmet": {
+   "command_schedule_completed": false
+  }
+ },
+ "summary_zh": "评分回合为最后的Episode 11：Spot已用足端支撑平地并踏步、平移、转向，但仅执行267/1000个控制步，唯一行进节点未完成，最终0/1。视频约5.3 s处的行进因服务判定倾倒而结束，两项原生跟踪误差也超阈值，根因支持到控制层。视频未显示躯干触地，且没有逐帧状态及接触力记录，无法确定具体失稳机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标签与视频、日志一致：最终回合Episode 11在267步被服务判fell，视频末帧机身明显倾斜压低，节点未完成、断点、平衡现象和控制根因均有证据支持。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:12:17Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--nnhxg5r/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--nnhxg5r/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "5579b9b0c966d83ce36d53ecae56e40393ba8c27451cf71d8af0554516fbdb52",
+   "video_sha256": "10a8d63ebcb255c9c506b24e601e8197f8ae5f48fd3aceaf2ee74244b8ffef80",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--obbgswv.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--obbgswv.json
new file mode 100644
index 0000000000000000000000000000000000000000..b708e4e1d18ca6f924d077a6cbf8a7b7b2629da3
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--obbgswv.json
@@ -0,0 +1,216 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-standard--obbgswv",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 967,
+   "sim_steps": 967,
+   "physics_steps": 4835,
+   "steps": 4847,
+   "elapsed_s": 1751.1,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.7541,
+   "worst_yaw_err": 1.2796,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 8,
+   "reset_budget": 10,
+   "episodes": 9,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-standard--obbgswv/replay.mp4",
+  "duration_s": 19.857,
+  "fps": 7.0,
+  "frames": 139,
+  "sha256": "0b411b953bde5cb0bdc372f79c7bac9f8bee1371d44749ba70bd1c6019f80b10"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 3.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.0–4.0 s，Spot四足反复踏步、机身降低，朝向只小幅改变，地面格线相对移动很少；未表现出首段持续前进并转向所需的运动。视频8.0–12.0 s亦主要低姿态踏步，视频13.0–16.0 s近似停留。"
+     },
+     {
+      "t": 19.714,
+      "log_step": 158,
+      "text_zh": "观察：视频19.571 s至19.714 s，Spot机身突然侧倾并下沉，足腿支撑姿态破坏，录像随即结束。log step 158读取最后一次运行的服务返回：episode=9、control_steps=967、sim_time_s=19.34、done=true、reason=fell；该回合与reward的967步、fell=1一致。仿真19.34 s和视频19.714 s分别保留，不以agent时间换算。"
+     },
+     {
+      "t": 19.714,
+      "log_step": null,
+      "text_zh": "观察：评分器reward报告worst_lin_err=0.7541 m/s、worst_yaw_err=1.2796 rad/s，均超过DAG阈值0.2 m/s与0.3 rad/s；n_actions=967，小于要求的1000控制步。"
+     },
+     {
+      "t": 19.714,
+      "log_step": 158,
+      "text_zh": "推断：结合最终回合的服务终止记录、视频末段倾斜和原生误差，完整指令跟踪事件从未完成；中途踏步或零指令段停留不能单独使此事件成立。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.0–19.571 s，Spot的四条腿足端交替或同时接触平地并承重，机器人踏步和小幅转动；视频19.714 s支撑姿态失稳。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.7541
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 1.2796
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 19.714,
+   "log_step": 158,
+   "where": "操作中",
+   "evidence_zh": "观察：Spot正在执行最后一段行进指令，视频19.571–19.714 s机身突发侧倾、下沉并结束；log step 158显示最后评分回合episode=9在967控制步以reason=fell终止，尚未跑满1000步。评分器同时报告两项跟踪误差超标，按指令行进始终未完成。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 3.0,
+    "log_step": null,
+    "evidence_zh": "观察：视频0.0–4.0 s，Spot虽然持续抬落足、机身起伏，却只产生很少的格线相对位移和朝向变化；视频8.0–12.0 s也主要踏步，未形成要求的持续前进。推断：这是行进推进和转向不足，区别于末帧的失衡；评分器worst_lin_err=0.7541 m/s、worst_yaw_err=1.2796 rad/s为跟踪不达标的独立读数。"
+   },
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 19.714,
+    "log_step": 158,
+    "evidence_zh": "观察：视频19.714 s相比视频19.571 s，Spot机身明显向侧面倾斜、下沉，足腿不能维持此前支撑姿态；log step 158的episode=9服务状态为done=true、reason=fell，评分器fell=1。这里标失衡，未把倾倒重复记为违反约束，也未声称机器人已完全翻倒或躯干触地。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志 / 视频 / 评分器字段",
+    "evidence_zh": "推断：agent在log step 155恢复关节位置步态及IMU姿态修正参数后运行最终回合；log step 158的服务返回确认动作实际执行、ok=true，但末批右前腿外展目标被限位裁剪，随后以fell结束。结合视频0.0–12.0 s的低效踏步、视频19.714 s的失衡及评分器两项原生误差超标，可归于所用步态控制未实现所需运动和稳定性；证据不足以将裁剪认定为直接原因，或进一步区分接触设计、逆解、反馈及具体参数错误。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志，跟随视角及7 fps画面不能恢复逐段精确速度、足端接触力和失衡的具体控制机制；原生误差取评分器字段。log step 158的agent脚本自算统计随后出现索引异常，且其采样统计不等同于评分器口径，故不用于节点或原生指标判定；该异常在服务已返回fell后发生，不是本次物理失败的原因。"
+ },
+ "summary_zh": "最后评分回合中，Spot以足端支撑并持续尝试行进，但前段主要低姿态踏步，唯一的完整指令跟踪节点始终未完成，最终0/1。视频19.714 s机身侧倾下沉，服务在967/1000控制步以fell终止；原生最差线速度误差0.7541 m/s、偏航角速度误差1.2796 rad/s均超标，支持控制失败。没有逐帧状态日志，无法进一步定位接触、逆解或反馈机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "唯一节点未完成、断点在最后一段指令行进中、末帧倾倒（平衡）与前段推进不足（步态）、根因停在“控制”均有视频、log step 158服务返回及评分器字段支持；未发现需要修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:20:01Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--obbgswv/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--obbgswv/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "054e74b655a256e050108c7145979fac197a80f175eb4ac120447095baa6c0ce",
+   "video_sha256": "0b411b953bde5cb0bdc372f79c7bac9f8bee1371d44749ba70bd1c6019f80b10",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--yxsfyxa.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--yxsfyxa.json
new file mode 100644
index 0000000000000000000000000000000000000000..81ac8f42e1ab209b0b5ae6e8faae99375ee4084c
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--yxsfyxa.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-standard--yxsfyxa",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 1000,
+   "sim_steps": 1000,
+   "physics_steps": 5000,
+   "steps": 6415,
+   "elapsed_s": 1156.7,
+   "video_rendered": 1,
+   "fell": 0,
+   "worst_lin_err": 0.1249,
+   "worst_yaw_err": 0.5125,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.152,
+   "peak_processes": 12,
+   "peak_disk_gb": 0.239,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-standard--yxsfyxa/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "bf46c75e2db9815283e9204bb554e1b06fc9c3241fdea6e724f9e4dd092557db"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 1.0,
+      "log_step": null,
+      "text_zh": "观察：视频0.00–4.00 s中，黄色Spot从站立开始交替抬脚，机体相对地面格线平移并转向；视频4.14–8.00 s中转向反向，视频8.14–12.00 s中继续平移，视频12.43–16.00 s中近似原地踏步，视频16.57–20.29 s中再次明显平移和转向。整段并非原地不动。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 21,
+      "text_zh": "观察：视频19.43–20.29 s逐帧可见Spot继续以足端支撑地面，躯干未倒地；log step 21的实时服务状态为episode=11、control_steps=1000、done=True、reason=horizon、resets_left=0。任务文本规定只评分最后一个回合，故此前测试回合的跌倒不属于本录像。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：inputs/run.json的评分器字段为success=0、n_actions=1000、fell=0、worst_lin_err=0.1249 m/s、worst_yaw_err=0.5125 rad/s、live_matches_replay=1。推断：平面速度误差满足0.2 m/s阈值，但偏航角速度误差超过0.3 rad/s阈值；该节点要求整段各段跟踪均达标，因此尽管跑满时域且未跌倒，完成事件始终未成立。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00–20.29 s中，Spot四条腿的足端交替接触平地并承重，换腿踏步使机体平移与转向；视频12.43–16.00 s的近似原地阶段仍以足端支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.1249,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.5125,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 12.714,
+   "log_step": 21,
+   "where": "操作中",
+   "evidence_zh": "观察：视频12.43–14.00 s中，Spot在近似原地踏步时机身仍随换腿小幅往复扭摆；视频19.43–20.29 s中它仍站立并继续行进。log step 21的服务状态确认最后回合跑满1000个控制步，评分器记录fell=0而worst_yaw_err=0.5125 rad/s超过0.3 rad/s阈值。推断：未修复的失败位于整段按指令行进步骤的转向跟踪；视频12.71 s是可见扭摆的导航锚点，不代表首次数值超阈值的时刻。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 12.714,
+    "log_step": 23,
+    "evidence_zh": "观察：视频12.43–14.00 s的近似原地段中，Spot反复换腿，躯干朝向随步伐小幅来回摆动而未翻倒；视频19.43–20.29 s仍有足端支撑。agent在log step 23自述最后回合的偏航角速度随步周期振荡约±0.7 rad/s，评分器同时给出worst_yaw_err=0.5125 rad/s、fell=0。推断：结合角速度误差读数，失误是步行中转向不够平稳、未达到指令跟踪要求；不能仅凭小幅踏步本身判失败。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "evidence_zh": "观察：log step 5的控制脚本按给定速度和偏航指令计算足端运动，并以本体传感器反馈修正；log step 16保留该方案并修改支撑/摆动足轨迹；log step 21以T=0.35、H=0.09运行最后回合，脚本自算的五段偏航误差均超过0.3 rad/s。agent在log step 23自述每0.1 s反馈一次且偏航随步周期振荡。视频12.43–14.00 s可见换腿时躯干小幅扭摆，评分器确认最差段偏航误差0.5125 rad/s。推断：脚本已按正确指令执行行进，但未抑制步行中的转向波动，支持控制能力不足；现有证据不能进一步区分接触设计、轨迹生成、反馈或某个参数的问题。",
+    "source_zh": "日志 / agent 自述 / 视频 / 评分器字段"
+   }
+  ],
+  "gap_zh": "没有逐控制步状态日志或评分器的逐段误差表，7 fps跟随/侧视录像只能显示小幅扭摆，不能精确重建偏航角速度及最早超差时刻。log step 21的分段数值是agent脚本对反馈块均值的自算结果，不等同于评分器逐步统计；log step 23的振荡幅度也只是agent自述。故以评分器最终指标判定节点失败，断点时间仅作导航，根因停在控制而不指定参数或反馈机制。",
+  "unmet": {
+   "command_schedule_completed": false
+  }
+ },
+ "summary_zh": "Spot完成最后一个回合的1000个控制步，视频0.00–20.29 s可见平移、转向、近似原地踏步且未倒地，但唯一的整段按指令行进节点未达成，最终0/1。评分器的最差段平面速度误差0.1249 m/s达标，偏航角速度误差0.5125 rad/s超标；视频12.43–14.00 s可见换腿伴随的小幅扭摆，结合日志归为步态现象及控制根因。缺少逐步状态和精确分段评分，无法定位首次超差或确认具体控制机制。",
+ "conf": 7,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成、断点和步态/控制标注与评分器字段（yaw误差0.5125>0.3，fell=0，1000步）及视频一致，未发现需修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:12:45Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--yxsfyxa/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--yxsfyxa/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "70c6e0de202966ae3f566eb957891dbf81d103c72a8776b37613bf2ab5b64c54",
