diff --git a/masters/roboolympics-fun/alpine_slalom/meta.json b/masters/roboolympics-fun/alpine_slalom/meta.json
index 6a0add1928bfa338182b47145242d864a85df27e..930c64637d3d678147c92607ee8abc96caa536e9 100644
--- a/masters/roboolympics-fun/alpine_slalom/meta.json
+++ b/masters/roboolympics-fun/alpine_slalom/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-alpine-slalom-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-alpine-slalom-i00-privileged/embodied-olympics-fun-alpine-sla__eV2gD7X",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; snow venue: a snow shader (bright blue-white, rough, slight subsurface, fine bump) on the course's snow geoms, snow-white valley floor; a chase camera: per-frame fit of the robot (height included, so jumps and spins stay framed), at least 6 m away, 24 degree elevation, smoothed in time",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; snow venue: a snow shader (bright blue-white, rough, slight subsurface, fine bump) on the course's snow geoms, snow-white valley floor; a director camera: a 0.0 s establishing shot framing the robot and the task's key equipment (gate0_turn, gate10_turn, gate11_turn, gate12_turn, gate13_turn, gate14_turn, gate15_turn, gate16_turn, gate17_turn, gate18_turn, gate19_turn, gate1_turn, gate20_turn, gate21_turn, gate22_turn, gate23_turn, gate24_turn, gate25_turn, gate26_turn, gate27_turn, gate28_turn, gate29_turn, gate2_turn, gate30_turn, gate31_turn, gate3_turn, gate4_turn, gate5_turn, gate6_turn, gate7_turn, gate8_turn, gate9_turn; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 24.0-degree elevation, a slow eased 0.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -90,16 +90,56 @@
   "az_rule": "robot-travel-70",
   "el_deg": 24.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^gate\\d+_turn",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 0.0,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "gate0_turn",
+   "gate10_turn",
+   "gate11_turn",
+   "gate12_turn",
+   "gate13_turn",
+   "gate14_turn",
+   "gate15_turn",
+   "gate16_turn",
+   "gate17_turn",
+   "gate18_turn",
+   "gate19_turn",
+   "gate1_turn",
+   "gate20_turn",
+   "gate21_turn",
+   "gate22_turn",
+   "gate23_turn",
+   "gate24_turn",
+   "gate25_turn",
+   "gate26_turn",
+   "gate27_turn",
+   "gate28_turn",
+   "gate29_turn",
+   "gate2_turn",
+   "gate30_turn",
+   "gate31_turn",
+   "gate3_turn",
+   "gate4_turn",
+   "gate5_turn",
+   "gate6_turn",
+   "gate7_turn",
+   "gate8_turn",
+   "gate9_turn"
+  ],
   "distance_m": [
    6.24,
-   9.304
+   15.771
   ],
   "robot_root": "pelvis",
   "travel_m": 307.125,
   "focus": {
-   "robot+object": 742,
-   "robot": 18
+   "robot": 18,
+   "robot+object": 83,
+   "robot+object+equip": 659
   },
   "objects": {
    "gate1_turn": 0.0,
@@ -128,7 +168,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 4.72,
+ "seconds_per_frame": 3.65,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
@@ -136,5 +176,5 @@
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:37-0500"
+ "created": "2026-10-25T09:57:13-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/alpine_slalom/poster.png b/masters/roboolympics-fun/alpine_slalom/poster.png
index dd4e30c8895c94dbd6bc444f2b1777e6981a2783..156e566b85fabe395a0a9bbb7925510889a47a78 100644
--- a/masters/roboolympics-fun/alpine_slalom/poster.png
+++ b/masters/roboolympics-fun/alpine_slalom/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9eddb864bc3a96372cf583bf7e9fad3f1d3ffbd2889f72335a581166795cf5a8
-size 3036918
+oid sha256:3fbfa3b79be207e76320a9d5483ad03339558fc41803dc784db83cf68fbb12d5
+size 3015010
diff --git a/masters/roboolympics-fun/alpine_slalom/video.mp4 b/masters/roboolympics-fun/alpine_slalom/video.mp4
index 4085c473cab225f8c53535c4558d5e43c0c5f0bd..961e41e53565f609538ddb450becf8d604d0dc5b 100644
--- a/masters/roboolympics-fun/alpine_slalom/video.mp4
+++ b/masters/roboolympics-fun/alpine_slalom/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:43c60360d5509c0445d422f0ff9ae6618a2b2cabc9aa9dc28ab2903cd655c8f9
-size 12693590
+oid sha256:36491130d419af32c0fdf4ce932cdf67892eb91b6bf63c1c9db56be020ae2caa
+size 10388552
diff --git a/masters/roboolympics-fun/backflip/meta.json b/masters/roboolympics-fun/backflip/meta.json
index 5b585041434195724f427939531012538746a00a..55435c09e04a97c761a7dc18859cc3e851520b39 100644
--- a/masters/roboolympics-fun/backflip/meta.json
+++ b/masters/roboolympics-fun/backflip/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-backflip-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-backflip-i00-privileged/embodied-olympics-fun-backflip-i__L6om7HM",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,10 +57,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.094,
-   3.358
+   3.283,
+   3.556
   ],
   "robot_root": "pelvis",
   "travel_m": 0.813,
@@ -77,7 +83,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.62,
+ "seconds_per_frame": 2.0,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -85,5 +91,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:37-0500"
+ "created": "2026-10-25T09:57:13-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/backflip/poster.png b/masters/roboolympics-fun/backflip/poster.png
index 605635917cdf8166cf2752adeb2e3d9fa5889df2..ac9ebb10cfc5a93d0877caeacb93498c8dbcb252 100644
--- a/masters/roboolympics-fun/backflip/poster.png
+++ b/masters/roboolympics-fun/backflip/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eb44a96f36ebbe06ea2963219cdea3fc168b5cae53d3ec30b62d4aa063c37fc8
-size 3111651
+oid sha256:00a52d70fc7c1bd56bab117f4dc79bf89e1b1d408b8c36177143e8d0776799cf
+size 3097991
diff --git a/masters/roboolympics-fun/backflip/video.mp4 b/masters/roboolympics-fun/backflip/video.mp4
index d5a04ca7e2a63eb2dcbc95010423f195300e5ef0..1befc1757d80821575324eaab3ed44b566f3e04a 100644
--- a/masters/roboolympics-fun/backflip/video.mp4
+++ b/masters/roboolympics-fun/backflip/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f9c39256f792c5d233bc2545f7807b7157727972fe35cb0cea667e6253e12067
-size 4084672
+oid sha256:9efc9de6d9a4a06cd61edac50f1f8e626b4aeb0f9d3c47734dcd31351a9dcdce
+size 3951934
diff --git a/masters/roboolympics-fun/badminton_serve/meta.json b/masters/roboolympics-fun/badminton_serve/meta.json
index 4dd412922623363f8387ae3dcc83dcd15d5908a1..25a8382c9e007ab3713e79596ddd598eaed34b4b 100644
--- a/masters/roboolympics-fun/badminton_serve/meta.json
+++ b/masters/roboolympics-fun/badminton_serve/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-badminton-serve-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-badminton-serve-i00-privileged/embodied-olympics-fun-badminton__dU3SuCW",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (net, net_cord, net_seg0, net_seg1, net_seg10, net_seg11, net_seg2, net_seg3, net_seg4, net_seg5, net_seg6, net_seg7, net_seg8, net_seg9; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,20 +56,64 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -100.81,
-  "az_rule": "object-travel-115",
+  "az_deg": -110.42,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^net",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "net",
+   "net_cord0",
+   "net_cord1",
+   "net_cord10",
+   "net_cord11",
+   "net_cord2",
+   "net_cord3",
+   "net_cord4",
+   "net_cord5",
+   "net_cord6",
+   "net_cord7",
+   "net_cord8",
+   "net_cord9",
+   "net_seg0_0",
+   "net_seg0_1",
+   "net_seg10_0",
+   "net_seg10_1",
+   "net_seg11_0",
+   "net_seg11_1",
+   "net_seg1_0",
+   "net_seg1_1",
+   "net_seg2_0",
+   "net_seg2_1",
+   "net_seg3_0",
+   "net_seg3_1",
+   "net_seg4_0",
+   "net_seg4_1",
+   "net_seg5_0",
+   "net_seg5_1",
+   "net_seg6_0",
+   "net_seg6_1",
+   "net_seg7_0",
+   "net_seg7_1",
+   "net_seg8_0",
+   "net_seg8_1",
+   "net_seg9_0",
+   "net_seg9_1"
+  ],
   "distance_m": [
-   3.576,
-   6.469
+   8.644,
+   8.674
   ],
   "robot_root": "pelvis",
   "travel_m": 0.242,
   "focus": {
-   "robot+object": 259,
-   "robot": 40
+   "robot+object+equip": 259,
+   "robot+equip": 40
   },
   "objects": {
    "shuttle": 5.81,
@@ -84,7 +128,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 3.03,
+ "seconds_per_frame": 2.23,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -92,5 +136,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:38-0500"
+ "created": "2026-10-25T09:57:13-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/badminton_serve/poster.png b/masters/roboolympics-fun/badminton_serve/poster.png
index 1ade24eed48b6ce743f7f6f244f1cca06d92a39d..3563c02677d7b3e8c13f9291ce14b9f91a0df8d0 100644
--- a/masters/roboolympics-fun/badminton_serve/poster.png
+++ b/masters/roboolympics-fun/badminton_serve/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6318a8e21a3950506ca30ce5eccb3c30e39e2e24205317238b7cd3bf01012e28
-size 3209522
+oid sha256:1113f863fb9e47d762f6551c7448b94fe45962a2556592666af19b5b720e799a
+size 3267099
diff --git a/masters/roboolympics-fun/badminton_serve/video.mp4 b/masters/roboolympics-fun/badminton_serve/video.mp4
index 3f4a8fe796264bf193541014aade24df159a31e6..fe766554f4792c2e6566a4ee00f2d000d117e08a 100644
--- a/masters/roboolympics-fun/badminton_serve/video.mp4
+++ b/masters/roboolympics-fun/badminton_serve/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e139dff6733b24ac6bfc6a8baf4b85417431909909ba556a02ee8e95e90cb815
-size 6319444
+oid sha256:d0c872197d8695c24ab40360f3eb9006794eeac047caaa00cc7e0e3104d806ad
+size 5684610
diff --git a/masters/roboolympics-fun/balance_beam/meta.json b/masters/roboolympics-fun/balance_beam/meta.json
index bc26e87492f64b63904a6603d8ca5bdb631e9c1b..81da2a7df5cd63823d00af35d86342b74f599476 100644
--- a/masters/roboolympics-fun/balance_beam/meta.json
+++ b/masters/roboolympics-fun/balance_beam/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-balance-beam-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-balance-beam-i00-privileged/embodied-olympics-fun-balance-be__FJ5XCVd",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (beam, beam_base, beam_leg; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,15 +56,27 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^beam",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "beam",
+   "beam_base0",
+   "beam_base1",
+   "beam_leg0",
+   "beam_leg1"
+  ],
   "distance_m": [
-   3.087,
-   3.22
+   6.682,
+   8.181
   ],
   "robot_root": "pelvis",
   "travel_m": 4.451,
   "focus": {
-   "robot": 720
+   "robot+equip": 720
   },
   "objects": {}
  },
@@ -76,7 +88,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.68,
+ "seconds_per_frame": 2.05,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -84,5 +96,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:38-0500"
+ "created": "2026-10-25T09:57:13-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/balance_beam/poster.png b/masters/roboolympics-fun/balance_beam/poster.png
index 2061004a7859f1e21dfb33ac37daab4714a7d960..7e7d1cda81626cf429b052cf9c2eea61baadc28d 100644
--- a/masters/roboolympics-fun/balance_beam/poster.png
+++ b/masters/roboolympics-fun/balance_beam/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:20b1f9384aea1bca2899f1731f88f40a4320b74ed93004d993ca23b4a110021b
-size 3100072
+oid sha256:163b0bc126c0f90586d8c0efd608b3547efd00295972a371a37e72db97e277b6
+size 3062709
diff --git a/masters/roboolympics-fun/balance_beam/video.mp4 b/masters/roboolympics-fun/balance_beam/video.mp4
index 011d7a2404bb8789f7f96a66269d160d8831f268..fcbb5ccb24fda6a23f64ca691d9da1d6f61848f4 100644
--- a/masters/roboolympics-fun/balance_beam/video.mp4
+++ b/masters/roboolympics-fun/balance_beam/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b3cc3f0294c632d4838dc59f7d742a0875ebc3c2ed19d141e1a16a1aeb0ea08b
-size 11572565
+oid sha256:00eb47c62358df38f39f0208002fe82507d10ef55df9b2c031b73abb2722171e
+size 8600909
diff --git a/masters/roboolympics-fun/balance_board/meta.json b/masters/roboolympics-fun/balance_board/meta.json
index 14e8785edd1ddf854d284dc968303d28d093e7d3..db8617ae61cc765aea9c92959ffbc360e89e0e16 100644
--- a/masters/roboolympics-fun/balance_board/meta.json
+++ b/masters/roboolympics-fun/balance_board/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-balance-board-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-balance-board-i00-privileged/embodied-olympics-fun-balance-bo__RkHCQBb",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,10 +56,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.737,
-   3.927
+   3.962,
+   4.161
   ],
   "robot_root": "pelvis",
   "travel_m": 0.039,
@@ -79,7 +85,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.58,
+ "seconds_per_frame": 1.99,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -87,5 +93,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:38-0500"
+ "created": "2026-10-25T09:57:13-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/balance_board/poster.png b/masters/roboolympics-fun/balance_board/poster.png
index ed8d60c480147f7db1cca3c228a11165f3383a11..4cba09a30d947074e05512944052c994901dc5a7 100644
--- a/masters/roboolympics-fun/balance_board/poster.png
+++ b/masters/roboolympics-fun/balance_board/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:99ddf6692308f33f64b438c76ed742eb3e1f0228ddfe8f702add9cb0dcfb91b0
-size 3105285
+oid sha256:6e63edd740ebd09846b9a3285be220478966a9959833e1f157619321a561f3e2
+size 3099425
diff --git a/masters/roboolympics-fun/balance_board/video.mp4 b/masters/roboolympics-fun/balance_board/video.mp4
index c8589acd2dcfd45fa3de877374eb4c0e432ad52b..5b2f212f3d7e00216b5494302a2b71767d2a792a 100644
--- a/masters/roboolympics-fun/balance_board/video.mp4
+++ b/masters/roboolympics-fun/balance_board/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b5671816567883b76efe463240635c7a0e9180c6da0177c90d62a908a532a637
-size 8246896
+oid sha256:07df53821677a9b4cfd7d3bfa1990fed7fcb32053a9ca03e9fa696b8010c2f69
+size 8112913
diff --git a/masters/roboolympics-fun/basketball_dribble/meta.json b/masters/roboolympics-fun/basketball_dribble/meta.json
index b71dfc687314e3cb5c8e2bc37ea573277827acb7..5866f3ad27c3355317cb5f2a046f87f945d0a686 100644
--- a/masters/roboolympics-fun/basketball_dribble/meta.json
+++ b/masters/roboolympics-fun/basketball_dribble/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-basketball-dribble-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-basketball-dribble-i00-privileged/embodied-olympics-fun-basketball__yqWK2fP",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -58,10 +58,16 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   4.298,
-   4.388
+   4.555,
+   4.647
   ],
   "robot_root": "pelvis",
   "travel_m": 13.877,
@@ -81,7 +87,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.86,
+ "seconds_per_frame": 2.06,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
@@ -89,5 +95,5 @@
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:38-0500"
+ "created": "2026-10-25T09:57:13-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/basketball_dribble/poster.png b/masters/roboolympics-fun/basketball_dribble/poster.png
index fab084b9155bd982b1603e66a5b58d95148ae078..9b54a13a8e82ad5ad32f6553b1aeb24be37dd70b 100644
--- a/masters/roboolympics-fun/basketball_dribble/poster.png
+++ b/masters/roboolympics-fun/basketball_dribble/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:391e7d5d80e1a7cb92dbe2a82e33618b997171e0a2febe691f13e1086a2d7439
-size 3144397
+oid sha256:738486526aeca6a5126d335beb0d30fd34c2217eaeb3e1032a4b84894cfdd0f8
+size 3139272
diff --git a/masters/roboolympics-fun/basketball_dribble/video.mp4 b/masters/roboolympics-fun/basketball_dribble/video.mp4
index 026e2cf8aac43aa969431617c0a35f3dffc7aa33..e7eeb45f933dc3fac9c4b6321a3a36e629195951 100644
--- a/masters/roboolympics-fun/basketball_dribble/video.mp4
+++ b/masters/roboolympics-fun/basketball_dribble/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4ba3ff70303fe1825d1f8c4488897762a9a6789a7cc4399e4eff0cb596e96077
-size 11305885
+oid sha256:dead69d34d5812e8307c19b0700db5b488f0163633f9efbd75162503afc45b8e
+size 10840379
diff --git a/masters/roboolympics-fun/bowling/meta.json b/masters/roboolympics-fun/bowling/meta.json
index c6a77b351fefafde6de65e46df9a9c77465bf069..bda1a33a29671b34a1f3d18e8646a502e30e7a2b 100644
--- a/masters/roboolympics-fun/bowling/meta.json
+++ b/masters/roboolympics-fun/bowling/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-bowling-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-bowling-i00-privileged/embodied-olympics-fun-bowling-i0__nD6fgoT",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule object-travel-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -59,19 +59,25 @@
   "az_rule": "object-travel-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   4.697,
-   36.677
+   4.972,
+   36.696
   ],
   "robot_root": "pelvis",
   "travel_m": 0.086,
   "focus": {
+   "object": 173,
    "robot+landed": 81,
-   "robot+object": 49,
+   "robot(incoming)": 10,
    "robot": 29,
-   "object": 173,
-   "robot(incoming)": 10
+   "robot+object": 49
   },
   "objects": {
    "pin1": 0.59,
@@ -95,13 +101,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.97,
+ "seconds_per_frame": 2.1,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 38 (same scene, camera and settings as the video)",
+ "poster": "frame 38 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:38-0500"
+ "created": "2026-10-25T09:57:13-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/bowling/poster.png b/masters/roboolympics-fun/bowling/poster.png
index d72608f26f4d5776a7c3da71c3be4d0da882d9b6..c434cea2649ea0f83f3a2234f7d05cda8ad14d39 100644
--- a/masters/roboolympics-fun/bowling/poster.png
+++ b/masters/roboolympics-fun/bowling/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c9c1d981c877d1564e218c009658f8f5ae790bfd115551bc11ae9ad8bb2bfcbd
-size 3132893
+oid sha256:d4f9f1a5916b9cc7b9d6a062984b5ce15bab0cffa83cb4318feca1d21c50374f
+size 3129430
diff --git a/masters/roboolympics-fun/bowling/video.mp4 b/masters/roboolympics-fun/bowling/video.mp4
index cb03df1a2a1aad02745a2d570ad362e3265b8343..789ac57ee1171d47821d6b72d34e75f0e8e25ea1 100644
--- a/masters/roboolympics-fun/bowling/video.mp4
+++ b/masters/roboolympics-fun/bowling/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d600112ba6c9fa9ff5651698830de3ad4142a4eda94e1ae2811c7972ad859a44
-size 7445595
+oid sha256:b6e70fb0d6ae07b6e2ad81ce2101e0bb5d3d81ec09d237b4a0c5ebdc6fda2f38
+size 7242968
diff --git a/masters/roboolympics-fun/box_jump/meta.json b/masters/roboolympics-fun/box_jump/meta.json
index ab200a5ec0e38d1277e147ef72254ef1fd622280..ba5e07db93b80efaf37b8c4cb57c4912b26293ba 100644
--- a/masters/roboolympics-fun/box_jump/meta.json
+++ b/masters/roboolympics-fun/box_jump/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-box-jump-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-box-jump-i00-privileged/embodied-olympics-fun-box-jump-i__o9U84U8",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,10 +57,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   5.354,
-   5.638
+   5.673,
+   5.981
   ],
   "robot_root": "pelvis",
   "travel_m": 0.958,
@@ -79,7 +85,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.68,
+ "seconds_per_frame": 2.03,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -87,5 +93,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:38-0500"
+ "created": "2026-10-25T09:57:13-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/box_jump/poster.png b/masters/roboolympics-fun/box_jump/poster.png
index 3f0318bc7e9b9172f62b9465d1efde205a0fe9e5..0ce2938c5372b15bd97969430bcf83edb28ad597 100644
--- a/masters/roboolympics-fun/box_jump/poster.png
+++ b/masters/roboolympics-fun/box_jump/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e78d9e90ba155cd4f925170fa20b8feae37ecdff414eb42a67f06cf4fd4bd797
-size 3046876
+oid sha256:2bf770f3ac31b9c0067fa10d69f6826adf0cad6fcf41137e3b49371aa5cb502b
+size 3036577
diff --git a/masters/roboolympics-fun/box_jump/video.mp4 b/masters/roboolympics-fun/box_jump/video.mp4
index 57e4399a65ea3788085922ddade847d021ace8e0..1c35054a3ec1495e2b2b1b81a6f522a18831ac6a 100644
--- a/masters/roboolympics-fun/box_jump/video.mp4
+++ b/masters/roboolympics-fun/box_jump/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c5b723c54603a03565367ec78d3927231dd6646a899e4ce70bc415911ed4370c
-size 3383986
+oid sha256:cb7b8aae20beaf529a9b34935037c99ad54f2b26d86c4326004e245975616f27
+size 3291728
diff --git a/masters/roboolympics-fun/boxing/meta.json b/masters/roboolympics-fun/boxing/meta.json
index aad1b9466e2d7e79cd927f1767796bd0e2643be4..02e5c1a338a108f31dd78fb78e39e50262fa7c54 100644
--- a/masters/roboolympics-fun/boxing/meta.json
+++ b/masters/roboolympics-fun/boxing/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-boxing-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-boxing-i00-privileged/embodied-olympics-fun-boxing-i00__pTjF4SX",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,10 +57,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   4.872,
-   5.158
+   5.136,
+   5.448
   ],
   "robot_root": "pelvis",
   "travel_m": 0.762,
@@ -79,7 +85,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.66,
+ "seconds_per_frame": 2.03,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -87,5 +93,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:38-0500"
+ "created": "2026-10-25T09:57:13-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/boxing/poster.png b/masters/roboolympics-fun/boxing/poster.png