+   "video_sha256": "bf46c75e2db9815283e9204bb554e1b06fc9c3241fdea6e724f9e4dd092557db",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--zzufyks.json b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--zzufyks.json
new file mode 100644
index 0000000000000000000000000000000000000000..e39c13843ae69d0e3da9813e6ca6cab0ae9c0b65
--- /dev/null
+++ b/phase2/mujoco-playground/spot_flat_terrain_joystick/trajectories/mujoco_playground_spot_flat_terrain_joystick-standard--zzufyks.json
@@ -0,0 +1,206 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_flat_terrain_joystick-standard--zzufyks",
+ "taskKey": "mujoco-playground/spot_flat_terrain_joystick",
+ "dag_sha256": "7bf159e4a8c9bf216a30dc0e2689c4883a5a3c3913693c824ecc9ed37cc674b0",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 1000,
+   "sim_steps": 1000,
+   "physics_steps": 5000,
+   "steps": 8499,
+   "elapsed_s": 1410.4,
+   "video_rendered": 1,
+   "fell": 0,
+   "worst_lin_err": 0.6442,
+   "worst_yaw_err": 0.7678,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_flat_terrain_joystick-standard--zzufyks/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "c59cc559770f978452cfc98e5f180c865f3ec47d4d43926604cfa7703f6ac922"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "command_schedule_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 67,
+      "text_zh": "观察：视频 0.0 s Spot以四足站立；视频 1.0–12.0 s 可见交替迈腿、相对地面格线移动及机身转向，视频 13.0–16.4 s 近似原地停留。log step 67记录最后一次复位至episode 11、control_steps=0、resets_left=0；此前复位掉的测试局不计入本轨迹。"
+     },
+     {
+      "t": 17.429,
+      "log_step": 67,
+      "text_zh": "观察：视频 17.4–19.3 s Spot仍由足端支撑，但腿部动作变小、机身朝向来回摆动，未表现出持续的明显侧移转向。log step 67的末局末段指令为(vx,vy,wz)=(-0.281,-0.7415,-0.6673)，脚本打印的服务传感器批次均值在control step 860为vxy=(-0.098,-0.010)、wz=+1.184，在control step 870为vxy=(-0.080,+0.187)、wz=-0.009；这些控制步读数只作为末局证据，不直接换算视频时刻。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 68,
+      "text_zh": "观察：视频最后一秒至末帧（视频 20.286 s）Spot仍站立并有小幅迈腿；log step 68记录episode 11以control_steps=1000、reason=horizon结束。评分器报告n_actions=1000、fell=0、live_matches_replay=1、worst_lin_err=0.6442 m/s、worst_yaw_err=0.7678 rad/s。推断：虽跑完整段且未跌倒，两个最差段平均误差分别超过0.2 m/s和0.3 rad/s，故整段按指令行进事件从未完成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–20.286 s Spot用四条腿的足端接触平地承重，行进时交替抬足、落足，近似停留时多足驻地支撑；末段仍有小幅换步，未见躯干着地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.6442
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.7678
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "command_schedule_completed",
+   "t": 17.429,
+   "log_step": 67,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 17.4–19.3 s Spot在末段行进尝试中迈步幅度很小、朝向反复摆动，随后直到视频 20.286 s末帧仍未形成持续的明显侧移转向；log step 67和log step 68对应episode 11的末段速度读数持续偏离指令。评分器的worst_lin_err=0.6442 m/s、worst_yaw_err=0.7678 rad/s均超阈值，fell=0。推断：断在整段按指令行进，视频 17.429 s取作末段明显偏差的导航锚点，不代表已定位最早超标的指令段。"
+  },
+  "unmet": {
+   "command_schedule_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 17.429,
+    "log_step": 67,
+    "evidence_zh": "观察：视频 17.4–19.3 s Spot的足端大多驻地，仅小幅换步，机身朝向有往返摆动；视频 19.4–20.286 s继续小步调整，未倒下。log step 67的末段指令要求持续向机体右侧移动并负向偏航，而control step 860、865的横向速度分别为-0.010、+0.144 m/s，偏航速度分别为+1.184、+1.146 rad/s。推断：行进与转向跟踪不足，属于未跌倒的步态失败。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": 17.429,
+    "log_step": 66,
+    "source_zh": "日志",
+    "evidence_zh": "观察：log step 66给出的末局控制循环按预设joint_targets生成关节位置，末段只依据倾斜量增减motion_gain；读取的机体平面速度lin和偏航速度wz用于打印，所示循环没有按速度误差纠偏。log step 67、log step 68的episode 11传感器读数显示末段横向速度明显不足、部分批次反向偏航；视频 17.4–19.3 s可见对应的小步及朝向摆动。推断：该控制器未能将指令转换为稳定的机体速度跟踪，已有偏差也未被消除；不足以进一步区分足端轨迹、接触安排或反馈增益的问题，根因保留在控制大类。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志；现有日志中部分输出已截断，评分器只给最差段误差，未报告各段平均误差，因而不能确定最先不合格的指令段。跟随视角与侧视的低帧率视频能支持末段步态和转向不足，但不能独立测量精确速度、接触力或控制子类；断点时间为可见偏差锚点。"
+ },
+ "summary_zh": "评分局为log step 67复位后的episode 11，Spot完成1000个控制步，经历行走转向与近似停留，并在视频 20.286 s末帧保持站立，但整段按指令行进节点未完成（0/1）。视频 17.4–19.3 s可见末段碎步和转向不足，评分器报告最差段线速度误差0.6442 m/s、偏航误差0.7678 rad/s均超阈值，日志支持控制器未能消除速度跟踪偏差。缺少逐帧状态及完整逐段误差，无法定位最早不合格的指令段或进一步确定控制子类。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "评分局、未完成节点、断点（行进整段）、步态现象和控制根因均与评分器读数及视频一致，未发现需要修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:25:26Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--zzufyks/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_flat_terrain_joystick-standard--zzufyks/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "c5d66ef1af89e91aad781019fbd2d72aa48e7c143ac08393a362b05aa0dd4fd2",
+   "video_sha256": "c59cc559770f978452cfc98e5f180c865f3ec47d4d43926604cfa7703f6ac922",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--b8vf5hz.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--b8vf5hz.json
new file mode 100644
index 0000000000000000000000000000000000000000..5e5d247b708e6f9b08355550fb3ffd0146ff032f
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--b8vf5hz.json
@@ -0,0 +1,206 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-privileged--b8vf5hz",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 1,
+   "fell_at_step": 468.0,
+   "worst_lin_err": 1.3325,
+   "worst_yaw_err": 0.9441,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 48,
+   "agent_resets": 4605,
+   "agent_sim_steps": 1036815,
+   "agent_replays": 2,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.892,
+   "peak_processes": 11,
+   "peak_disk_gb": 0.241,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--b8vf5hz/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "34d0f71bdabf0453b454f4780c4bfde73de8fdf49e562f05c387fe00db2047ca"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.714,
+      "log_step": null,
+      "text_zh": "观察：视频 0.000–1.857 s 的跟随与侧面双视图中，Spot 从四足站立开始抬足交替踏步，足端落在方格地面上；视频 2–8 s 仍在行走，说明本节点已实际尝试，但完整日程尚未完成。"
+     },
+     {
+      "t": 9.571,
+      "log_step": null,
+      "text_zh": "观察：视频 9.286–9.429 s Spot 躯干俯仰倾斜逐渐增大，视频 9.571 s 已明显失稳，视频 9.714–10.143 s 翻转至四腿朝上；视频 19.286–20.286 s 的全部末尾帧仍为倒置，未恢复行进。"
+     },
+     {
+      "t": null,
+      "log_step": 90,
+      "text_zh": "观察：log step 90 对最终交付 trajectory.npz 的评分器 CLI 回放报告 instance_ok=true、steps=1000、success=false、fell_at=468、tracking_ok=false；各段平面速度误差依次为 0.5554、0.6225、1.3325、1.0035、0.8859 m/s，均超过 0.2 m/s，worst_yaw_err=0.9441 rad/s、worst_gait_err=0.0915 m 也超限。后二者与 native_metrics 对应，足高误差取此交付文件的评分器回放输出；reward 未单列 worst_gait_err。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 93,
+      "text_zh": "推断：log step 92 的两个后续试验均未满足保存条件，log step 93 再次回放原交付文件仍报告全部 1000 步和 fell_at=468，与 reward 的 deterministic=1、fell_at_step=468 及本次视频翻倒相符；因此不能采用调参中截到跌倒前的误差，也不能将执行满行数记为本节点完成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.000–8.857 s，Spot 四条腿的足端交替落在平坦方格地面，承重并推动机身行进；视频 9.429–10.143 s 该支撑失稳，机身翻转。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 1.3325,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.9441,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_gait_err",
+   "value": 0.0915,
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 9.571,
+   "log_step": 90,
+   "where": "操作中",
+   "evidence_zh": "观察：Spot 在视频 9.286–9.571 s 的行进中出现越来越大的躯干俯仰，视频 9.714–10.143 s 翻倒，直到视频 20.286 s 仍未恢复。log step 90 的最终交付文件回放和 reward 均确认控制步 468 触发跌倒判据，且速度跟踪超限；未完成的是“按指令行进”这一步。视频锚点来自逐帧观察，不将 agent 工作时间换算为视频时间。"
+  },
+  "unmet": {
+   "commanded_gait_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 9.571,
+    "log_step": 90,
+    "evidence_zh": "观察：视频 9.429 s Spot 躯干已明显倾斜，视频 9.571 s 倾斜进一步增大，视频 9.714 s 接近竖直翻转，视频 10.000–10.143 s 四腿朝上；后续直至视频 20.286 s 没有重新建立正常四足支撑。log step 90 和 reward 的跌倒字段与画面一致。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": 9.571,
+    "log_step": 90,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 61、70 的脚本修改显示 agent 用解析腿部逆解、相位摆腿、支撑足扫动和姿态/速度修正生成关节目标，log step 78、86 记录调参并保存候选；但 log step 90、93 对最终交付文件的回放仍报告跟踪不达标及控制步 468 跌倒，视频 9.429–10.143 s 也显示连续失稳翻倒。这支持控制器未能实现稳定的指令跟踪，无法进一步证明是某个增益、逆解或接触参数单独导致，故保留在“控制”。"
+   }
+  ],
+  "gap_zh": "没有逐步状态日志、IMU 曲线或足端接触力记录；7 fps 双视图存在步态采样混叠，不能从画面量化足高相位误差或确定失稳的具体控制子因。agent 在 log step 94 自述可能受摩擦限制，但没有排除其他控制因素的证据，未据此标注摩擦或任务环境根因。"
+ },
+ "summary_zh": "Spot 已实际抬足行进，但视频约 9.6 s 在换向后的行进过程中失稳翻倒，直到视频 20.286 s 仍倒置，“按指令行进”节点最终为 0/1。最终交付文件虽执行满 1000 个控制步，评分器回放仍确认跌倒且平面速度、偏航和足高误差均超限，根因保守归为控制。缺少逐步传感器和接触力证据，无法确定具体增益或摩擦机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标注与视频、log 和评分器字段一致：节点未达成，约视频 9.4–10 s 翻倒并保持倒置，断点、平衡现象、控制根因均有证据支持；未发现需要修改的内容。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:07:46Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--b8vf5hz/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--b8vf5hz/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "70665819db785c988bcb1ec158d6ef63fcb763bce3f5aa3ff0b93ec4bb1b1efa",