index 854ad0d2fcb691ca53d095f35233b12323ff232a..d8b1e33ca277c5e3718d15de3fa83338632ddb6f 100644
--- a/masters/roboolympics-fun/boxing/poster.png
+++ b/masters/roboolympics-fun/boxing/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:56ba6fec874e0f8cc9719b27b2ec8dbbc7f3592c3badbebb550969babaa8ca87
-size 3155145
+oid sha256:277356329a1b6eccc15d7323e24f9fbd6619d8cc850382c1032485cdd2e274c2
+size 3144204
diff --git a/masters/roboolympics-fun/boxing/video.mp4 b/masters/roboolympics-fun/boxing/video.mp4
index e25ae10e802796fbac0cc5a94f1e4a5f83e67bd1..53ff766dc783de9c59d4fba41dda9850cb76a09d 100644
--- a/masters/roboolympics-fun/boxing/video.mp4
+++ b/masters/roboolympics-fun/boxing/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ed2d114a79d4593712748a225191dec068cfbe47d9c4ff2cd0850c0c56015666
-size 4041069
+oid sha256:206fdd478f09dadbb1c9fe4462253fba184dffc99a568a150444f669e769f3ed
+size 3962335
diff --git a/masters/roboolympics-fun/breaking_windmill/meta.json b/masters/roboolympics-fun/breaking_windmill/meta.json
index cad8f4603a114481fd6a88b1234374d918433641..3751593baf6cd3c887cde03cdde15139e430738d 100644
--- a/masters/roboolympics-fun/breaking_windmill/meta.json
+++ b/masters/roboolympics-fun/breaking_windmill/meta.json
@@ -7,7 +7,7 @@
  "source": "scene-orbit",
  "why_no_video": "No evaluated agent has a success on this task version yet. Site rule (boards 0161 / 0252): show GPT-6 Astra's failed attempt with the most progress instead. This video is that attempt replayed exactly (not a solved run): the task's own referee at the end says it failed, as the verifier did.",
  "shows": "agent-attempt (failed):codex-gpt6_astra_quick-high-embodied-olympics-fun-breaking-windmill-i00-privileged/embodied-olympics-fun-breaking-w__mxtoRx8",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,10 +57,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   2.072,
-   3.389
+   2.181,
+   3.411
   ],
   "robot_root": "pelvis",
   "travel_m": 0.655,
@@ -77,7 +83,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.64,
+ "seconds_per_frame": 2.02,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
@@ -85,5 +91,5 @@
  "trial_choice": "no success on this task version; GPT-6 Astra's failed attempt (most progress)",
  "poster": "last frame (end of the failed attempt)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:39-0500"
+ "created": "2026-10-25T09:57:13-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/breaking_windmill/poster.png b/masters/roboolympics-fun/breaking_windmill/poster.png
index f4d2c1d69fd7a7cdc9eb8006174983d2d027dca1..37593c803a4e4653ab879061160ad3918514eeb9 100644
--- a/masters/roboolympics-fun/breaking_windmill/poster.png
+++ b/masters/roboolympics-fun/breaking_windmill/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1118a0dfac387175aaf8c838d1efc2dca72a8a34ae0875c5f534b6d55f3c1499
-size 3052344
+oid sha256:f0654778f61d10e1111804226f736108f699ab05aedf335b3509f1561f1b340f
+size 3044501
diff --git a/masters/roboolympics-fun/breaking_windmill/video.mp4 b/masters/roboolympics-fun/breaking_windmill/video.mp4
index 7adddc40a0a3a0c4ed09933cb0cb3e643a8c8a23..49eb6def872e1102a127e5217be1a95491f6e6fe 100644
--- a/masters/roboolympics-fun/breaking_windmill/video.mp4
+++ b/masters/roboolympics-fun/breaking_windmill/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d15b4138e5fae82de7c099ee1833b5ff6ead3771b547b78dffd8e394cc4d646f
-size 15904376
+oid sha256:1b1aed767aef7ce00b17ac0cce3b53da61da24338074785fc4f7b02576b85b1c
+size 15520522
diff --git a/masters/roboolympics-fun/catch/meta.json b/masters/roboolympics-fun/catch/meta.json
index d1631b9f8a3279347e1f9037d54e66a351ad6a22..60475e0454e3c2269f5a7fd0adc0ee87bd43e702 100644
--- a/masters/roboolympics-fun/catch/meta.json
+++ b/masters/roboolympics-fun/catch/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-catch-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-catch-i00-privileged/embodied-olympics-fun-catch-i00__qBKtgfp",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule object-travel-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -55,17 +55,23 @@
   "az_rule": "object-travel-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   4.356,
-   16.306
+   4.615,
+   16.326
   ],
   "robot_root": "pelvis",
   "travel_m": 1.66,
   "focus": {
-   "robot+object": 48,
+   "robot(incoming)": 86,
    "robot": 33,
-   "robot(incoming)": 86
+   "robot+object": 48
   },
   "objects": {
    "ball": 38.23
@@ -79,7 +85,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 3.92,
+ "seconds_per_frame": 3.15,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
@@ -87,5 +93,5 @@
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:39-0500"
+ "created": "2026-10-25T09:57:14-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/catch/poster.png b/masters/roboolympics-fun/catch/poster.png
index e0c2128a34684173c75937960d9f79e72aa7ae9c..d1469687d5fa956836a39bf7fe987b6899d66d83 100644
--- a/masters/roboolympics-fun/catch/poster.png
+++ b/masters/roboolympics-fun/catch/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d4edd616756ca828a97062c9a7f9d14bac9671236591d4de49d11de8c5e6b59a
-size 3049411
+oid sha256:adb1048fd6163144508e2e93d463364285799098acdbce7cb7e3d409a15a0e97
+size 3044674
diff --git a/masters/roboolympics-fun/catch/video.mp4 b/masters/roboolympics-fun/catch/video.mp4
index e748a2f0e2f7a3a7771e73ece4b0823a18a856f5..e3694af3e2b30ebbb6563cd484f484a511af112a 100644
--- a/masters/roboolympics-fun/catch/video.mp4
+++ b/masters/roboolympics-fun/catch/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4fcf5e3a498b657b438460f17ac9e231f2076972e0ab9d577e309b8fcd5d895b
-size 3564063
+oid sha256:e333abe61b088acfb8a25b270015ebb3f6493af56a952eee32db92d04477a44f
+size 3537862
diff --git a/masters/roboolympics-fun/catch_and_shoot/meta.json b/masters/roboolympics-fun/catch_and_shoot/meta.json
index d34d00fffb4bef815599200a0a7cf9afa417f941..a4d818b0a9f7c8022137e1937f5829373e7b25dc 100644
--- a/masters/roboolympics-fun/catch_and_shoot/meta.json
+++ b/masters/roboolympics-fun/catch_and_shoot/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-catch-and-shoot-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-catch-and-shoot-i00-privileged/embodied-olympics-fun-catch-and__GZgBxmp",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (hoop; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -52,20 +52,28 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -155.62,
-  "az_rule": "object-travel-115",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^hoop_(rim|board|plate|net)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "hoop"
+  ],
   "distance_m": [
-   5.886,
-   16.553
+   13.13,
+   18.111
   ],
   "robot_root": "pelvis",
   "travel_m": 0.024,
   "focus": {
-   "robot+object": 217,
-   "robot": 5
+   "robot+equip": 5,
+   "robot+object+equip": 217
   },
   "objects": {
    "passer": 0.03,
@@ -80,13 +88,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.91,
+ "seconds_per_frame": 2.06,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
  "mode": "privileged",
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
- "poster": "frame 162 (same scene, camera and settings as the video)",
+ "poster": "frame 162 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:39-0500"
+ "created": "2026-10-25T09:57:14-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/catch_and_shoot/poster.png b/masters/roboolympics-fun/catch_and_shoot/poster.png
index b9c8d1b6bd5c36097e92dee6ee10f6184b6d6dcc..3e72e3678a750ed0e84169ad9e2bab59901551da 100644
--- a/masters/roboolympics-fun/catch_and_shoot/poster.png
+++ b/masters/roboolympics-fun/catch_and_shoot/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5cb18a0b34045617126dbc9bf727d299d5013342193e0bb9bad65df92ed5c07c
-size 3143831
+oid sha256:fa1836a160246e658ad7a0b1575ecf9d0e5af7d15da0577c3ca9a61e9df66e5d
+size 3121888
diff --git a/masters/roboolympics-fun/catch_and_shoot/video.mp4 b/masters/roboolympics-fun/catch_and_shoot/video.mp4
index c2e67a8154fad460cf5fa5dec6b9bb389112b89e..65931449e3d5794d7debfe103c113fa92a262659 100644
--- a/masters/roboolympics-fun/catch_and_shoot/video.mp4
+++ b/masters/roboolympics-fun/catch_and_shoot/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:28001c0fb9bd2e9d63658dc845f8755d317c2451ceacb388f69b45f908a86cda
-size 5684331
+oid sha256:5c73ca704c7df1b8fe319c2ecd41d49148ea7490fcee4d0fae53c34049b97acd
+size 4720084
diff --git a/masters/roboolympics-fun/clean_and_jerk/meta.json b/masters/roboolympics-fun/clean_and_jerk/meta.json
index 9125f384a7a69d1014543021143669adad742434..0ab0fb7172abbbc37f3f529499dcec1df985a7c5 100644
--- a/masters/roboolympics-fun/clean_and_jerk/meta.json
+++ b/masters/roboolympics-fun/clean_and_jerk/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-clean-and-jerk-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-clean-and-jerk-i00-privileged/embodied-olympics-fun-clean-and__HRohaub",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -60,16 +60,22 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   4.151,
-   6.408
+   4.396,
+   6.788
   ],
   "robot_root": "pelvis",
   "travel_m": 0.076,
   "focus": {
-   "robot": 191,
-   "robot+object": 136
+   "robot+object": 136,
+   "robot": 191
   },
   "objects": {
    "bar": 0.21
@@ -83,7 +89,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.64,
+ "seconds_per_frame": 2.03,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
@@ -91,5 +97,5 @@
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:39-0500"
+ "created": "2026-10-25T09:57:14-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/clean_and_jerk/poster.png b/masters/roboolympics-fun/clean_and_jerk/poster.png
index 2a3c11f8da1edddffeaa73e07b3abaed4a7872f0..861e0920ae157bc1f1a3194662597c4c2618c810 100644
--- a/masters/roboolympics-fun/clean_and_jerk/poster.png
+++ b/masters/roboolympics-fun/clean_and_jerk/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:18d98d68879eb9777501b8e346b1036397a08e7f540fcf2717d351b6996f041b
-size 3088548
+oid sha256:5ce61c85f020ece3e59206812880ffe906e609b519f77e54dcea4deb7baa611c
+size 3082184
diff --git a/masters/roboolympics-fun/clean_and_jerk/video.mp4 b/masters/roboolympics-fun/clean_and_jerk/video.mp4
index 35ecaae0f22b7f0aef36a300162817d38cee6ad1..61a1b07e5e8753a870ea70e3522476aeae397d41 100644
--- a/masters/roboolympics-fun/clean_and_jerk/video.mp4
+++ b/masters/roboolympics-fun/clean_and_jerk/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ad7cdfce402d206c89478423922feb7322dbd479f5e3da34ebceb3b96caa2b52
-size 5720370
+oid sha256:9777c53972d3cc65c57fd585affc725bef712f31b29e3972b6b5e7e63b7598fe
+size 5591664
diff --git a/masters/roboolympics-fun/crawl/meta.json b/masters/roboolympics-fun/crawl/meta.json
index 2db0bab191bd69b0edd5e8cfcc675663cdb6e791..ff79a6480c55efbbd0635ed76afdb765bbf39f15 100644
--- a/masters/roboolympics-fun/crawl/meta.json
+++ b/masters/roboolympics-fun/crawl/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-crawl-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-crawl-i00-privileged/embodied-olympics-fun-crawl-i00__YDvTC8q",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,10 +56,16 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   2.141,
-   3.303
+   2.265,
+   3.495
   ],
   "robot_root": "pelvis",
   "travel_m": 11.431,
@@ -76,7 +82,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.81,
+ "seconds_per_frame": 2.13,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -84,5 +90,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:39-0500"
+ "created": "2026-10-25T09:57:14-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/crawl/poster.png b/masters/roboolympics-fun/crawl/poster.png
index 67908f916f541b7b596cffdd108aa5eedbe95dca..6459670155fbbe6a99e532ce9c6b1467921e6488 100644
--- a/masters/roboolympics-fun/crawl/poster.png
+++ b/masters/roboolympics-fun/crawl/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:01ce4dc4205fd6d5464e2b72b0204baf7255b1f77a7377bdc823973194ad4faa
-size 3311344
+oid sha256:467a37afbdcd3e9f12dfa025e274b8567b8b3386ac9011a5e778d73a07d1ee11
+size 3303767
diff --git a/masters/roboolympics-fun/crawl/video.mp4 b/masters/roboolympics-fun/crawl/video.mp4
index f8b16300226f638567ffa1f2fbd1613434063733..0bc5b2bea2ac7af644833a35e2b4ee24698bf0a4 100644
--- a/masters/roboolympics-fun/crawl/video.mp4
+++ b/masters/roboolympics-fun/crawl/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb0a674a9f857c644ae9c9df03b7d39e26d6e2baa09326728770e56e1417aa14
-size 28413111
+oid sha256:a0e4930961a37d9876b174750dd6c8513b592864a90b4488e875ef3225c0d63b
+size 27062366
diff --git a/masters/roboolympics-fun/discus/meta.json b/masters/roboolympics-fun/discus/meta.json
index 66fabd07c3adc697bd44976a58870789521089f5..bfffd8552760322c27c019ab571a1a9639fbe5ad 100644
--- a/masters/roboolympics-fun/discus/meta.json
+++ b/masters/roboolympics-fun/discus/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-discus-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-discus-i00-privileged/embodied-olympics-fun-discus-i00__Uhxr9X9",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule object-travel-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -60,17 +60,23 @@
   "az_rule": "object-travel-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   5.578,
-   14.867
+   5.889,
+   14.897
   ],
   "robot_root": "pelvis",
   "travel_m": 1.332,
   "focus": {
-   "robot": 7,
+   "robot+object": 51,
    "object": 171,
-   "robot+object": 51
+   "robot": 7
   },
   "objects": {
    "discus": 25.02
@@ -84,13 +90,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.7,
+ "seconds_per_frame": 2.02,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
  "mode": "privileged",
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
- "poster": "frame 41 (same scene, camera and settings as the video)",
+ "poster": "frame 41 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:40-0500"
+ "created": "2026-10-25T09:57:14-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/discus/poster.png b/masters/roboolympics-fun/discus/poster.png
index 1e5d40946130866fb63a00cc8a03abefce73ad85..be1826f0ea888d5971613982b9e4d39a75b0e452 100644
--- a/masters/roboolympics-fun/discus/poster.png
+++ b/masters/roboolympics-fun/discus/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a2e9bbed96c92f81f5713939192d201936a0d65817d9fd6119d80f29401edcbd
-size 3027679
+oid sha256:943c06383814f409b0c8b99f29e00df9b0fcf19eeca760698a2b3247e2a65398
+size 3028481
diff --git a/masters/roboolympics-fun/discus/video.mp4 b/masters/roboolympics-fun/discus/video.mp4
index b53ce90c0866d4442671cd5de4fa22dcd50839ba..34ffbfc5623eefb5fa70a937aadda6f7291c4ad6 100644
--- a/masters/roboolympics-fun/discus/video.mp4
+++ b/masters/roboolympics-fun/discus/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:898a064a6519084d2ce9a32d2cda3f4779613d68916e3ecbd401592011041d09
-size 6559398
+oid sha256:07ef2634550c7ca43b7cc053e867c1b5404e2ec66714cb0a88c6df61cbf72390
+size 5609016
diff --git a/masters/roboolympics-fun/dodgeball/meta.json b/masters/roboolympics-fun/dodgeball/meta.json
index ff2f39281fd57edd2f5d55acf9ff21491d2b3310..760290d74677d66552dad6cd0e2cafa7c1b0b9a5 100644
--- a/masters/roboolympics-fun/dodgeball/meta.json
+++ b/masters/roboolympics-fun/dodgeball/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-dodgeball-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-dodgeball-i00-privileged/embodied-olympics-fun-dodgeball__sZ5uVQa",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (pedestal; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,18 +57,32 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^pedestal\\d*$",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "pedestal1",
+   "pedestal2",
+   "pedestal3",
+   "pedestal4",
+   "pedestal5"
+  ],
   "distance_m": [
-   7.133,
-   19.085
+   9.093,
+   21.821
   ],
   "robot_root": "pelvis",
   "travel_m": 2.383,
   "focus": {
-   "robot+object": 277,
    "robot": 1,
-   "object": 106,
-   "robot(incoming)": 53
+   "object+equip": 34,
+   "robot+object": 199,
+   "object": 72,
+   "robot(incoming)+equip": 53,
+   "robot+object+equip": 78
   },
   "objects": {
    "ball1": 20.24,
@@ -91,13 +105,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.64,
+ "seconds_per_frame": 1.99,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
  "mode": "privileged",
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
- "poster": "frame 84 (same scene, camera and settings as the video)",
+ "poster": "frame 84 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:40-0500"
+ "created": "2026-10-25T09:57:14-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/dodgeball/poster.png b/masters/roboolympics-fun/dodgeball/poster.png
index f61fc7ebf9f6e7200409adef8141d02b47474c7b..c4cf4db5f634c1a057fb2085cd9eefeaefcf9945 100644
--- a/masters/roboolympics-fun/dodgeball/poster.png
+++ b/masters/roboolympics-fun/dodgeball/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4030ec93781a3e111ec8923c758390c44d9b284f21378b5b410180d5196c2cd5
-size 3081999
+oid sha256:199bed299771b8385a1594b8969679bffe3500eb51052d461e527a2c9ad71be8
+size 3321424
diff --git a/masters/roboolympics-fun/dodgeball/video.mp4 b/masters/roboolympics-fun/dodgeball/video.mp4
index dab9cfa7143b2d11aa286c87df01ff042505628f..2060e63916cca3411af51efbdb0e9d6afb6744ea 100644
--- a/masters/roboolympics-fun/dodgeball/video.mp4
+++ b/masters/roboolympics-fun/dodgeball/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9bb5f06833231633b5804c87b60983589cb09ea9e4f35e4d92772aff601f8639
-size 9637449
+oid sha256:2f4e392cfb0a19f26858f7c819fe154670ca81a626535f6125650d6fcf5f1217
+size 10592251
diff --git a/masters/roboolympics-fun/double_dutch/meta.json b/masters/roboolympics-fun/double_dutch/meta.json
index 9f14f8294957bccbdb021b40bfcb7d37540e4bb6..b8afa499dcb12ccfeb21bc777818f71e87d5433c 100644
--- a/masters/roboolympics-fun/double_dutch/meta.json
+++ b/masters/roboolympics-fun/double_dutch/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-double-dutch-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-double-dutch-i00-privileged/embodied-olympics-fun-double-dut__uSMXSEu",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -64,10 +64,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   5.949,
-   6.393
+   6.276,
+   6.717
   ],
   "robot_root": "pelvis",
   "travel_m": 1.448,
@@ -90,7 +96,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.7,
+ "seconds_per_frame": 2.02,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -98,5 +104,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:40-0500"
+ "created": "2026-10-25T09:57:14-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/double_dutch/poster.png b/masters/roboolympics-fun/double_dutch/poster.png
index 2c08cac88c2025dcd413c80df31c9b33821c37ca..52d41e3518304a8e3f1d4fdc3ce206d2bb1abd9b 100644
--- a/masters/roboolympics-fun/double_dutch/poster.png
+++ b/masters/roboolympics-fun/double_dutch/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:912a9bf49848149d00bd9c9446f3cc386b3b47902e7ff21a184336df6abb3689
-size 3121155
+oid sha256:6b6528c016c2b83103ec2a2588508d2465b8e2189029a1c7d6f13effe44bb955
+size 3112120
diff --git a/masters/roboolympics-fun/double_dutch/video.mp4 b/masters/roboolympics-fun/double_dutch/video.mp4
index 6490d7d36c8374ba0422de64d3074b2303c7cecb..6a9b093654269a004b70efdb5c9a7c8bcd7fe15f 100644
--- a/masters/roboolympics-fun/double_dutch/video.mp4
+++ b/masters/roboolympics-fun/double_dutch/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f168ac797460bce38ec7b1750f4ae53dd92e880d44a66db065998c525c1857f4
-size 6032125
+oid sha256:ee58caaa69d0ab9fb3981d4f43c2842d39f881cebfc4bd8a19b19f92e53b8442
+size 5772955
diff --git a/masters/roboolympics-fun/dribble_slalom/meta.json b/masters/roboolympics-fun/dribble_slalom/meta.json
index fe218af9586c87b43c546221c2f53391baf721c2..4ebd095f99a1619af8fb4532e7be872a403df8b7 100644
--- a/masters/roboolympics-fun/dribble_slalom/meta.json
+++ b/masters/roboolympics-fun/dribble_slalom/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-dribble-slalom-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-dribble-slalom-i00-privileged/embodied-olympics-fun-dribble-sl__rfFDxRp",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (cone; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -64,20 +64,33 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -50.0,
-  "az_rule": "facing-50",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^cone",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "cone1",
+   "cone2",
+   "cone3",
+   "cone4",
+   "cone5",
+   "cone6"
+  ],
   "distance_m": [
-   3.24,
-   4.325
+   6.598,
+   17.355
   ],
   "robot_root": "pelvis",
   "travel_m": 1.194,
   "focus": {
-   "robot+object": 154,
-   "robot": 577
+   "robot+equip": 577,
+   "robot+object+equip": 154
   },
   "objects": {
    "ball": 0.98
@@ -91,7 +104,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.61,
+ "seconds_per_frame": 2.0,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -99,5 +112,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:40-0500"
+ "created": "2026-10-25T09:57:14-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/dribble_slalom/poster.png b/masters/roboolympics-fun/dribble_slalom/poster.png
index 623ff806941670c3551fb97b28562136b39ed9e8..3cacc49498e958d8148910a467845d33a2ae1fd6 100644
--- a/masters/roboolympics-fun/dribble_slalom/poster.png