+   "video_sha256": "34d0f71bdabf0453b454f4780c4bfde73de8fdf49e562f05c387fe00db2047ca",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--goxqmft.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--goxqmft.json
new file mode 100644
index 0000000000000000000000000000000000000000..2484338590a734eb561cf4c5cddc493171fababd
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--goxqmft.json
@@ -0,0 +1,204 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-privileged--goxqmft",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.5005,
+   "worst_yaw_err": 0.8541,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 1027,
+   "agent_resets": 1033,
+   "agent_sim_steps": 2086685,
+   "agent_replays": 3,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 1.858,
+   "peak_processes": 17,
+   "peak_disk_gb": 0.241,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--goxqmft/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "27ac19a81aed6dfd17aad527d11471d38661c4b628490ee20e971e8bc6d9d79f"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.143,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00–8.29 s，黄色 Spot 从站立开始交替摆腿，足端轮流触地，机身在方格地面上移动并改变朝向；实际已开始行进。"
+     },
+     {
+      "t": 8.429,
+      "log_step": null,
+      "text_zh": "观察：视频 8.43–16.43 s，Spot 的腿反复大幅前后伸展，机身长轴在双视图中大致维持横向，转向很少；成功参考在对应后退区段明显连续转弯。视频 16.86 s 后机身再次前移，视频 19.43–20.29 s 仍有足端支撑且躯干未倒地。"
+     },
+     {
+      "t": null,
+      "log_step": 28,
+      "text_zh": "观察：log step 22 保存交付文件；log step 28 使用评分器回放命令检查该文件，报告 steps=1000、fell_at=null、tracking_ok=false。五段平面速度误差分别为 0.2377、0.3419、0.3194、0.4036、0.5005 m/s，均超过 0.2 m/s；后面三段偏航误差为 0.7324、0.8541、0.5055 rad/s，超过 0.3 rad/s；第4段足高误差为 0.051 m，超过 0.05 m。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 28,
+      "text_zh": "推断：评分器 reward 的 n_actions=1000、fell=0、deterministic=1、worst_lin_err=0.5005、worst_yaw_err=0.8541 与 log step 28 的交付回放检查一致，也与视频完整行进而未倒地相符；本节点要求全部五段验收通过，因此全程没有达到完成条件，终值为 false。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.14–20.29 s，Spot 四条腿的黑色足端交替接触方格地面，承托躯干并在摆腿与落脚过程中推进机身；视频 8.43–16.43 s 腿伸展幅度较大，躯干仍由足端支撑。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.5005
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.8541
+  },
+  {
+   "name": "worst_gait_err",
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better",
+   "value": 0.051
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 8.429,
+   "log_step": 28,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 8.43 s 起 Spot 在后退区段出现大幅伸腿，视频 9.00–16.43 s 机身转向有限，仍持续尝试行进；log step 28 对交付文件的回放检查报告五段速度误差全部超限，且后面三段偏航误差超限。推断：失败发生在“按指令行进”的执行中；视频 8.43 s 是可见偏差加重的导航锚点，缺少逐步读数，不能确定首次误差越限的时刻。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 8.429,
+    "log_step": 28,
+    "evidence_zh": "观察：视频 8.43–16.43 s，Spot 的足端反复大幅前后伸出，躯干保持离地但机身朝向变化很小；成功参考同一后退日程中明显连续转弯。视频 19.43–20.29 s 机器人仍站立行进。推断：结合 log step 28 的速度与偏航跟踪不达标，现象为行进和转向偏离指令的步态失败。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "source_zh": "日志",
+    "evidence_zh": "观察：log step 5 的轨迹生成脚本读取指令日程与 local_linvel，用速度误差修正落脚点并求腿部逆解；log step 11 加入平面速度反馈及偏航角速度误差反馈来分配支撑力，log step 20 改为分区段选取控制增益，log step 22 用该控制器保存轨迹。log step 28 的交付回放仍有持续速度及偏航误差，且最坏两项与最终 reward 一致。推断：生成交付轨迹的反馈控制未能使机身速度和转向跟踪收敛到任务容差内；现有证据支持反馈控制不足，不能定位到某个具体增益。"
+   }
+  ],
+  "gap_zh": "观察：inputs/ 没有逐帧状态日志；录像仅为 7 fps，不能从画面量化每步速度、偏航或足高相位误差，也无法精确定位最初偏离时刻。最坏足高误差 0.051 m 来自 log step 28 中评分器对交付文件的回放检查，reward 未单列该字段；其余最坏误差及未跌倒结果与 reward 相符。推断：具体反馈增益、足端几何和接触力对失败各自的贡献无法进一步分离。",
+  "unmet": {
+   "commanded_gait_completed": false
+  }
+ },
+ "summary_zh": "Spot 执行了完整 1000 控制步，足端交替支撑且未触发跌倒判据，但五段平面速度误差均超标，后面三段偏航误差超标，“按指令行进”最终完成 0/1。视频 8.43–16.43 s 的后退区段可见大幅伸腿及不足的转向，结合交付回放检查，失败归于步态偏离与反馈控制不足。没有逐帧状态日志，7 fps 录像无法量化精确相位或确定具体失效增益；足高最坏误差取自 log step 28 的交付回放检查。",
+ "conf": 7,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点 final=false、breakpoint、步态现象与反馈控制根因均有 log step 28 的评分器回放读数和 reward 的 worst 值支持；未发现需要修改的内容。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:08:05Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--goxqmft/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--goxqmft/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "49009d4df175aeb43e25765fe9324df2f625d63a0ca6fd95ca67b040f87ece97",
+   "video_sha256": "27ac19a81aed6dfd17aad527d11471d38661c4b628490ee20e971e8bc6d9d79f",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--hrxypef.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--hrxypef.json
new file mode 100644
index 0000000000000000000000000000000000000000..ebb30920e8806a1684b58dfa8c08878226509081
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--hrxypef.json
@@ -0,0 +1,184 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-privileged--hrxypef",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.2274,
+   "worst_yaw_err": 0.2216,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 44,
+   "agent_resets": 216,
+   "agent_sim_steps": 318655,
+   "agent_replays": 7,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 2.082,
+   "peak_processes": 13,
+   "peak_disk_gb": 0.249,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--hrxypef/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "b8d27dbd36c74207c6a557d785f7a66d15c072e132b0a5fb0590f697bf91e06e"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00–7.86 s 的跟随与侧面双视图中，Spot 四足反复抬起和落地，躯干随方格地面移动并转向，已开始执行行进；视频 7.86–8.86 s 及视频 15.86–16.86 s 的换向附近也持续迈步。"
+     },
+     {
+      "t": 20.286,
+      "log_step": null,
+      "text_zh": "观察：逐帧检查视频 0.00–20.286 s，Spot 持续行进，未见倒地；视频 19.43–20.286 s 的最后一秒仍由足端支撑。inputs/run.json 的 reward 报告 n_actions=1000、horizon=1000、fell=0、deterministic=1、replay_success=0、worst_lin_err=0.2274 m/s、worst_yaw_err=0.2216 rad/s。"
+     },
+     {
+      "t": null,
+      "log_step": null,
+      "text_zh": "推断：评分器报告的最大区段平面速度误差 0.2274 m/s 超过 DAG 的 0.2 m/s 阈值，因此即使已执行全部控制步且未跌倒，完整五段验收也未通过；此 event 节点从未完成。该判断依据具体误差字段，而非仅依据 success=0；不把会话测试成绩当作交付回放成绩。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–20.286 s，Spot 四条腿的黑色足端交替接触方格地面，承重、推进和转向；视频末尾仍保持足端支撑，未见躯干触地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.2274
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.2216
+  },
+  {
+   "name": "worst_gait_err",
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": null,
+   "log_step": null,
+   "where": "操作中",
+   "evidence_zh": "推断：断点为 Spot 正在执行的“按指令行进”。视频 0.00–20.286 s 显示持续迈步和转向、未见倒地；inputs/run.json 的 reward 确认已执行 1000/1000 步且 fell=0，但 worst_lin_err=0.2274 m/s 超过 0.2 m/s，行进过程验收失败。没有交付回放的逐段或逐步速度记录，无法定位误差属于哪一段、何时不可修复，故视频锚点留空。"
+  },
+  "unmet": {
+   "commanded_gait_completed": false
+  },
+  "phenomena": [],
+  "root_causes": [],
+  "gap_zh": "本次没有状态日志，7 fps 双视图也无速度或误差叠字。已逐帧检查本次视频及成功参考：两者均呈持续迈步、转向与足端支撑，未能可靠辨认本次超阈值误差对应的过慢、偏航或侧向偏移，故现象层留空，不将指标未达标改写成可见现象。reward 未报告 worst_gait_err 或各段误差，不能确认足高指标是否也失败。log step 43、46、48 的速度波动与调参输出属于会话测试；log step 54 写入按段优化、保存轨迹并回放的脚本，log step 55 仅返回后台启动，log step 56 自述已启动，log step 58 出现模型提供方过滤提示，未记录后台最后一次保存和验证的完成时点及最终参数。因此无法将测试中的具体控制缺陷或过滤中断与交付物失败建立可靠因果关系，根因层留空；也无法确定过滤是否发生于交付结果固定之前。"
+ },
+ "summary_zh": "Spot 在评分回放中执行了全部 1000 个控制步，视频 0.00–20.286 s 显示持续迈步、转向且未倒地，但评分器报告最大区段平面速度误差 0.2274 m/s，超过 0.2 m/s，唯一节点“按指令行进”未完成，最终为 0/1。断点属于该步操作中；低帧率视频和汇总成绩无法定位具体失准时刻或辨认明确的失败现象。会话测试不能替代交付物的逐步记录，且 log step 58 的过滤提示与后台优化保存的先后不明，故现象和根因均留空。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未完成、断点、时间钟和 gap_zh 均有证据支持：回放 1000 步无跌倒，worst_lin_err=0.2274>0.2；现象/根因留空有理由（无逐段误差，交付来自后台 gen.py，log 58 过滤发生时保存与验证时点不明）。未发现需改动的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:16:50Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--hrxypef/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--hrxypef/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "bd0c3649d4d5838b701896ee3c49bc566c73eba8f6372f1ee350cc60e4a8f1b9",
+   "video_sha256": "b8d27dbd36c74207c6a557d785f7a66d15c072e132b0a5fb0590f697bf91e06e",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--jnfvcno.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--jnfvcno.json
new file mode 100644
index 0000000000000000000000000000000000000000..b06171a96ec53e979c86575cae9dd82590b41314
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--jnfvcno.json
@@ -0,0 +1,205 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-privileged--jnfvcno",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.768,
+   "worst_yaw_err": 0.5408,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 108,
+   "agent_resets": 167,
+   "agent_sim_steps": 524450,
+   "agent_replays": 58,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 1.64,
+   "peak_processes": 10,
+   "peak_disk_gb": 0.832,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--jnfvcno/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "c26b33e4c421a339187e256279e7fc721cdaa13a9d6f6221f9de540fb45fd7e8"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 2.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0–4.0 s，Spot 四足持续交替抬落并向前移动，躯干保持直立；侧面视图的躯干朝向变化很小。对照成功参考录像视频 2.0–4.0 s 的明显转向，本次前段转向不足。"
+     },
+     {
+      "t": 13.571,
+      "log_step": null,
+      "text_zh": "观察：视频 12.143–16.0 s，Spot 仍反复迈腿，地面方格相对机身移动缓慢；视频 16.143–20.286 s 继续迈步，末尾一秒至最后一帧均未见翻倒。"
+     },
+     {
+      "t": null,
+      "log_step": 182,
+      "text_zh": "观察：log step 182 对当时保存的 trajectory.npz 进行独立 CLI 回放，检查器输出 steps=1000、fell_at=null、tracking_ok=false；五段 lin_err 为 0.5564、0.4573、0.4682、0.7680、0.5600 m/s，yaw_err 为 0.5408、0.4809、0.4563、0.3822、0.4135 rad/s，均超出各自 0.2 m/s 和 0.3 rad/s 阈值；worst_gait_err=0.0408 m。"
+     },
+     {
+      "t": null,
+      "log_step": 190,
+      "text_zh": "推断：log step 189 再次生成最终交付文件，log step 190 的检查结果及末态哈希 5ab2eb0b4b57 与 log step 182 相同；评分器 reward 同时确认 deterministic=1、n_actions=1000、fell=0、worst_lin_err=0.768、worst_yaw_err=0.5408。结合视频可见行进，完整执行了回合但未通过行进验收，该事件节点从未完成；会话中其他版本的跌倒不计入本次轨迹。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–20.286 s，Spot 四条腿的黑色足端交替落在方格地面上，支撑躯干并推进、转向；视频 2.0 s 和视频 13.571 s 的双视图可见着地支撑足和抬起的足端。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.768,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.5408,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_gait_err",
+   "value": 0.0408,
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 2.0,
+   "log_step": 190,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 1.429–4.0 s，Spot 已在迈步行进，但侧面视图中转向很少；后续仍持续迈步而未摔倒。log step 190 最终文件检查的首段 lin_err=0.5564 m/s、yaw_err=0.5408 rad/s，后四段也均超限，评分器 reward 确认未成功。视频 2.0 s 是前段偏差的粗略导航锚点，不是阈值首次越界时间。"
+  },
+  "unmet": {
+   "commanded_gait_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 2.0,
+    "log_step": 190,
+    "evidence_zh": "观察：视频 1.429–4.0 s，Spot 虽在迈步，躯干转向明显小于成功参考录像相同视频时段；视频 12.143–16.0 s，机器人反复迈腿但推进缓慢，全程至视频 20.286 s 未见倒地。log step 190 对最终轨迹的检查报告全部五段平面速度及偏航误差超限，评分器 reward 的 fell=0；表现为行进与转向跟踪不足。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 190,
+    "evidence_zh": "推断：agent 在 log step 11 说明按指令和固定步态相位生成关节目标轨迹，log step 55 记录增大步幅改善速度却使反向切换不稳、偏航随步态振荡的会话诊断。最终 log step 189–190 生成并检查的轨迹仍在首段控制步 100、200 分别仅有前向速度约 0.59、0.56 m/s，而指令约为 0.99 m/s；全部五段速度和偏航误差均超限，且评分器确认回放确定、未跌倒。结合视频 1.429–4.0 s 转向不足，支持关节运动控制未产生要求的躯干速度和偏航；不能据此进一步指定某个参数或反馈环节。",
+    "source_zh": "日志"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志，录像仅 7 fps 且无速度叠字，不能精确定位误差首次越界或从画面量化步态相位、足高与瞬时速度。原生足高指标 0.0408 m 取自 log step 182 的独立 CLI 检查及 log step 190 最终文件检查，两者末态哈希相同；reward 未单列该指标。部分编辑日志仅有 patch 调用占位，无法还原最终生成器的全部控制细节，因此根因停留在“控制”。未将会话诊断的控制步转换为视频时间。"
+ },
+ "summary_zh": "Spot 完整运行 1000 个控制步并保持直立，但“按指令行进”未通过，节点最终完成 0/1。视频前段转向不足，后退段推进缓慢；最终文件检查和评分器一致显示平面速度、偏航误差超限，归为步态现象及控制根因。缺少逐帧状态与完整编辑内容，不能进一步确定具体控制子类；足高最差误差由最终文件检查报告为 0.0408 m。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "单节点未完成、最终 0/1、断点、步态现象与控制根因均与 log step 182/190 的检查输出（五段 lin_err/yaw_err 全超限、fell_at=None）及视频帧一致；未发现需改内容的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:15:16Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--jnfvcno/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--jnfvcno/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "c867dbe8fcd6e3a19ff60192fce250e176895af1bb45b5a84805016089726094",
+   "video_sha256": "c26b33e4c421a339187e256279e7fc721cdaa13a9d6f6221f9de540fb45fd7e8",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--jrb4wxt.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--jrb4wxt.json
new file mode 100644
index 0000000000000000000000000000000000000000..67567800e1a688ef30ecf8c2daf456093c29b39d
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--jrb4wxt.json
@@ -0,0 +1,199 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-privileged--jrb4wxt",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 0,
+   "fell_at_step": -1.0,
+   "worst_lin_err": 0.7307,
+   "worst_yaw_err": 1.3843,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 231,
+   "agent_resets": 500,
+   "agent_sim_steps": 416365,
+   "agent_replays": 17,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 7.505,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.255,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--jrb4wxt/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "00dd05f0896a6ba94be3749555f20fc3d0efd70aefc810d585dfe1813b3b7527"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 3.0,
+      "log_step": null,
+      "text_zh": "观察：已逐帧检查视频 0.0–20.286 s 的全部143帧。视频 1.0–4.0 s，Spot四足交替落地并带动身体移动，但侧面视图中机身朝向变化较小；成功参考视频约 1.0–4.0 s 则呈现明显连续转向。视频 8.857–16.286 s 可见躯干和足端往复摆动，朝向推进仍很有限；视频 19.429–20.286 s 机器人仍以足端支撑，未见倒地。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 180,
+      "text_zh": "观察：log step 180 对最终交付的 trajectory.npz 使用评分器回放检查，报告 steps=1000、horizon=1000、success=false、tracking_ok=false、fell_at=null；五段平面误差依次为0.3198、0.3554、0.6308、0.7307、0.3445 m/s，五段偏航误差依次为0.4126、0.4268、1.3843、1.3308、0.3783 rad/s，均超过相应的0.2 m/s及0.3 rad/s阈值；后退两段足高误差为0.0505、0.0504 m。reward同时报告 deterministic=1、n_actions=1000、fell=0，最大平面与偏航误差与该回放一致。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 180,
+      "text_zh": "推断：完整执行控制步数且保持支撑不等于按指令行进完成；最终交付的评分器回放明确未通过五段跟踪验收，且视频存在一致的转向不足，因此该完整过程事件从未达成，终值为false。这里没有把日志中的agent工作时间换算为视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–20.286 s，Spot的四条腿摆动，足端交替落在方格地面，承重并推进机身；例如视频 0.143、1.429、9.143、19.857 s 可见足端着地支撑及其后摆腿。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.7307
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 1.3843
+  },
+  {
+   "name": "worst_gait_err",
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 3.0,
+   "log_step": 180,
+   "where": "操作中",
+   "evidence_zh": "观察：视频约 3.0 s，Spot已在持续摆腿行进，但侧面视图仍接近侧向，视频 1.0–4.0 s 的连续转向幅度较小；成功参考在相应视频区间明显转向。log step 180 的最终交付回放报告首段 lin_err=0.3198 m/s、yaw_err=0.4126 rad/s，已经不满足验收，且后续四段也未达标。推断：最早一段的行进跟踪失败一直没有被修复；视频约 3.0 s 仅作该过程偏差的导航锚点，不代表精确越阈时刻。"
+  },
+  "unmet": {
+   "commanded_gait_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "步态",
+    "t": 3.0,
+    "log_step": 180,
+    "evidence_zh": "观察：视频 1.0–4.0 s，Spot四足持续摆动、机身发生位移，但朝向改变缓慢；视频 8.857–16.286 s，身体反复俯仰和摆动，连续转向仍较弱。相同双视图中的成功参考在对应视频区间有明显连续转向；本次视频末尾 19.429–20.286 s 仍保持足端支撑，reward的fell=0。推断：这表现为行进与转向未跟上指令的步态偏差；log step 180 的首段及后续速度、偏航跟踪失败支持该判断。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志",
+    "evidence_zh": "观察：log step 134 保存的最终生成脚本按指令、固定scale=2.0及kx=0.04、ky=0.02生成周期足端目标，再用数值逆运动学转换为12关节位置；该脚本直接生成整段动作。log step 175–176 重新生成并回放交付文件，log step 180 再次确认同一末态哈希48dd82f9ac24及各段跟踪失败，reward的deterministic=1且最大误差一致。推断：该控制实现虽能维持支撑，却未把目标足端运动转化为所需的机身平面速度与转向；归为控制能力不足。未有证据把问题唯一定位到某个增益、逆解误差或接触布局，故不细分子类。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志；视频仅7 fps，对约3.66 Hz足端运动存在采样混叠，也没有速度或姿态数值叠字，因此无法仅靠画面量化足高相位误差或定位最早瞬时偏差。精确分段误差依据log step 180的最终交付评分器回放；reward没有worst_gait_err字段，因此对应native_metrics.value保留null，回放检查自身报告0.0505 m。现有日志不足以进一步确定具体控制失效环节。"
+ },
+ "summary_zh": "Spot完整执行了1000个控制步，视频全程可见足端交替支撑，末尾也未倒地，但“按指令行进”节点未完成（0/1）。视频前段已呈现转向不足，后退阶段机身反复摆动；最终交付回放证实五段速度与偏航均超标，根因归于未实现指令跟踪的控制轨迹。缺少逐帧状态且视频仅7 fps，无法再细分控制失效环节或由画面量化足高相位误差。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "节点未达成、final=false、断点（首段跟踪失败）、步态现象和控制根因均有 log step 180 的评分器回放（五段误差超阈、fell_at=null）及视频支持；与成功参考对比，本次视频朝向变化明显偏小，标注一致。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:02:44Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--jrb4wxt/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--jrb4wxt/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "7a9ac7de08cce7caa4c2bbb3b1f3d47a4c12269da9092cae3c53f896726b0929",
+   "video_sha256": "00dd05f0896a6ba94be3749555f20fc3d0efd70aefc810d585dfe1813b3b7527",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--xgunypy.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--xgunypy.json
new file mode 100644
index 0000000000000000000000000000000000000000..ba3e45323f23ab55cb17741928b40672bcd55a84
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-privileged--xgunypy.json
@@ -0,0 +1,200 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-privileged--xgunypy",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "deterministic": 1,
+   "grader_error": 0,
+   "infra_error": 0,
+   "replay_success": 0,
+   "n_actions": 1000,
+   "horizon": 1000,
+   "fell": 1,
+   "fell_at_step": 419.0,
+   "worst_lin_err": 2.3374,