+++ b/masters/roboolympics-fun/dribble_slalom/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9040880a23184153d404d9cbd38158419b1cb06fe6c08a1a6411aab41a715f88
-size 3125980
+oid sha256:41a926e812abbc57d0c75cf40ed14b4a80b6cad48af935f9077d3c5523afb3f8
+size 3019152
diff --git a/masters/roboolympics-fun/dribble_slalom/video.mp4 b/masters/roboolympics-fun/dribble_slalom/video.mp4
index 9768f12392ad73124443b19528780bc6885ad1c1..c8277aeda2154ac061b323fac4a12e864ebb919a 100644
--- a/masters/roboolympics-fun/dribble_slalom/video.mp4
+++ b/masters/roboolympics-fun/dribble_slalom/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fe10e3fd016685d7f90b6689e72358130479ca5ea344bf2b708a810576e7e30e
-size 19081995
+oid sha256:b056c960a33be93a02251b5079681def0b9fb20d0c335a99039d961fc49e4471
+size 10555296
diff --git a/masters/roboolympics-fun/dunk/meta.json b/masters/roboolympics-fun/dunk/meta.json
index ba7631aea14ee94f4bdb6fb177a4cc77da193a41..a5d94f3885e1e5ee7e32d5936e84ea3e2136d3a7 100644
--- a/masters/roboolympics-fun/dunk/meta.json
+++ b/masters/roboolympics-fun/dunk/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-dunk-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@aef3b33b59",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-dunk-i00-privileged/embodied-olympics-fun-dunk-i00-p__eruDdGs",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (hoop; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -52,20 +52,28 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -50.0,
-  "az_rule": "facing-50",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^hoop_(rim|board|plate|net)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "hoop"
+  ],
   "distance_m": [
-   4.352,
-   5.863
+   8.636,
+   8.807
   ],
   "robot_root": "pelvis",
   "travel_m": 0.213,
   "focus": {
-   "robot+object": 212,
-   "robot": 80
+   "robot+equip": 80,
+   "robot+object+equip": 212
   },
   "objects": {
    "ball": 1.91
@@ -79,13 +87,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.97,
+ "seconds_per_frame": 2.13,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success: its formal privileged result on this task version (the quick check of aef3b33b59, reviewed and promoted)",
- "poster": "frame 140 (same scene, camera and settings as the video)",
+ "poster": "frame 140 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T21:09:41-0500"
+ "created": "2026-10-25T09:57:14-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/dunk/poster.png b/masters/roboolympics-fun/dunk/poster.png
index 3b18d40764d6f595105a33683b8151a5f17273ab..5a217b6817f5dc5c164a839f39ae1e3ff216d35b 100644
--- a/masters/roboolympics-fun/dunk/poster.png
+++ b/masters/roboolympics-fun/dunk/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ea943b0c12b40b1a0d3b5fb414100439e526992c276565c72bf8f1516404304e
-size 3238032
+oid sha256:4aa08aa527e2a6c33cce9e7e29a4794452eaf8b7c407d8b7df7b69f12bf393be
+size 3215839
diff --git a/masters/roboolympics-fun/dunk/video.mp4 b/masters/roboolympics-fun/dunk/video.mp4
index 96502a579d845a4a3f6a03e899edd38711641b84..c6d5e7e12814b4fd7059c2615edbbc6bc104957d 100644
--- a/masters/roboolympics-fun/dunk/video.mp4
+++ b/masters/roboolympics-fun/dunk/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:734a7b104ac60fb16d6ec4f5109a39f409b96c2a5159811624aae6cbb4311e5f
-size 6716384
+oid sha256:c48a72c4cc636bbcb0e4f7a35123d277663de715057cceef837b3d41ed510aa6
+size 5216438
diff --git a/masters/roboolympics-fun/dunk_between_legs/meta.json b/masters/roboolympics-fun/dunk_between_legs/meta.json
index e8a3e64d7da96017c3ab9f051dd9f4728e0b3190..45a0cf32c80671e6afdb29a43d76c6199a5d19f9 100644
--- a/masters/roboolympics-fun/dunk_between_legs/meta.json
+++ b/masters/roboolympics-fun/dunk_between_legs/meta.json
@@ -7,7 +7,7 @@
  "source": "scene-orbit",
  "why_no_video": "No evaluated agent has a success on this task version yet. Site rule (boards 0161 / 0252): show GPT-6 Astra's failed attempt with the most progress instead. This video is that attempt replayed exactly (not a solved run): the task's own referee at the end says it failed, as the verifier did.",
  "shows": "agent-attempt (failed):codex-gpt6_astra-high-embodied-olympics-fun-dunk-between-legs-i00-privileged/embodied-olympics-fun-dunk-betwe__xbhRyY5",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (hoop; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -61,16 +61,24 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^hoop_(rim|board|plate|net)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "hoop"
+  ],
   "distance_m": [
-   4.498,
-   6.681
+   6.525,
+   8.124
   ],
   "robot_root": "pelvis",
   "travel_m": 3.589,
   "focus": {
-   "robot+object": 320,
-   "robot": 170
+   "robot+object+equip": 320,
+   "robot+equip": 170
   },
   "objects": {
    "ball": 3.64
@@ -84,7 +92,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.91,
+ "seconds_per_frame": 2.11,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -92,5 +100,5 @@
  "trial_choice": "no success on this task version; GPT-6 Astra's failed attempt (most progress)",
  "poster": "last frame (end of the failed attempt)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T21:09:41-0500"
+ "created": "2026-10-25T09:57:14-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/dunk_between_legs/poster.png b/masters/roboolympics-fun/dunk_between_legs/poster.png
index 3717850fadd3b004491d15485ef5c71231e0d9ad..4e9b8bcebf87f8c12f36d9b782f328fae230000d 100644
--- a/masters/roboolympics-fun/dunk_between_legs/poster.png
+++ b/masters/roboolympics-fun/dunk_between_legs/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:902fef322d56f111582cb0a2221db8fa3cdb6ecf9327a8f482bb4739a8bf0c2e
-size 3302388
+oid sha256:0e24477af0d70157d889b5ace4fc26f1c1a740b809943ae6592d5e144a789faa
+size 3262348
diff --git a/masters/roboolympics-fun/dunk_between_legs/video.mp4 b/masters/roboolympics-fun/dunk_between_legs/video.mp4
index 4eac7d8f06555b3ff3aceb1e92220ac4bbb1f37f..00605989a9de663c240b2eb94cbe1259646a63f7 100644
--- a/masters/roboolympics-fun/dunk_between_legs/video.mp4
+++ b/masters/roboolympics-fun/dunk_between_legs/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:219ecca0e641d22d4f989d8927ff82b76831fc337dd5bda1fcab92554672b306
-size 10614174
+oid sha256:9c10ed9561c0f4cff96fcc337c0adc41523d901418cc9b91dce249b955f393fc
+size 8785435
diff --git a/masters/roboolympics-fun/first_time_shot/meta.json b/masters/roboolympics-fun/first_time_shot/meta.json
index 92c85f594c980554a5341f7f0fb0c83a7a8c39ef..d087ed5ad016ddea16fa9ef73e7716e61074c345 100644
--- a/masters/roboolympics-fun/first_time_shot/meta.json
+++ b/masters/roboolympics-fun/first_time_shot/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-first-time-shot-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-first-time-shot-i00-privileged/embodied-olympics-fun-first-time__PQr3kXP",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (goal; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -54,22 +54,31 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -149.16,
-  "az_rule": "object-travel-115",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^goal$",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "goal"
+  ],
   "distance_m": [
-   3.302,
-   11.686
+   5.544,
+   18.349
   ],
   "robot_root": "pelvis",
   "travel_m": 0.592,
   "focus": {
-   "robot": 83,
-   "robot+object": 94,
+   "object+equip": 14,
+   "robot+object+equip": 5,
+   "robot+object": 89,
    "robot(incoming)": 11,
-   "object": 14
+   "robot": 83
   },
   "objects": {
    "gate": 0.08,
@@ -84,13 +93,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 3.14,
+ "seconds_per_frame": 2.24,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 171 (same scene, camera and settings as the video)",
+ "poster": "frame 171 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:40-0500"
+ "created": "2026-10-25T09:57:15-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/first_time_shot/poster.png b/masters/roboolympics-fun/first_time_shot/poster.png
index 2e2d64481a88af7b16abaae42f220cd4a7caf039..ec858a41aead8ca487ebe6bc8c55fe4f3759ea40 100644
--- a/masters/roboolympics-fun/first_time_shot/poster.png
+++ b/masters/roboolympics-fun/first_time_shot/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ccb68fd9b4aa49e1d2b133b416b487a8233555cbb4ea0dd3f44f4157777eadd3
-size 3157732
+oid sha256:cbe985dfbfd43d3e0a63f20bc6be4925e19d80fc590f6e5b56e2ae096f957425
+size 3053836
diff --git a/masters/roboolympics-fun/first_time_shot/video.mp4 b/masters/roboolympics-fun/first_time_shot/video.mp4
index 22dcda68ae6b968d23e91f9fedd1e14022ac2649..d43fbc68a4aef99486ee386eb6180eb44a29f0d4 100644
--- a/masters/roboolympics-fun/first_time_shot/video.mp4
+++ b/masters/roboolympics-fun/first_time_shot/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f6d33ba8a2fe0f2633bcb42eb2021822f27ef130ba0f10c9b218d761f1f89b81
-size 5047706
+oid sha256:406dc0ab026e1184a4d0d27d79a8d7146eff6fa53dca786d4e72e5dcbce5bddc
+size 4572513
diff --git a/masters/roboolympics-fun/free_kick_wall/meta.json b/masters/roboolympics-fun/free_kick_wall/meta.json
index c532c8794f2dab93aeebd0cbf44b08a224cf1f4d..d53d39ea625573a85ec69bba58db0dd33447fbe3 100644
--- a/masters/roboolympics-fun/free_kick_wall/meta.json
+++ b/masters/roboolympics-fun/free_kick_wall/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-free-kick-wall-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-free-kick-wall-i00-privileged/embodied-olympics-fun-free-kick__rReLRzw",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (goal, mannequin; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -53,21 +53,35 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -100.18,
-  "az_rule": "object-travel-115",
+  "az_deg": -105.74,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(goal$|mannequin)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "goal",
+   "mannequin0",
+   "mannequin1",
+   "mannequin2",
+   "mannequin3"
+  ],
   "distance_m": [
-   3.979,
-   14.151
+   14.126,
+   24.077
   ],
   "robot_root": "pelvis",
   "travel_m": 0.224,
   "focus": {
-   "robot+object": 36,
    "robot": 45,
-   "object": 100
+   "object": 1,
+   "robot+object+equip": 19,
+   "object+equip": 99,
+   "robot+object": 17
   },
   "objects": {
    "ball": 20.88
@@ -81,13 +95,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.75,
+ "seconds_per_frame": 2.36,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 51 (same scene, camera and settings as the video)",
+ "poster": "frame 51 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:41-0500"
+ "created": "2026-10-25T09:57:15-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/free_kick_wall/poster.png b/masters/roboolympics-fun/free_kick_wall/poster.png