+   "worst_yaw_err": 1.0033,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "video_rendered": 1,
+   "agent_sessions": 68,
+   "agent_resets": 4426,
+   "agent_sim_steps": 595465,
+   "agent_replays": 3,
+   "agent_cost_self_reported": 1,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 1.737,
+   "peak_processes": 14,
+   "peak_disk_gb": 0.981,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--xgunypy/replay.mp4",
+  "duration_s": 20.429,
+  "fps": 7.0,
+  "frames": 143,
+  "sha256": "4a4bc8b88f2d20cfaec9012ce9f6552321899bae1357c7a70edbbfd4adc6555f"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 1.0,
+      "log_step": null,
+      "text_zh": "观察：视频 0.00–8.29 s 的跟随与侧面双视图显示黄色 Spot 四足交替落地支撑、机身相对方格地面行进并转向，已实际尝试按指令行进。"
+     },
+     {
+      "t": 8.43,
+      "log_step": 153,
+      "text_zh": "观察：视频 8.43–8.71 s，Spot 躯干明显前俯、部分足端离地，随后视频 8.86–9.14 s 侧翻；log step 153 对最终保存的 trajectory.npz 执行评分工具 replay，返回 steps=1000、fell_at=419、success=false，与评分 reward 的 fell_at_step=419、deterministic=1 一致。"
+     },
+     {
+      "t": 20.286,
+      "log_step": 153,
+      "text_zh": "观察：视频 19.29–20.29 s 中 Spot 仍呈四腿朝上、机身翻转姿态，未恢复正常行进；log step 153 的最终文件回放检查显示前两段三项误差均达标，但后三段未达标，最差平面速度误差 2.3374 m/s、偏航误差 1.0033 rad/s、足高误差 0.0649 m。因此完整日程且全程不触发倾斜判据的事件从未完成。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–8.29 s，Spot 的四条腿以足端交替接触方格地面，承重并推进、转向；视频 8.43 s 后失衡翻倒。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 2.3374
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 1.0033
+  },
+  {
+   "name": "worst_gait_err",
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better",
+   "value": 0.0649
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 8.43,
+   "log_step": 153,
+   "where": "操作中",
+   "evidence_zh": "观察：Spot 正在执行行进日程，视频 8.29 s 尚由足端支撑，视频 8.43–8.71 s 明显前俯失衡，视频 9.14 s 已翻倒；此后至视频 20.29 s 未恢复正常行进。log step 153 的最终交付文件回放检查及 reward 均记录控制步 419 触发倾斜跌倒判据，故断点为正在执行的“按指令行进”。"
+  },
+  "unmet": {
+   "commanded_gait_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 8.43,
+    "log_step": 153,
+    "evidence_zh": "观察：视频 8.43–8.71 s，Spot 躯干前端下沉、后端抬起，足端支撑失稳；视频 8.86–9.14 s 随即侧翻，视频 19.29–20.29 s 仍可见四腿朝上。log step 153 的回放检查返回 fell_at=419，评分字段 fell=1。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "log_step": 153,
+    "source_zh": "日志",
+    "evidence_zh": "推断：agent 在 log step 143 将足端轨迹及逆运动学控制器生成的关节目标记录为开环文件，在 log step 153 使用最终参数重新生成并回放同一交付文件；评分工具明确复现控制步 419 的失衡，且后三段速度、偏航与足高跟踪均失败，和视频 8.43–9.14 s 的失衡翻倒一致。这支持交付的关节目标轨迹未能稳定应对指令反向的控制问题；无法仅由这些证据区分足端布局、轨迹动态或反馈增益的具体责任。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志；录像仅 7 fps 且未叠加速度、IMU 或足高误差，不能精确重建倾斜越界时刻、足端相位及失稳机制。视频 9.29 s 后部分画面被地面遮挡或裁切，但末秒可见翻转姿态。worst_gait_err 未出现在 reward 汇总中，其数值 0.0649 m 来自 log step 153 对最终交付文件的评分工具 replay 输出；其余最差误差和跌倒控制步均与 reward 一致。agent 在 log step 156 关于刚度与拖脚的机理自述未在最终轨迹上获得独立验证，故根因仅标至“控制”。"
+ },
+ "summary_zh": "Spot 实际尝试了完整行进轨迹；log step 153 的最终文件回放检查显示前两段误差达标，但完整五段验收未通过，唯一节点“按指令行进”最终为 0/1。视频约 8.43 s 在反向指令切换附近出现前俯失衡，随后翻倒，末秒仍未恢复，评分也记录控制步 419 触发倾斜判据；证据支持控制层面的失败。缺少逐帧传感器状态，无法进一步确认具体失稳机制。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "单节点未达成、断点 8.43 s 视频时钟（控制步 419≈8.38 s）与翻倒帧一致，平衡现象和控制根因有 log step 153 回放与评分字段支持，未发现需改动之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:04:32Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--xgunypy/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-privileged--xgunypy/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "acd00a4331d557e937e9ed81bac79e9ffa95bcb05b2450fd40915ab5a4f5cd9b",
+   "video_sha256": "4a4bc8b88f2d20cfaec9012ce9f6552321899bae1357c7a70edbbfd4adc6555f",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--9lhx5tc.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--9lhx5tc.json
new file mode 100644
index 0000000000000000000000000000000000000000..55ddd46688bb743f77c937ea0f8ad806a6db19b7
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--9lhx5tc.json
@@ -0,0 +1,211 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-standard--9lhx5tc",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 43,
+   "sim_steps": 43,
+   "physics_steps": 215,
+   "steps": 113,
+   "elapsed_s": 930.9,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 1.0547,
+   "worst_yaw_err": 0.5157,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 1,
+   "reset_budget": 10,
+   "episodes": 2,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.337,
+   "peak_processes": 6,
+   "peak_disk_gb": 0.189,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-standard--9lhx5tc/replay.mp4",
+  "duration_s": 0.88,
+  "fps": 25.0,
+  "frames": 22,
+  "sha256": "47b9c2a9fa24e788ece6ff671d921f40c471dfdd234cf6b7475754eb5e5d8f6a"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 21,
+      "text_zh": "观察：log step 21 将服务重置到第2回合；视频0.00 s 的Spot从四足站姿开始，随后抬足踏步。评分器记录episodes=2、resets=1、n_actions=43，live_matches_replay=1；本标签只评价最后的第2回合，不把重置前的调试运动计入。"
+     },
+     {
+      "t": 0.8,
+      "log_step": 22,
+      "text_zh": "观察：视频0.00–0.84 s 全部22帧中，Spot短暂抬落四足，后段躯干明显侧倾，未出现完整20秒的行进和指令切换。log step 22 的第2回合控制输出在control_steps=40时因向上轴z分量0.8738暂停；log step 23 从相同40步状态尝试扶正后，服务返回EpisodeOver: fell。评分器记录本回合仅43个控制步、fell=1，故按指令行进节点未完成。"
+     },
+     {
+      "t": 0.8,
+      "log_step": 22,
+      "text_zh": "推断：log step 3 服务规格给出50 Hz，log step 22 的本回合40个控制步对应视频约0.80 s；该处画面确有明显侧倾，与重力读数一致。评分器43步对应约0.86 s，与视频总长0.88 s一致；此映射仅使用控制步数，不使用agent工作时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频0.00–0.84 s，Spot四条腿的足端在方格地面上抬起、落下，由着地足支撑躯干；视频0.64–0.84 s躯干逐渐侧倾，着地足未能维持稳定姿态。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 1.0547
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.5157
+  },
+  {
+   "name": "worst_gait_err",
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 0.64,
+   "log_step": 23,
+   "where": "操作中",
+   "evidence_zh": "观察：视频0.64 s起Spot在踏步时躯干侧倾加重，视频0.76–0.84 s仍未恢复直立；log step 22在第2回合40个控制步时暂停步态，log step 23尝试姿态恢复后服务返回fell。评分器确认本回合43步即结束，按指令行进没有完成或修复。"
+  },
+  "unmet": {
+   "commanded_gait_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 0.76,
+    "log_step": 23,
+    "evidence_zh": "观察：视频0.64–0.84 s，Spot躯干向一侧明显倾倒，两侧腿的支承高度不对称，末帧仍侧倾；log step 22读取向上轴z分量仅0.8738，随后log step 23服务以fell终止，评分器fell=1。画面尚未显示躯干完全倒地，但已发生触发终止的失衡。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 22的第2回合脚本以5步一批读取IMU重力、角速度和平移速度，并对足端轨迹施加姿态补偿；实测平均平面速度为[0.1839,-0.2558] m/s，指令为[0.9939,0.3707] m/s，40个控制步后向上轴z分量降至0.8738。log step 23继续用重力、角速度反馈计算扶正关节目标，仍触发fell；结合视频0.64–0.84 s持续加重的侧倾，说明该闭环步态与姿态恢复未能稳定机身。现有证据不定位到某个增益或补偿符号。"
+   }
+  ],
+  "gap_zh": "观察：inputs/未提供逐帧状态日志，评分器也未报告worst_gait_err；因此无法复核每步接触、精确越过0.85阈值的时刻及各段足高误差。log step 22的控制代码有一段被截略，不能据此确定具体反馈增益或补偿公式的错误；断点视频时间只作近似导航。"
+ },
+ "summary_zh": "第2回合中Spot已开始抬足踏步，但仅执行43/1000个控制步，按指令行进节点始终未达成，最终0/1。视频0.64–0.84 s侧倾加重，随后服务以fell终止；日志支持闭环步态与姿态恢复未能维持平衡。缺少逐帧状态及评分器足高误差，不能进一步定位具体控制参数。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "标签与视频、日志一致：仅评价第2回合，43个控制步即fell，节点未达成，最终0；断点、平衡现象与反馈控制根因有证据支撑，控制步到视频时间的映射已说明。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:05:06Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--9lhx5tc/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--9lhx5tc/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "a65e97a9f069d627ee2e58e48b64eb7c094ed0d2ec68d8ba23b6ec772e0f8de4",
+   "video_sha256": "47b9c2a9fa24e788ece6ff671d921f40c471dfdd234cf6b7475754eb5e5d8f6a",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--b2w9ddy.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--b2w9ddy.json
new file mode 100644
index 0000000000000000000000000000000000000000..0380179c2cd491e4b8435da2aa100469d8e3bee1
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--b2w9ddy.json
@@ -0,0 +1,213 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-standard--b2w9ddy",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 465,
+   "sim_steps": 465,
+   "physics_steps": 2325,
+   "steps": 3862,
+   "elapsed_s": 1412.1,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.3196,
+   "worst_yaw_err": 0.4461,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.585,
+   "peak_processes": 11,
+   "peak_disk_gb": 0.831,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-standard--b2w9ddy/replay.mp4",
+  "duration_s": 9.75,
+  "fps": 12.0,
+  "frames": 117,
+  "sha256": "09464d373a8510697313fda95ec85f67f0e8809e97ca6aa9a9062588a6f27347"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 90,
+      "text_zh": "观察：视频 0.0–8.5 s 的跟随与侧面双视图中，Spot 从站立开始交替抬落四足，在方格地面上行进并转向，随后进入后退运动；log step 90 启动最后一次重置后的回合，log step 95 确认它是 episode=11。这里只标该末回合，不计前十个已重置回合的动作。"
+     },
+     {
+      "t": 9.33,
+      "log_step": 95,
+      "text_zh": "观察：视频 9.333–9.667 s 中 Spot 后退时躯干侧倾加剧，四足支撑姿态失稳，录像在倾斜状态结束，未显示后续日程。log step 95 的机器人服务终态为 episode=11、control_steps=465、sim_time_s=9.3、done=True、reason=fell；评分器同报 sim_steps=465、fell=1、success=0。"
+     },
+     {
+      "t": 9.667,
+      "log_step": 95,
+      "text_zh": "推断：末回合只执行 465/1000 个控制步，且已触发跌倒终止，没有完成完整五段指令日程，故“按指令行进”从未达到、final=false。视频末帧为视频 9.667 s，服务终止时间为仿真 9.30 s，两者分别引用，不把仿真时间直接当视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–9.667 s，Spot 四条腿的足端反复落到方格地面并交替离地，承重、推进和转向；末段躯干侧倾，未见躯干已经着地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.3196,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.4461,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_gait_err",
+   "value": null,
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 9.33,
+   "log_step": 95,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 9.333–9.667 s，Spot 正在后退行进，躯干持续侧倾，未恢复稳定就结束录像；log step 95 的末回合服务输出明确为 control_steps=465、reason=fell，评分器 fell=1。失败发生在执行行进步骤期间，完整日程尚未完成。"
+  },
+  "unmet": {
+   "commanded_gait_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 9.33,
+    "log_step": 95,
+    "evidence_zh": "观察：视频 9.333–9.667 s 的双视图可见 Spot 机身侧倾逐渐增大，足端支撑姿态失稳，但未显示完全倒地。log step 95 的机器人服务终态 IMU 重力方向 z 分量为 -0.847821，且 done=True、reason=fell；评分器 fell=1。该任务按躯干倾斜阈值判跌倒，不要求躯干落地。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "t": 9.33,
+    "log_step": 94,
+    "evidence_zh": "推断：后退时姿态反馈未能抑制横滚失稳。log step 84 的代码根据足端接触选择支撑腿，并用 IMU 重力方向与陀螺仪构造横滚纠正；log step 90–91 在末回合启用后退专用阻尼和倾斜恢复。末回合传感读数（log step 93–94）仍从 control step 441 的 gravity_y=-0.169、gravity_z=-0.982，发展到 control step 464 的 gravity_y=-0.467、gravity_z=-0.858、gyro_x=2.94 rad/s，随后服务在 control step 465 判 fell（log step 95）；视频 9.333–9.667 s 同见机身侧倾加剧。证据支持姿态闭环失稳，不能据此认定某一增益数值或接触布局必然错误。",
+    "source_zh": "日志 / 视频 / 评分器字段"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志，末回合传感记录仅有日志中的抽样读数；无法进一步区分具体增益、足端落点或执行器饱和对失稳的贡献。评分器未报告 worst_gait_err，视频也不能精确量化逐段足高误差，因此该指标保留 null。服务终止于仿真 9.30 s，录像末帧是视频 9.667 s；未建立逐帧时钟映射，断点使用实际检查的画面作约略锚点。"
+ },
+ "summary_zh": "末回合 Spot 已交替支撑行进、转向并进入后退，但仅执行 465/1000 个控制步即触发跌倒终止，“按指令行进”最终为 0/1。视频 9.333–9.667 s 可见躯干侧倾加剧，结合末回合 IMU 与控制代码，推断后退时姿态反馈未能稳定机身；评分器平面速度和偏航误差亦超阈值。缺少逐帧状态与评分器足高误差，不能确定具体失稳参数或完整步态误差。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "末回合（episode 11）465/1000 控制步因 IMU 向上轴 z=-0.858（<0.85 阈值，日志 step 95 reason=fell，评分器 fell=1）而终止，commanded_gait_completed=false、断点、平衡现象与反馈控制推断均有证据支持；时钟分别引用视频时间与仿真时间，未发现需要修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:14:57Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--b2w9ddy/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--b2w9ddy/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "f97a9550e93e0965a71c6772b7a7cbfd302b97c0a02959fe90ddf675687708bb",
+   "video_sha256": "09464d373a8510697313fda95ec85f67f0e8809e97ca6aa9a9062588a6f27347",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--bya5epm.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--bya5epm.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf41d7c1341ae3500586cb1c9847a4675bcb663d
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--bya5epm.json
@@ -0,0 +1,213 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-standard--bya5epm",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "graded_by_live": 1,
+   "replays": 0,
+   "finalized": 1,
+   "missing_log": 0,
+   "sim_error": 0,
+   "grader_error": 0,
+   "infra_error": 0,
+   "n_actions": 362,
+   "sim_steps": 362,
+   "physics_steps": 1810,
+   "steps": 1479,
+   "elapsed_s": 1259.9,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.7671,
+   "worst_yaw_err": 0.7621,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "instance_mismatch": 0,
+   "no_verdict": 0,
+   "gave_up": 0,
+   "sim_crashed_on_call": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.368,
+   "peak_processes": 12,
+   "peak_disk_gb": 0.284,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-standard--bya5epm/replay.mp4",
+  "duration_s": 7.118,
+  "fps": 17.0,
+  "frames": 121,
+  "sha256": "269bf276265f284eab90d50a0811352fc7172c8d55bdb2f7cda7cc36e48ee4ea"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.5,
+      "log_step": null,
+      "text_zh": "观察：视频 0.0–6.5 s 的双视图中，黄色 Spot 持续摆腿，足端交替落在方格地面上，机身移动并转向，已实际开始行进；视频 6.765–7.059 s 躯干降低并明显倾斜，录像随后结束，未呈现完整的 20 秒指令过程。"
+     },
+     {
+      "t": 7.059,
+      "log_step": 114,
+      "text_zh": "观察：log step 111 启动最后一次控制器运行；log step 114 的机器人服务状态为 episode=11、control_steps=362、done=True、reason=fell，且 inputs/run.json 的评分器字段 n_actions=362、fell=1、success=0 与之相符。此前 log step 106 的 episode=10 是另一测试回合；log step 117 的代码修改发生在最终回合结束后。"
+     },
+     {
+      "t": 7.059,
+      "log_step": 114,
+      "text_zh": "推断：最终回合只执行了要求的 1000 个控制步中的 362 步且触发跌倒判据，按指令行进节点从未完成。节点时刻依据视频帧，不将日志的 agent 时间或控制步数直接换算为视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–6.5 s，Spot 的四条腿交替摆动，多个足端先后接触方格地面承重并推进行进；视频 6.765–7.059 s 躯干明显降低、倾斜，足端仍可见地面接触。不据此断言躯干已经触地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "value": 0.7671,
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_yaw_err",
+   "value": 0.7621,
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better"
+  },
+  {
+   "name": "worst_gait_err",
+   "value": null,
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better"
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 6.8,
+   "log_step": 114,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 6.765–7.059 s，正在摆腿行进的 Spot 躯干下沉并明显倾斜，未恢复便结束录像；log step 114 的机器人服务确认最终 episode=11 在 control_steps=362 时以 reason=fell 结束。失败发生在按指令行进过程中，完整指令日程没有执行完。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 6.8,
+    "log_step": 114,
+    "evidence_zh": "观察：视频 6.765–7.059 s 的双视图显示 Spot 躯干逐渐贴近地面并明显侧倾；log step 114 的服务状态 reason=fell 与评分器 fell=1 支持失衡判定。画面结束时仍有足端着地，未显示躯干完全倒地；此任务的跌倒判据是过度倾斜，无须躯干触地。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": 6.8,
+    "log_step": 109,
+    "source_zh": "日志 / 视频 / 评分器字段",
+    "evidence_zh": "推断：log step 109 的 agent 控制脚本按足端相位曲线和解析逆解生成关节位置命令，log step 111 执行该脚本；视频 6.765–7.059 s 机器人在平坦地面上摆腿时躯干下沉、倾斜，随后服务在 log step 114 报告 fell。结合评分器 worst_lin_err=0.7671 m/s、worst_yaw_err=0.7621 rad/s，控制器既未实现要求的运动跟踪，也未维持行进中的稳定支撑；可归因至控制，但现有证据不足以锁定逆解、反馈或某个参数。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志，也未提供最终回合的连续 IMU、足端高度、接触力和关节跟踪记录；log step 109 的脚本正文中段亦被输入日志截断。因此无法精确定位倾斜阈值首次越界的控制步，或进一步区分控制子类。断点用视频约 6.8 s 的末段失衡画面作导航锚点；不把服务控制步数换算成视频时刻。评分器未报告 worst_gait_err，该原生指标保留 null。",
+  "unmet": {
+   "commanded_gait_completed": false
+  }
+ },
+ "summary_zh": "本次评分为最后的第 11 回合：Spot 已实际摆腿行进、转向并以足端支撑地面，但在完成 362/1000 个控制步后因跌倒判据结束，唯一的按指令行进节点未完成（0/1）。视频约 6.8–7.059 s 可见躯干下沉和明显倾斜，服务的 fell 判定支持平衡失败；相位曲线与逆解生成的关节控制未维持稳定行进，根因保留在控制层级。缺少连续姿态、接触力及足高误差数据，不能进一步确定具体控制子类或首次越界时刻。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "标注与证据一致：唯一节点未完成（final=false，0/1），评分回合为第 11 回合，视频 6.8 s 后躯干下沉倾斜，服务 reason=fell、评分器 fell=1 支持；断点、平衡现象和停在“控制”的根因均有依据，未发现需修改之处。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:07:55Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--bya5epm/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--bya5epm/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "105f964a6046a237584511c6a79fee139731735e5250246d0d698d5a349b35ec",
+   "video_sha256": "269bf276265f284eab90d50a0811352fc7172c8d55bdb2f7cda7cc36e48ee4ea",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--f4x6niz.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--f4x6niz.json
new file mode 100644
index 0000000000000000000000000000000000000000..7cd5ffd127d68bf317d29ee5ce5443c7553b93fc
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--f4x6niz.json
@@ -0,0 +1,212 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-standard--f4x6niz",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 475,
+   "sim_steps": 475,
+   "physics_steps": 2375,
+   "steps": 766,
+   "elapsed_s": 915.4,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.5066,
+   "worst_yaw_err": 0.5553,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-standard--f4x6niz/replay.mp4",
+  "duration_s": 9.917,
+  "fps": 12.0,
+  "frames": 119,
+  "sha256": "588b398900860c37365668323df8fc3375cab1883c2ad20e17d21aa25e9db18a"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.5,
+      "log_step": null,
+      "text_zh": "观察：视频 0.5–8.0 s 的跟随与侧面双视图中，黄色 Spot 四足交替落地，机器人随方格背景变化持续行进并转向，已开始执行行进步骤。"
+     },
+     {
+      "t": 9.833,
+      "log_step": 47,
+      "text_zh": "观察：视频 8.0–9.833 s 中机器人继续迈步，末段躯干降低并侧倾，画面结束时未完成后续行进日程；log step 47 的服务状态为 episode=11、control_steps=475、done=True、reason=fell、resets_left=0，评分器同样记录 n_actions=475、fell=1、live_matches_replay=1。"
+     },
+     {
+      "t": 9.833,
+      "log_step": 46,