index 4226dc7126fc811d0a9e5b1e5d555765e3254550..828ca497b38bf69d7b17e9fa33ffd973d16a11fd 100644
--- a/masters/roboolympics-fun/free_kick_wall/poster.png
+++ b/masters/roboolympics-fun/free_kick_wall/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dd28a6a1872c243a865d6d68e80c4dfcf7f8a3c22f95bff4e35d4f8680377e1e
-size 3050730
+oid sha256:5350c7318db5599fd27651e8d13f35b4a30b5d97e6a8f4acbd13c1c954827327
+size 3068727
diff --git a/masters/roboolympics-fun/free_kick_wall/video.mp4 b/masters/roboolympics-fun/free_kick_wall/video.mp4
index 32709677ad9c016ad7f406500ee542109d21c008..49f953258cb6542b3d7d2f64cb91f185d32286f9 100644
--- a/masters/roboolympics-fun/free_kick_wall/video.mp4
+++ b/masters/roboolympics-fun/free_kick_wall/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4fbd5bc50fd2e104f117d944d01096d847da945693026fd5b75bcff2ee84f430
-size 4084171
+oid sha256:3b5cdfe0d977f267154dc800099298af237a24b715ea296ad8d662c686b4bb5c
+size 3957548
diff --git a/masters/roboolympics-fun/free_throw/meta.json b/masters/roboolympics-fun/free_throw/meta.json
index 168792b5b5148d16adfa70d4e1004708fa5eac31..f98011ce5b503741e76807ab9fac1e92ba208a61 100644
--- a/masters/roboolympics-fun/free_throw/meta.json
+++ b/masters/roboolympics-fun/free_throw/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-free-throw-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-free-throw-i00-privileged/embodied-olympics-fun-free-throw__6uq7st2",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (hoop; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -51,20 +51,28 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -114.98,
-  "az_rule": "object-travel-115",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^hoop_(rim|board|plate|net)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "hoop"
+  ],
   "distance_m": [
-   13.177,
-   14.35
+   14.974,
+   16.052
   ],
   "robot_root": "pelvis",
   "travel_m": 0.311,
   "focus": {
-   "robot+object": 178,
-   "robot": 3
+   "robot+equip": 3,
+   "robot+object+equip": 178
   },
   "objects": {
    "ball": 4.28
@@ -78,13 +86,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.84,
+ "seconds_per_frame": 2.04,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
  "mode": "privileged",
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
- "poster": "frame 23 (same scene, camera and settings as the video)",
+ "poster": "frame 23 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:41-0500"
+ "created": "2026-10-25T09:57:15-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/free_throw/poster.png b/masters/roboolympics-fun/free_throw/poster.png
index 5d552b135f863ff31d77232921555a3acf69b444..a92247399a2f10c212b166131709e4c678586de8 100644
--- a/masters/roboolympics-fun/free_throw/poster.png
+++ b/masters/roboolympics-fun/free_throw/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5ab28719dfacc14410c0e14c71517fbdf1fb4682cc605cdc8b4f219716ffe5ce
-size 3106870
+oid sha256:49d7859e95a794810dcd890f03da226a7c27f08b16fc51dd3e41a6f9b6037aef
+size 3113502
diff --git a/masters/roboolympics-fun/free_throw/video.mp4 b/masters/roboolympics-fun/free_throw/video.mp4
index d6e51ba9a0e90e7fe08e56766d5fb2c0c27c26f8..6608a9a8ba0c1cd5f31832a7b2ee858eddafb3bb 100644
--- a/masters/roboolympics-fun/free_throw/video.mp4
+++ b/masters/roboolympics-fun/free_throw/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:acb4cb377e1d02a6704334b35deeb254e3709e41220dba1c8d8639cb209b1919
-size 3735262
+oid sha256:70a66be627945e0e61d9b1ea5f743ab1218d4dd492f07d7361e858650327efd2
+size 3387922
diff --git a/masters/roboolympics-fun/freeski_360/meta.json b/masters/roboolympics-fun/freeski_360/meta.json
index 393c88298e0ffa477913bbb811f9770e77bc78ad..c835555bce4268b7926e2ad4aa44b5ae179df498 100644
--- a/masters/roboolympics-fun/freeski_360/meta.json
+++ b/masters/roboolympics-fun/freeski_360/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-freeski-360-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@aef3b33b59",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-freeski-360-i00-privileged/embodied-olympics-fun-freeski-36__pA2yUNn",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; snow venue: a snow shader (bright blue-white, rough, slight subsurface, fine bump) on the course's snow geoms, snow-white valley floor; a chase camera: per-frame fit of the robot (height included, so jumps and spins stay framed), at least 6 m away, 24 degree elevation, smoothed in time",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; snow venue: a snow shader (bright blue-white, rough, slight subsurface, fine bump) on the course's snow geoms, snow-white valley floor; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 24.0-degree elevation, a slow eased 0.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -72,10 +72,16 @@
   "az_rule": "robot-travel-70",
   "el_deg": 24.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(kicker|table|landing)$",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
    6.24,
-   14.608
+   14.61
   ],
   "robot_root": "pelvis",
   "travel_m": 143.407,
@@ -96,7 +102,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 4.88,
+ "seconds_per_frame": 3.88,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -104,5 +110,5 @@
  "trial_choice": "a counted GPT-6 Astra success: its formal privileged result on this task version (the quick check of aef3b33b59, reviewed and promoted)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T21:09:41-0500"
+ "created": "2026-10-25T09:57:15-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/freeski_360/poster.png b/masters/roboolympics-fun/freeski_360/poster.png
index 77f51828b7f0d5c4265cedcdfa301052c4ce46b7..ed68b82c67f9631e5807d07031e231745ebcd515 100644
--- a/masters/roboolympics-fun/freeski_360/poster.png
+++ b/masters/roboolympics-fun/freeski_360/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6f2af7b16c170f67907e404ed935b1f3ebb14c31c97283abdb13898b9e421fca
-size 3093822
+oid sha256:1b9e37891466b6ca2b281d26a0ca1595882a57348a4a5f08bb7c898a0a76b646
+size 3093682
diff --git a/masters/roboolympics-fun/freeski_360/video.mp4 b/masters/roboolympics-fun/freeski_360/video.mp4
index faa4566f4467756618d1097180383088f3a18c6a..28af366b67a280e7e922e2c33981828f6defdca8 100644
--- a/masters/roboolympics-fun/freeski_360/video.mp4
+++ b/masters/roboolympics-fun/freeski_360/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a33bccfef63001ce1cecbe29fe0d1e2f403b5e062164d0c2150e01d39025df16
-size 13328293
+oid sha256:11d08bcb3eb5fe20421421f82315d43a783a941fd8e49f83e2f4850b1e431705
+size 12551422
diff --git a/masters/roboolympics-fun/freeski_straight_air/meta.json b/masters/roboolympics-fun/freeski_straight_air/meta.json
index bc3be43926f5f130ff80b56dd9fb7a55f865e072..4053e30ec87c414f87fa1180a471a80d0b568a94 100644
--- a/masters/roboolympics-fun/freeski_straight_air/meta.json
+++ b/masters/roboolympics-fun/freeski_straight_air/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-freeski-straight-air-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@aef3b33b59",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-freeski-straight-air-i00-privileged/embodied-olympics-fun-freeski-st__uJxLXfG",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; snow venue: a snow shader (bright blue-white, rough, slight subsurface, fine bump) on the course's snow geoms, snow-white valley floor; a chase camera: per-frame fit of the robot (height included, so jumps and spins stay framed), at least 6 m away, 24 degree elevation, smoothed in time",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; snow venue: a snow shader (bright blue-white, rough, slight subsurface, fine bump) on the course's snow geoms, snow-white valley floor; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 24.0-degree elevation, a slow eased 0.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -72,10 +72,16 @@
   "az_rule": "robot-travel-70",
   "el_deg": 24.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(kicker|table|landing)$",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
    6.24,
-   14.96
+   14.962
   ],
   "robot_root": "pelvis",
   "travel_m": 143.424,
@@ -96,7 +102,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 4.94,
+ "seconds_per_frame": 3.89,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -104,5 +110,5 @@
  "trial_choice": "a counted GPT-6 Astra success: its formal privileged result on this task version (the quick check of aef3b33b59, reviewed and promoted)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T21:09:41-0500"
+ "created": "2026-10-25T09:57:15-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/freeski_straight_air/poster.png b/masters/roboolympics-fun/freeski_straight_air/poster.png
index 747bb818714b4477ba926e83f0e9b22957e236f1..6b9d3f312fceecbf4233664a2ce504b71b15ebd9 100644
--- a/masters/roboolympics-fun/freeski_straight_air/poster.png
+++ b/masters/roboolympics-fun/freeski_straight_air/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1fbd710e5d1a2f59ccf42f20fa273a1373acde2a66084ffb01c62d1f999a525b
-size 3103881
+oid sha256:584200b2b83c34a0086ce01b7aec17796575b850aa8c9cc6716d0cf641b1edf0
+size 3105711
diff --git a/masters/roboolympics-fun/freeski_straight_air/video.mp4 b/masters/roboolympics-fun/freeski_straight_air/video.mp4
index 6ec55e924580e0169116a5b058a8fec3aef43c35..dbdeb55ac085f9aea4197f98d62f2827ae2259a6 100644
--- a/masters/roboolympics-fun/freeski_straight_air/video.mp4
+++ b/masters/roboolympics-fun/freeski_straight_air/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eee12262695d1c0d7a9c5479f0cfcfe61bef95a64c7d8c66abcbe7cc76e7b41f
-size 12635346
+oid sha256:32da7799902648d1e5c1aa891063d93a88da17f28beae799616ed7351442d016
+size 11988843
diff --git a/masters/roboolympics-fun/giant_slalom/meta.json b/masters/roboolympics-fun/giant_slalom/meta.json
index 950549fc69aadf1b5f683a1efa49ea18b4aa91f1..d587b52eeed6ea980266cba5206c2d2669085a81 100644
--- a/masters/roboolympics-fun/giant_slalom/meta.json
+++ b/masters/roboolympics-fun/giant_slalom/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-giant-slalom-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-giant-slalom-i00-privileged/embodied-olympics-fun-giant-slal__rNyiV8M",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; snow venue: a snow shader (bright blue-white, rough, slight subsurface, fine bump) on the course's snow geoms, snow-white valley floor; a chase camera: per-frame fit of the robot (height included, so jumps and spins stay framed), at least 6 m away, 24 degree elevation, smoothed in time",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; snow venue: a snow shader (bright blue-white, rough, slight subsurface, fine bump) on the course's snow geoms, snow-white valley floor; a director camera: a 0.0 s establishing shot framing the robot and the task's key equipment (gate0_turn, gate10_turn, gate11_turn, gate12_turn, gate13_turn, gate14_turn, gate15_turn, gate16_turn, gate17_turn, gate18_turn, gate19_turn, gate1_turn, gate20_turn, gate21_turn, gate22_turn, gate23_turn, gate24_turn, gate25_turn, gate2_turn, gate3_turn, gate4_turn, gate5_turn, gate6_turn, gate7_turn, gate8_turn, gate9_turn; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 24.0-degree elevation, a slow eased 0.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -84,16 +84,50 @@