+      "text_zh": "推断：log step 45–46 明确选择并执行最后一次实时回合，log step 47 与评分器确认只执行475/1000个控制步，故完整20秒五段行进从未完成，节点始终为假；该判断依据实际执行长度和终止原因，并非仅依据零分。log step 46 的脚本打印476，与服务计数差一步，以服务及评分器的475为准；未把脚本计数或agent用时换算成视频时间。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.5–9.833 s，Spot 各腿足端反复接触方格地面、交替支撑躯干并推进、转向，末段支撑姿态失稳；未见躯干完全倒地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.5066
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.5553
+  },
+  {
+   "name": "worst_gait_err",
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 9.6,
+   "log_step": 47,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 9.333–9.833 s，Spot 在迈步时躯干进一步降低并侧倾，未恢复便结束录像；log step 47 的实时服务确认最后回合在control_steps=475时以fell终止，评分器fell=1。机器人正在执行按指令行进，未完成要求的1000控制步。视频锚点取末段可见失衡处，不将控制步数直接换算成视频时间。"
+  },
+  "unmet": {
+   "commanded_gait_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 9.6,
+    "log_step": 47,
+    "evidence_zh": "观察：视频 9.333–9.833 s 中机器人躯干降低并明显侧倾；log step 47 的服务状态为done=True、reason=fell，评分器fell=1。log step 46 的本回合脚本从传感器计算并打印min up z=0.8474228，低于DAG的0.85判据；画面没有显示完全侧翻落地，失败是先触发过度倾斜的跌倒判据。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "反馈控制",
+    "log_step": 46,
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 25、36 的控制器代码依据实测平面速度、偏航角速度及躯干姿态/角速度修正足端与关节目标，log step 45 将该控制器设为每次提交2个控制步后用于最后实时回合。log step 46 的本回合传感器统计显示，前两段平均前向速度约0.66、0.64 m/s，低于约0.994、0.924 m/s指令，且最终up z降至0.8474228；log step 47 服务确认fell，与视频 9.333–9.833 s 的倾斜相符。这支持速度与姿态反馈未能稳定跟踪的归因，尚不能定位到某个增益或逆解错误。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志，双视图未叠加传感器数据，不能精确确定倾斜首次越过阈值的帧、各足接触力或足高相位误差；评分器未报告worst_gait_err。log step 46 的脚本步计数为476，服务与评分器为475，录像时长也不等于475×0.02秒，故仅用所见视频帧定位失衡，不建立精确控制步到视频时间映射。具体反馈参数、模型差异或足端布局如何导致失稳尚无充分证据。"
+ },
+ "summary_zh": "最后评分回合中，Spot 已通过足端交替支撑行进并转向，但只执行475/1000个控制步，“按指令行进”节点未完成（0/1）。视频约 9.6 s 出现躯干降低和侧倾，服务确认以fell终止；评分器的平面速度误差0.5066 m/s和偏航误差0.5553 rad/s亦超阈值，日志支持反馈控制未能稳定跟踪。缺少逐帧状态及评分足高误差，无法进一步确定具体参数原因。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "节点未达成、断点（行进中跌倒，约9.6 s视频锚点）、平衡现象与反馈控制根因均有视频帧、log step 46/47及评分器字段支持；未把控制步换算成视频时间，未发现需改内容的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:25:04Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--f4x6niz/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--f4x6niz/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "cba03ff2a2d50b7d509237208448a45f9673d87b394f07d10368694df501757b",
+   "video_sha256": "588b398900860c37365668323df8fc3375cab1883c2ad20e17d21aa25e9db18a",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--gqotxv6.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--gqotxv6.json
new file mode 100644
index 0000000000000000000000000000000000000000..258994866771bff8a85553b69c1a869c353f3244
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--gqotxv6.json
@@ -0,0 +1,206 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-standard--gqotxv6",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 36,
+   "sim_steps": 36,
+   "physics_steps": 180,
+   "steps": 377,
+   "elapsed_s": 787.6,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.6769,
+   "worst_yaw_err": 0.3968,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-standard--gqotxv6/replay.mp4",
+  "duration_s": 0.76,
+  "fps": 25.0,
+  "frames": 19,
+  "sha256": "c630ff027c119242c677a830e7d7b92968ee38a50d532b04e0b3f1a15dcd3a23"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.0,
+      "log_step": 67,
+      "text_zh": "观察：视频 0.00 s 的跟随与侧面双视图中，Spot 四足着地站立；视频 0.04–0.72 s 中开始抬落足和小幅移动，但后半段躯干逐渐侧倾，直到末帧仍未恢复；完整检查了全部19帧。log step 67 的机器人服务状态为 episode=11、control_steps=36、done=true、reason=fell，评分器 n_actions=36、fell=1、live_matches_replay=1，确认这是最后受评回合；该回合没有执行完要求的1000步。"
+     },
+     {
+      "t": 0.72,
+      "log_step": 112,
+      "text_zh": "观察：视频 0.72 s 末帧可见 Spot 躯干明显侧倾，尚未完全倒地，录像随即结束；log step 112 的机器人服务状态仍为同一 episode=11、control_steps=36、reason=fell，说明后续修改没有恢复或重跑该回合。成功参考在参考视频 0–20.29 s 持续移动、转向并保持支撑，仅用于对照完成形态，不作为本回合进度证据。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00–0.72 s，Spot 以足端接触方格地面承重，四腿抬落足尝试行进；视频后半段足端支撑未能维持躯干直立。未见躯干已经触地，不额外声称躯干与地面的接触。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.6769
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.3968
+  },
+  {
+   "name": "worst_gait_err",
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 0.6,
+   "log_step": 67,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 0.40、0.60、0.72 s 的双视图显示 Spot 正在抬落足行进，躯干侧倾随后明显加剧，末帧没有恢复直立；log step 67 的机器人服务确认最后受评回合在 control_steps=36 因 fell 结束，未完成1000步。断点取视频约 0.60 s 的失稳过程，而非把跌倒约束当作步骤。"
+  },
+  "unmet": {
+   "commanded_gait_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 0.6,
+    "log_step": 67,
+    "evidence_zh": "观察：视频 0.60–0.72 s，Spot 的躯干相对起始站姿明显侧倾，左右腿的支撑姿态失去对称，未能恢复；log step 67 的机器人服务状态报告 reason=fell，评分器字段 fell=1。此处是过度倾斜触发的失衡终止，视频没有显示机器人完全侧躺或躯干着地。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 67 的实际执行脚本每10个控制步更新一次观测，以测得的 roll/pitch 构造 R 并生成足端目标；同一回合诊断记录的横滚角从控制步10的 -0.069 rad 增长到控制步30的 -0.344 rad，随后机器人服务报告 fell，结合视频 0.60–0.72 s 的侧倾，说明该控制器未能稳定姿态。另有独立代码证据：log step 94–95 显示 pos_i_prev 在逐腿循环中才以零数组创建，使左后腿首次抬足起点误用零坐标；log step 90 的离线输出也显示左后足初始平面目标为(0,0)，而 log step 92 的中立足点约为(-0.3293,0.1658)。这些控制缺陷在受评回合终止后才被修改，分别对失衡的贡献未经实际重跑验证，故不细分控制子类。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志，也没有受评回合末端的精确 IMU 向上轴数值；视频在过度倾斜时结束，不能声称发生了躯干触地。评分器未报告 worst_gait_err，且短回合没有覆盖后四段指令；log step 77、97–100 的姿态和足端起点修改均发生在受评回合结束后，只做离线检查，不能证明哪一项单独修正即可避免失稳。log step 71 的网络错误发生在 log step 67 已完成跌倒终止之后，不改变该物理失败的判定。"
+ },
+ "summary_zh": "最后受评的第11回合中，Spot 已通过足端支撑启动行进，但视频约 0.60–0.72 s 躯干侧倾加剧，机器人服务在第36个控制步以 fell 终止，唯一节点“按指令行进”未完成，最终为0/1。日志中的控制代码与诊断支持控制器失稳，并发现左后足首次抬足起点初始化错误；回合结束后的修补没有重新执行。缺少逐帧状态和末端精确IMU数据，无法区分各控制缺陷的贡献，也没有报告足高误差指标。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "唯一节点未完成、最终0/1、断点在该节点、平衡现象和控制根因均有视频与log step 67的诊断输出支持（roll -0.069→-0.344 rad，control_steps=36，reason=fell）；未发现需要修改的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:08:02Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--gqotxv6/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--gqotxv6/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "3fcefbefa25089a881a4738e323489b46da7e187441ab81e89b26b6cbbf0a97e",
+   "video_sha256": "c630ff027c119242c677a830e7d7b92968ee38a50d532b04e0b3f1a15dcd3a23",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--n9mu8gq.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--n9mu8gq.json
new file mode 100644
index 0000000000000000000000000000000000000000..2f746c085c5e11874a837cc2beb205cbb8d8db4c
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--n9mu8gq.json
@@ -0,0 +1,218 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-standard--n9mu8gq",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 700,
+   "sim_steps": 700,
+   "physics_steps": 3500,
+   "steps": 5410,
+   "elapsed_s": 1145.9,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 0.6751,
+   "worst_yaw_err": 0.7371,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 1,
+   "peak_gpu_memory_gb": 0.0,
+   "peak_memory_gb": 0.0,
+   "peak_processes": 0.0,
+   "peak_disk_gb": 0.0,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-standard--n9mu8gq/replay.mp4",
+  "duration_s": 14.1,
+  "fps": 10.0,
+  "frames": 141,
+  "sha256": "921961e429efcefd17785fe20931cb83c00fae9623007aa5135be7e42531a909"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.4,
+      "log_step": 78,
+      "text_zh": "观察：视频 0.0–0.8 s 双视图中，Spot 已开始交替抬足和落足，足端接触方格地面支撑躯干；视频 1.0–8.0 s 可见机器人持续迈步并改变朝向。"
+     },
+     {
+      "t": 8.4,
+      "log_step": 79,
+      "text_zh": "观察：视频 7.8–8.8 s 的连续帧中，Spot 在迈步过程中由前进转为后退；log step 79 的当前回合控制步 450 记录指令 vx=-0.9386 m/s、实测 vx=-0.3691 m/s，后续控制步 650 已处于第四段后退指令。"
+     },
+     {
+      "t": 14.0,
+      "log_step": 79,
+      "text_zh": "观察：视频 13.6–14.0 s 中 Spot 躯干明显侧倾，录像于该姿态结束；log step 79 的服务状态为 episode=11、control_steps=700、done=true、reason=fell，log step 80 再次确认。评分器报告 n_actions=700、fell=1、live_matches_replay=1，未执行完整 1000 步，故此完成事件从未达到。"
+     },
+     {
+      "t": 14.0,
+      "log_step": 79,
+      "text_zh": "推断：log step 77 启动最后一次重置回合，log step 79–80 确认它是最终第 11 回合。按 log step 7 的 50 Hz 控制率，当前回合控制步 700 × 0.02 s = 视频 14.0 s，已与视频 14.0 s 的侧倾终止画面核对；此前各回合仅为会话调试，不计入本节点。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.0–0.8 s、7.8–8.8 s 及 12.0–14.0 s 中，Spot 四条腿的足端交替落在平坦方格地面上，承重并推动机身前进、后退和转向；log step 78–79 的当前回合足端传感器也分别记录左右侧足端接触。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 0.6751
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.7371
+  },
+  {
+   "name": "worst_gait_err",
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 13.7,
+   "log_step": 79,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 13.6–14.0 s，Spot 正在后退转向，躯干出现明显侧倾，未恢复便结束回合。log step 79 的最终第 11 回合在控制步 700 记录 done=true、reason=fell；当时指令仍为第四段 (-0.9597,-0.2951,0.5688)，机器人没有完成整段行进日程。"
+  },
+  "unmet": {
+   "commanded_gait_completed": false
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 13.7,
+    "log_step": 79,
+    "evidence_zh": "观察：视频 13.6–14.0 s 中，Spot 躯干向侧面倾斜、两侧腿的支撑姿态明显不对称，录像在躯干尚未完全倒地时结束；log step 79 记录当前回合控制步 700 的 IMU 重力方向为 [-0.15295,0.50843,-0.84741]，服务判定 reason=fell，评分器 fell=1。推断：向上轴的世界 z 分量约为 0.84741，低于任务的 0.85 倾斜阈值，因此属于行进中的失衡。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "t": 13.7,
+    "log_step": 79,
+    "source_zh": "日志 / 视频 / 评分器字段",
+    "evidence_zh": "推断：agent 在 log step 77 声明使用局部腿运动学、指令滤波和较小后退步幅后启动最终回合；log step 79 的同一回合控制步 650 重力 z=-0.98676，控制步 700 已变为 -0.84741 并触发 fell，与视频 13.0–14.0 s 的侧倾相符。机器人已实际执行前进及后退指令，但行进控制未能维持躯干稳定；评分器还报告 worst_lin_err=0.6751 m/s、worst_yaw_err=0.7371 rad/s，均超阈值。证据支持控制能力不足，不能进一步确定具体控制子类。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志；日志保留了最终回合的稀疏传感读数，但部分中间输出已截断，控制器修改仅显示 patch 调用而无完整补丁内容，无法验证具体反馈公式、增益或足端承载布局为何失稳。评分器未报告 worst_gait_err，且 10 fps 视频不足以精确量化足高误差或相位跟踪；该指标填 null。"
+ },
+ "summary_zh": "最终评分回合中，Spot 已用四足交替支撑行进，并在视频约 8.0–8.8 s 转入后退，但视频 13.6–14.0 s 出现明显侧倾，在 700/1000 控制步触发跌倒终止，“按指令行进”节点未完成（0/1）。现象为全身失衡，结合最终回合传感读数及超阈值的速度、偏航误差，根因归为控制。缺少逐帧状态与完整控制器修改内容，无法细分失稳机制，评分器亦未报告足高误差。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 8,
+    "summary_zh": "终局回合 11 在控制步 700（视频 14.0 s）因 fell 终止，节点未达成、断点、平衡现象与控制根因均有 log step 79/80 的服务读数（reason=fell、重力 z=-0.847<0.85）和视频支持；时钟换算 700×0.02=14.0 s 与视频长度一致。未发现需改动的错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:22:12Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--n9mu8gq/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--n9mu8gq/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "a8cc70ef141b0f1aba061286b14ca097db5ecb8e3e7b555d7f8db68144a5f714",
+   "video_sha256": "921961e429efcefd17785fe20931cb83c00fae9623007aa5135be7e42531a909",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}
diff --git a/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--rw2kqfa.json b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--rw2kqfa.json
new file mode 100644
index 0000000000000000000000000000000000000000..0f1f1a93843bafdd5fdde705fe66c72ed4c87484
--- /dev/null
+++ b/phase2/mujoco-playground/spot_joystick_gait_tracking/trajectories/mujoco_playground_spot_joystick_gait_tracking-standard--rw2kqfa.json
@@ -0,0 +1,206 @@
+{
+ "schema": "efe-trajectory-labels/1",
+ "taxonomy": "v2",
+ "runId": "mujoco_playground_spot_joystick_gait_tracking-standard--rw2kqfa",
+ "taskKey": "mujoco-playground/spot_joystick_gait_tracking",
+ "dag_sha256": "22e6e8628238710b795d2a4684cc3564332dcf288461487eb66d709b89147299",
+ "outcome": {
+  "success": 0,
+  "reward": 0.0,
+  "source": "catalog",
+  "raw": {
+   "success": 0,
+   "live_success": 0,
+   "replay_success": 0,
+   "deterministic": 1,
+   "live_matches_replay": 1,
+   "finalized": 1,
+   "n_actions": 28,
+   "sim_steps": 28,
+   "physics_steps": 140,
+   "steps": 733,
+   "elapsed_s": 1420.7,
+   "video_rendered": 1,
+   "fell": 1,
+   "worst_lin_err": 1.3165,
+   "worst_yaw_err": 0.3904,
+   "end_tilt_deg": -1.0,
+   "end_imu_height": -1.0,
+   "end_pose_angle_deg": -1.0,
+   "grader_error": 0,
+   "instance_mismatch": 0,
+   "infra_error": 0,
+   "no_verdict": 0,
+   "missing_log": 0,
+   "gave_up": 0,
+   "sim_error": 0,
+   "sim_startups": 1,
+   "resumes": 0,
+   "resets": 10,
+   "reset_budget": 10,
+   "episodes": 11,
+   "resume_state_mismatch": 0,
+   "resume_max_joint_dev": 0.0,
+   "resource_enforced": 0,
+   "limits_violation_observed": 0,
+   "resource_monitor_missing": 0,
+   "peak_gpu_memory_gb": 0,
+   "peak_memory_gb": 0.31,
+   "peak_processes": 9,
+   "peak_disk_gb": 0.241,
+   "resource_emergency_kills": 0,
+   "resource_oom_kills": 0,
+   "resource_host_oom_kills": 0,
+   "resource_service_oom_kills": 0
+  }
+ },
+ "video": {
+  "kind": "hf",
+  "url": "https://huggingface.co/datasets/MLL-Lab/embodied-exam-media/resolve/main/runs/mujoco_playground_spot_joystick_gait_tracking-standard--rw2kqfa/replay.mp4",
+  "duration_s": 0.6,
+  "fps": 25.0,
+  "frames": 15,
+  "sha256": "21e2c6d898e9390773e2dbe70b1fd42ff2501bfaf6a37486438a3b2ede04f157"
+ },
+ "state": null,
+ "progress": {
+  "method": "visual",
+  "denominator": 1,
+  "base": 0,
+  "peak": 0,
+  "final": 0,
+  "nodes": {
+   "commanded_gait_completed": {
+    "method": "visual",
+    "intervals": [],
+    "final": false,
+    "evidence": [
+     {
+      "t": 0.24,
+      "log_step": null,
+      "text_zh": "观察：已逐帧检查视频 0.00–0.56 s 的全部15帧；Spot从四足站立开始屈伸腿部，足端有落地支撑和离地摆动，躯干随后逐渐侧倾。整段录像没有完成持续行进、后退及恢复前进的完整指令日程。"
+     },
+     {
+      "t": 0.56,
+      "log_step": 47,
+      "text_zh": "观察：log step 47 中机器人服务先报告重置至 episode=11、control_steps=0、resets_left=0，执行最终控制脚本后报告 control_steps=28、done=True、reason=fell；run.json 的评分字段 n_actions=28、fell=1、live_matches_replay=1 与此一致。该回合远未执行节点要求的1000个控制步。"
+     }
+    ],
+    "contact_actual": [
+     {
+      "who": "腿/躯干",
+      "how": "支撑",
+      "note_zh": "观察：视频 0.00 s 四个足端落在方格地面上；视频 0.12–0.56 s Spot屈伸四腿，部分足端离地、其余足端继续触地承重，随后躯干侧倾。未见躯干本身触地。"
+     }
+    ]
+   }
+  },
+  "constraints": {}
+ },
+ "native_metrics": [
+  {
+   "name": "worst_lin_err",
+   "unit": "m/s",
+   "threshold": 0.2,
+   "direction": "lower_better",
+   "value": 1.3165
+  },
+  {
+   "name": "worst_yaw_err",
+   "unit": "rad/s",
+   "threshold": 0.3,
+   "direction": "lower_better",
+   "value": 0.3904
+  },
+  {
+   "name": "worst_gait_err",
+   "unit": "m",
+   "threshold": 0.05,
+   "direction": "lower_better",
+   "value": null
+  }
+ ],
+ "failure": {
+  "breakpoint": {
+   "ref": "commanded_gait_completed",
+   "t": 0.48,
+   "log_step": 47,
+   "where": "操作中",
+   "evidence_zh": "观察：视频 0.36–0.56 s，Spot在启动步态过程中躯干侧倾加重，末帧仍倾斜且未恢复；log step 47 的第11回合由机器人服务以 reason=fell 在28个控制步后终止。失败发生在按指令行进期间。"
+  },
+  "phenomena": [
+   {
+    "cat": "全身",
+    "item": "平衡",
+    "t": 0.48,
+    "log_step": 47,
+    "evidence_zh": "观察：视频 0.24 s 躯干已开始倾斜，视频 0.48–0.56 s 可见更明显的侧倾和腿部支撑姿态失衡；log step 47 服务终态为 done=True、reason=fell，评分字段 fell=1。判据是躯干过度倾斜，画面并未显示机器人已躺倒在地。"
+   }
+  ],
+  "root_causes": [
+   {
+    "cat": "控制",
+    "item": "控制",
+    "source_zh": "日志",
+    "evidence_zh": "推断：log step 41 的控制代码依据机器人IMU重力方向、角速度和线速度生成足端轨迹及姿态修正，再提交关节位置目标；log step 47 调整滤波、前馈和修正增益后运行最终版本，服务仍在28个控制步后以fell终止，视频 0.36–0.56 s 也显示侧倾持续加重。该步态控制未能稳定机身支撑；现有证据不能进一步区分落足布局、轨迹跟踪或反馈参数的具体责任。"
+   }
+  ],
+  "gap_zh": "没有逐帧状态日志；log step 47 只输出最后回合的起止状态，没有该回合逐步IMU、足地接触力或实际关节跟踪数据。log step 46 的诊断属于已重置的第10回合，不能当作第11回合的传感器记录；agent在log step 51 提出的支撑足拖曳等解释尚未独立核实。评分字段未报告worst_gait_err，不能量化最终回合足高误差。",
+  "unmet": {
+   "commanded_gait_completed": false
+  }
+ },
+ "summary_zh": "观察：评分采用log step 47 的第11回合；视频 0.00–0.56 s 显示Spot已启动腿部运动和足地支撑，但躯干逐渐侧倾，服务在28个控制步后以fell终止，“按指令行进”完成0/1。推断：最终步态控制未能维持机身稳定，断点为该节点操作中，现象为全身平衡失效。最后回合缺少逐步传感器与接触力记录，无法确认具体失稳机制，足高误差也未报告。",
+ "conf": 8,
+ "review": {
+  "status": "agreed",
+  "rounds": 1,
+  "reviewers": [
+   {
+    "engine": "claude",
+    "model": "claude-sonnet-5-5",
+    "round": 1,
+    "verdict": "agree",
+    "conf": 7,
+    "summary_zh": "视频仅0.6 s/15帧，Spot屈伸腿部后躯干侧倾；log step 47 服务以 fell 终止（28控制步），与 run.json 评分字段一致。节点未达成、断点、平衡现象和控制根因（停在“控制”未细分）均有证据支持，未见错误。",
+    "n_issues": 0,
+    "n_major": 0
+   }
+  ],
+  "open_zh": []
+ },
+ "provenance": {
+  "prompt": "prompts/03_trajectory.md",
+  "prompt_sha256": "2385b35930667330fc4ea181a15ea0818b49cfeae0df1700b8b08c988e72ad8e",
+  "engine": "codex",
+  "model": "gpt-6.1-sol",
+  "effort": "high",
+  "created": "2026-10-30T03:09:30Z",
+  "review_prompt": "prompts/04_trajectory_review.md",
+  "review_prompt_sha256": "9afbc4ac19b1e1cf03732c8961ec46a618d682ee229343f66e3316794b14f9ba",
+  "reconcile_prompt": "prompts/05_reconcile.md",
+  "reconcile_prompt_sha256": "8c827af3ae3f77d09229399fb41283c50a8f725302cbe3b263b8051213dde2b2",
+  "calls": [
+   {
+    "stage": "traj",
+    "round": 0,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--rw2kqfa/03_label"
+   },
+   {
+    "stage": "traj_review",
+    "round": 1,
+    "ok": true,
+    "space": "work/runs/mujoco_playground_spot_joystick_gait_tracking-standard--rw2kqfa/04_review_r1"
+   }
+  ],
+  "review_engine": "claude",
+  "review_model": "claude-sonnet-5-5",
+  "inputs": {
+   "run_json_sha256": "83ec0bab1a0c12deb3f3f37be0bfaf08e13e9b4405d94c1e9b2580a14645a7f1",
+   "video_sha256": "21e2c6d898e9390773e2dbe70b1fd42ff2501bfaf6a37486438a3b2ede04f157",
+   "states_sha256": null,
+   "video_choice": "HF replay.mp4"
+  }
+ }
+}