   "az_rule": "robot-travel-70",
   "el_deg": 24.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^gate\\d+_turn",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 0.0,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "gate0_turn",
+   "gate10_turn",
+   "gate11_turn",
+   "gate12_turn",
+   "gate13_turn",
+   "gate14_turn",
+   "gate15_turn",
+   "gate16_turn",
+   "gate17_turn",
+   "gate18_turn",
+   "gate19_turn",
+   "gate1_turn",
+   "gate20_turn",
+   "gate21_turn",
+   "gate22_turn",
+   "gate23_turn",
+   "gate24_turn",
+   "gate25_turn",
+   "gate2_turn",
+   "gate3_turn",
+   "gate4_turn",
+   "gate5_turn",
+   "gate6_turn",
+   "gate7_turn",
+   "gate8_turn",
+   "gate9_turn"
+  ],
   "distance_m": [
    6.24,
-   8.667
+   12.678
   ],
   "robot_root": "pelvis",
   "travel_m": 611.005,
   "focus": {
    "robot": 35,
-   "robot+object": 746
+   "robot+object+equip": 359,
+   "robot+object": 387
   },
   "objects": {
    "left_ski": 611.14,
@@ -108,7 +142,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 5.06,
+ "seconds_per_frame": 3.78,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
@@ -116,5 +150,5 @@
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:41-0500"
+ "created": "2026-10-25T09:57:15-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/giant_slalom/poster.png b/masters/roboolympics-fun/giant_slalom/poster.png
index 2604b7afea64c2d0920009a090ecb6f020c1ab5a..4919a1c174701f2a7a22dbb2bfa61b0f884c43b3 100644
--- a/masters/roboolympics-fun/giant_slalom/poster.png
+++ b/masters/roboolympics-fun/giant_slalom/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c25e4ef2e97c01002caa58cad3371373ebe5fdb0da99cd2c4757b2169c0f5938
-size 2979842
+oid sha256:f93dd3f4c17113fc8441c6af0149f1b25f7b46b8ef0de2d35ce411afc5e0b78f
+size 2979347
diff --git a/masters/roboolympics-fun/giant_slalom/video.mp4 b/masters/roboolympics-fun/giant_slalom/video.mp4
index acecd18cfdcf074f1597718c1d8318b3bff2dca3..0a23085a97511d9439fc8779ff51f2eee2980dee 100644
--- a/masters/roboolympics-fun/giant_slalom/video.mp4
+++ b/masters/roboolympics-fun/giant_slalom/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0d1336d75ae24867d8020d805a213449d310e274ad1f67eee4a1408d774b1f06
-size 12654338
+oid sha256:e0a25cfdf07b62a3023d28cff97207f6dc50a9290ab64ee8e992694ad01003ea
+size 11535447
diff --git a/masters/roboolympics-fun/goalkeeper_save/meta.json b/masters/roboolympics-fun/goalkeeper_save/meta.json
index c4df413646350f71101ae9f12f34c2d272c1e14e..4219d6e26b292febef731781d0945ecf9cea39cd 100644
--- a/masters/roboolympics-fun/goalkeeper_save/meta.json
+++ b/masters/roboolympics-fun/goalkeeper_save/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-goalkeeper-save-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-goalkeeper-save-i00-privileged/embodied-olympics-fun-goalkeeper__46wnCQn",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (goal; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -51,22 +51,30 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -29.02,
-  "az_rule": "object-travel-115",
+  "az_deg": 65.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^goal$",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "goal"
+  ],
   "distance_m": [
-   4.104,
-   14.23
+   7.282,
+   16.232
   ],
   "robot_root": "pelvis",
   "travel_m": 0.735,
   "focus": {
-   "robot(outgoing)": 404,
-   "robot(incoming)": 22,
-   "robot": 152,
-   "robot+object": 91
+   "robot+equip": 152,
+   "robot(outgoing)+equip": 404,
+   "robot+object+equip": 91,
+   "robot(incoming)+equip": 22
   },
   "objects": {
    "machine1": 0.02,
@@ -85,7 +93,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 5.08,
+ "seconds_per_frame": 3.46,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
@@ -93,5 +101,5 @@
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:42-0500"
+ "created": "2026-10-25T09:57:15-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/goalkeeper_save/poster.png b/masters/roboolympics-fun/goalkeeper_save/poster.png
index e914b15c8c5657d4895418ea23d185cdad2bb77a..31d2b7b03f0cc822656f77ab376343fb05476e2d 100644
--- a/masters/roboolympics-fun/goalkeeper_save/poster.png
+++ b/masters/roboolympics-fun/goalkeeper_save/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9872149837696a459769340efc91b441a200c1f281347d227efe4e7201790617
-size 3208403
+oid sha256:d32882f762c74097c95bc452a5300b36435080b8967840cf64c3189ab179a2c7
+size 3134539
diff --git a/masters/roboolympics-fun/goalkeeper_save/video.mp4 b/masters/roboolympics-fun/goalkeeper_save/video.mp4
index f710bd2bfb32f4c33adc442f012f86dd2790da29..25b9b1ca1e1544393b2930a3d55aa1f6a73f6191 100644
--- a/masters/roboolympics-fun/goalkeeper_save/video.mp4
+++ b/masters/roboolympics-fun/goalkeeper_save/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0c11f9453b3814162213c61a18bd99e60b61349a60522f2baa48fa2df7e7a096
-size 14645562
+oid sha256:46a7cd222fae869c15e60e743aaac9d863a1ccefe95a38f27fd14fbf49c64390
+size 10870099
diff --git a/masters/roboolympics-fun/golf_putt/meta.json b/masters/roboolympics-fun/golf_putt/meta.json
index 4714c722a70697bec3caa27ae2f10f1464407ab1..7a1ea844586fa746bbc91d56c3022d52a4fb5605 100644
--- a/masters/roboolympics-fun/golf_putt/meta.json
+++ b/masters/roboolympics-fun/golf_putt/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-golf-putt-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-golf-putt-i00-privileged/embodied-olympics-fun-golf-putt__eCpKzxA",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (green_hole, hole_bottom, hole_liner; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -54,21 +54,77 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -25.12,
-  "az_rule": "object-travel-115",
+  "az_deg": -33.74,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(green_hole|hole)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "green_hole0",
+   "green_hole1",
+   "green_hole10",
+   "green_hole11",
+   "green_hole12",
+   "green_hole13",
+   "green_hole14",
+   "green_hole15",
+   "green_hole16",
+   "green_hole17",
+   "green_hole18",
+   "green_hole19",
+   "green_hole2",
+   "green_hole20",
+   "green_hole21",
+   "green_hole22",
+   "green_hole23",
+   "green_hole3",
+   "green_hole4",
+   "green_hole5",
+   "green_hole6",
+   "green_hole7",
+   "green_hole8",
+   "green_hole9",
+   "hole_bottom",
+   "hole_liner0",
+   "hole_liner1",
+   "hole_liner10",
+   "hole_liner11",
+   "hole_liner12",
+   "hole_liner13",
+   "hole_liner14",
+   "hole_liner15",
+   "hole_liner16",
+   "hole_liner17",
+   "hole_liner18",
+   "hole_liner19",
+   "hole_liner2",
+   "hole_liner20",
+   "hole_liner21",
+   "hole_liner22",
+   "hole_liner23",
+   "hole_liner3",
+   "hole_liner4",
+   "hole_liner5",
+   "hole_liner6",
+   "hole_liner7",
+   "hole_liner8",
+   "hole_liner9"
+  ],
   "distance_m": [
-   3.455,
-   3.587
+   5.057,
+   5.459
   ],
   "robot_root": "pelvis",
   "travel_m": 0.003,
   "focus": {
-   "robot+object": 141,
-   "robot": 19,
-   "robot+landed": 19
+   "robot+landed+equip": 19,
+   "robot+object+equip": 141,
+   "robot+equip": 19
   },
   "objects": {
    "putter": 0.4,
@@ -83,13 +139,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.67,
+ "seconds_per_frame": 1.95,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 73 (same scene, camera and settings as the video)",
+ "poster": "frame 73 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:42-0500"
+ "created": "2026-10-25T09:57:15-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/golf_putt/poster.png b/masters/roboolympics-fun/golf_putt/poster.png
index ca719ecfc819f1a2a211a5e257ae7583743f2936..02b4e9fbeebb0bd1aa6dffbc26e518cc6f50d8de 100644
--- a/masters/roboolympics-fun/golf_putt/poster.png
+++ b/masters/roboolympics-fun/golf_putt/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e8c0735ac30d8abe4a55d4189b3728882b49eaa4f68ed7f2c6bf0a74dda380e0
-size 3144933
+oid sha256:f14da57d4ee1205403d3a2d0076428cac4478dc71767c4c8ca83a8e12c9d3c23
+size 3135399
diff --git a/masters/roboolympics-fun/golf_putt/video.mp4 b/masters/roboolympics-fun/golf_putt/video.mp4
index 71449b7ee42817b48182a94d6026a4bc420c0277..f6769030f54130acd5e7eb1de63bbe79dbed8785 100644
--- a/masters/roboolympics-fun/golf_putt/video.mp4
+++ b/masters/roboolympics-fun/golf_putt/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c47f297c2a1372f1ea388e16fbb12bd9ebeb848cbca01a4a676e5e64faecbf76
-size 3984949
+oid sha256:e38ab5aaa33767031f88fc0d3a07539f1e2f991aa1183054082c24976f36eb94
+size 3779890
diff --git a/masters/roboolympics-fun/handstand/meta.json b/masters/roboolympics-fun/handstand/meta.json
index c9e4e4c7adb1a4bd4fd3cd8f87ea0b670fe302a0..cc82ce7454232e56e4314e011feeb1813cba44cd 100644
--- a/masters/roboolympics-fun/handstand/meta.json
+++ b/masters/roboolympics-fun/handstand/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-handstand-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@aef3b33b59",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-handstand-i00-privileged/embodied-olympics-fun-handstand__jNPsVm2",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,10 +56,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   2.132,
-   4.146
+   2.252,
+   4.384
   ],
   "robot_root": "pelvis",
   "travel_m": 0.015,
@@ -76,7 +82,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.62,
+ "seconds_per_frame": 2.01,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -84,5 +90,5 @@
  "trial_choice": "a counted GPT-6 Astra success: its formal privileged result on this task version (the quick check of aef3b33b59, reviewed and promoted)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T21:09:41-0500"
+ "created": "2026-10-25T09:57:15-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/handstand/poster.png b/masters/roboolympics-fun/handstand/poster.png
index e135f7a2283c69559853219ef51e899f2b2e56d8..d27a5fdc53161b1364d9d6390bdd24d631214471 100644
--- a/masters/roboolympics-fun/handstand/poster.png
+++ b/masters/roboolympics-fun/handstand/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:69a4f30263f13f52b8a315e2c35a92c49260c76e968c309257f143aa2af0fd27
-size 3119833
+oid sha256:24f9b2e72993393ceb6022457f90d150c5671b2cefb914d499d49049eb021b93
+size 3107154
diff --git a/masters/roboolympics-fun/handstand/video.mp4 b/masters/roboolympics-fun/handstand/video.mp4
index a36a286143a1fdb7d462475dcfb34c79c928483a..fd75659a9d526e289a3f2589056646b60d84832a 100644
--- a/masters/roboolympics-fun/handstand/video.mp4
+++ b/masters/roboolympics-fun/handstand/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a2252c46f846b36d70b7a7dbcdbedd6741fd3f99efa7bacc1998af7a721c67bf
-size 11327778
+oid sha256:b42a66009281e829c2d9fd26a9d2fffda3362904a917ab09b34aa257ca48a9fc
+size 11039787
diff --git a/masters/roboolympics-fun/header/meta.json b/masters/roboolympics-fun/header/meta.json
index cb4a76e2228d12ab949547dda95c301e71d52b7d..cee87d6397e7dde8e146256ad45e71fce6219a8d 100644
--- a/masters/roboolympics-fun/header/meta.json
+++ b/masters/roboolympics-fun/header/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-header-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-header-i00-privileged/embodied-olympics-fun-header-i00__rStdfSM",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (goal; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -53,22 +53,30 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -195.82,
-  "az_rule": "object-travel-115",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^goal$",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "goal"
+  ],
   "distance_m": [
-   8.312,
-   35.181
+   13.364,
+   20.47
   ],
   "robot_root": "pelvis",
   "travel_m": 0.28,
   "focus": {
-   "robot(incoming)": 83,
-   "object": 51,
-   "robot": 1,
-   "robot+object": 34
+   "robot+object+equip": 34,
+   "robot(incoming)+equip": 83,
+   "object+equip": 51,
+   "robot+equip": 1
   },
   "objects": {
    "machine": 0.04,
@@ -83,7 +91,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.94,
+ "seconds_per_frame": 2.5,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -91,5 +99,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:42-0500"
+ "created": "2026-10-25T09:57:15-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/header/poster.png b/masters/roboolympics-fun/header/poster.png
index 8e7f306459e56f4a33d8db0db1f142fac44388ed..8774ba94939f7dab539b8d3fd3a45bd33d61df24 100644
--- a/masters/roboolympics-fun/header/poster.png
+++ b/masters/roboolympics-fun/header/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e2d3a242ec9aa42ec160ab1b56d401915f540817bcc4e1ef4900cfebce429020
-size 3147948
+oid sha256:ba66648ae3be8e01be22dfbe5dcfdfd1fb164986a4f17d754fadbaf447df10d7
+size 3122672
diff --git a/masters/roboolympics-fun/header/video.mp4 b/masters/roboolympics-fun/header/video.mp4
index cb93fffdaa6db22d323d47237a894484151d1218..bf428c36c9daa54f7150f0e5d8bc53d567e9c129 100644
--- a/masters/roboolympics-fun/header/video.mp4
+++ b/masters/roboolympics-fun/header/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:507e6dd285254f9a579440575b0ee5445d60109d0c56f0b77c6afd4780fa4ad3
-size 4520312
+oid sha256:a09970cf51a3c252653c5590279966f5d24b217d9e566a884e33b0ce53467381
+size 3504529
diff --git a/masters/roboolympics-fun/high_bar_swing/meta.json b/masters/roboolympics-fun/high_bar_swing/meta.json
index 7c8bbb4b340fbb5f628df94772392b97d1a5d8eb..889047ee396e3fa486696d64e09f728988583850 100644
--- a/masters/roboolympics-fun/high_bar_swing/meta.json
+++ b/masters/roboolympics-fun/high_bar_swing/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-high-bar-swing-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-high-bar-swing-i00-privileged/embodied-olympics-fun-high-bar-s__QpTJC3u",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (bar; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -60,15 +60,23 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^bar$",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "bar"
+  ],
   "distance_m": [
-   3.642,
-   6.288
+   4.139,
+   5.94
   ],
   "robot_root": "pelvis",
   "travel_m": 0.435,
   "focus": {
-   "robot": 219
+   "robot+equip": 219
   },
   "objects": {}
  },
@@ -80,7 +88,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.65,
+ "seconds_per_frame": 2.0,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -88,5 +96,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:42-0500"
+ "created": "2026-10-25T09:57:15-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/high_bar_swing/poster.png b/masters/roboolympics-fun/high_bar_swing/poster.png
index 8483c66a2be8e5b85de11731c4e65efaea738c11..b5496980c4225aedaa93d5d73fe34e34f6037676 100644
--- a/masters/roboolympics-fun/high_bar_swing/poster.png
+++ b/masters/roboolympics-fun/high_bar_swing/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a6a9d8b9ea29834194644bf184a7618c1bebce9673fbec51d4361e8d6cdd6e34
-size 3096126
+oid sha256:53bc3330aa0599162fe05a7ea7a9c2752e3c56f8d171d9bdb6ef2470828be8ee
+size 3101788
diff --git a/masters/roboolympics-fun/high_bar_swing/video.mp4 b/masters/roboolympics-fun/high_bar_swing/video.mp4
index f75f2049c2d2f19f7fbd6f7bf3281a9a44cffc8b..ad76816712243924c49bee65fe18f3dc4a55eaf4 100644
--- a/masters/roboolympics-fun/high_bar_swing/video.mp4
+++ b/masters/roboolympics-fun/high_bar_swing/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a2568041bf4bbe5728925a195f2c19a4db83050e58cc60123e1c71daa5ecd9ef
-size 5421875
+oid sha256:c23c8258233715f7fc61c825a3bf38eee1c1de32a9fc547d89e61c577164689f
+size 5022713
diff --git a/masters/roboolympics-fun/high_jump/meta.json b/masters/roboolympics-fun/high_jump/meta.json
index 95b8021ae3c95a77dc45e00b3495cdaaf6597b96..0e698c9c7ff8a464057d42f3a30c429cf8c6a012 100644
--- a/masters/roboolympics-fun/high_jump/meta.json
+++ b/masters/roboolympics-fun/high_jump/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-high-jump-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@1e4fb6bf44",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-high-jump-i00-privileged/embodied-olympics-fun-high-jump__D6JwKzP",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (bar, bar_supportL, bar_supportR, bar_uprightL, bar_uprightR; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,15 +56,28 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(bar|bar_upright|bar_support)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "bar",
+   "bar_supportL",
+   "bar_supportR",
+   "bar_uprightL",
+   "bar_uprightR"
+  ],
   "distance_m": [
-   3.262,
-   4.873
+   5.635,
+   11.115
   ],
   "robot_root": "pelvis",
   "travel_m": 8.931,
   "focus": {
-   "robot": 647
+   "robot+equip": 510,
+   "robot": 137
   },
   "objects": {}
  },
@@ -76,13 +89,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.52,
+ "seconds_per_frame": 1.98,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success: its formal privileged result on this task version (the quick check of 1e4fb6bf44, reviewed and promoted)",
- "poster": "frame 567 (same scene, camera and settings as the video)",
+ "poster": "frame 567 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T23:24:49-0500"
+ "created": "2026-10-25T09:57:16-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/high_jump/poster.png b/masters/roboolympics-fun/high_jump/poster.png
index 7a80371cd1e5de1234e72bea437e9fe1e9529419..ed89367efd4099169a444cea8a13da461e0a1403 100644
--- a/masters/roboolympics-fun/high_jump/poster.png
+++ b/masters/roboolympics-fun/high_jump/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:62068adc72ae65eace81c939e6a99afd0fbb327f04fe60caa3e5a446e10e6ba8
-size 3212387
+oid sha256:cc4e58ac98c838ef46b37d637ad8b4bc2b57382869e49cb11f77e7ed64446f03
+size 3080383
diff --git a/masters/roboolympics-fun/high_jump/video.mp4 b/masters/roboolympics-fun/high_jump/video.mp4
index 167d97806ded5d001a94ea1829322f686b75e65a..b41a0c7f5f57938be17f737a4a7913579f1bc970 100644
--- a/masters/roboolympics-fun/high_jump/video.mp4
+++ b/masters/roboolympics-fun/high_jump/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dae6d59a0aaf8d81a304435bc6e6dadfe16a8d25a3cd15f2286f07532ee2f22e
-size 11043669
+oid sha256:71d0d4a0038b9bf61f0a4f1025a0c52a5515c915af60aa36c21f5462b48bba9e
+size 9836153
diff --git a/masters/roboolympics-fun/hurdles_freestanding/meta.json b/masters/roboolympics-fun/hurdles_freestanding/meta.json
index 03e2732bc12ed5d1ab13f2af533dd0bcb61a57f9..304970602aae3e773f90fb7268a0d1d4df41387a 100644
--- a/masters/roboolympics-fun/hurdles_freestanding/meta.json
+++ b/masters/roboolympics-fun/hurdles_freestanding/meta.json
@@ -7,7 +7,7 @@
  "source": "scene-orbit",
  "why_no_video": "No evaluated agent has a success on this task version yet. Site rule (boards 0161 / 0252): show GPT-6 Astra's failed attempt with the most progress instead. This video is that attempt replayed exactly (not a solved run): the task's own referee at the end says it failed, as the verifier did.",
  "shows": "agent-attempt (failed):codex-gpt6_astra-high-embodied-olympics-fun-hurdles-freestanding-i00-privileged/embodied-olympics-fun-hurdles-fr__wqHuDBn",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -59,17 +59,23 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   4.308,
-   9.634
+   4.565,
+   10.209
   ],
   "robot_root": "pelvis",
   "travel_m": 32.557,
   "focus": {
-   "robot+object": 65,
    "robot+landed": 228,
-   "robot": 448
+   "robot": 448,
+   "robot+object": 65
   },
   "objects": {
    "hurdle1": 2.75,
@@ -85,7 +91,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.65,
+ "seconds_per_frame": 1.98,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -93,5 +99,5 @@
  "trial_choice": "no success on this task version; GPT-6 Astra's failed attempt (most progress)",
  "poster": "last frame (end of the failed attempt)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:43-0500"
+ "created": "2026-10-25T09:57:16-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/hurdles_freestanding/poster.png b/masters/roboolympics-fun/hurdles_freestanding/poster.png
index 31cf417a2913949084fc067c16b06c105ceddfb5..8ea08e6744465b7f567a4d7a74cbedfed546ae86 100644
--- a/masters/roboolympics-fun/hurdles_freestanding/poster.png
+++ b/masters/roboolympics-fun/hurdles_freestanding/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ee56c1442f6010b75c84de54f802f6dfdc9ae81cf1cd9651e3fa6ed9e2cfd3a2
-size 3215729
+oid sha256:238bc7f3aac4ad28f65602ef60068a159771f8245f0f2afa05ad0ac2c3b3014c
+size 3209214
diff --git a/masters/roboolympics-fun/hurdles_freestanding/video.mp4 b/masters/roboolympics-fun/hurdles_freestanding/video.mp4
index b25f34c417f3825171ffe78d3a18f82db984b7c0..a0f14f4017c0430e789efc6cf522419b9af74be6 100644
--- a/masters/roboolympics-fun/hurdles_freestanding/video.mp4
+++ b/masters/roboolympics-fun/hurdles_freestanding/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c0c073fd3502353763eaea10bfc4ff49fca3572f19a54b7384657f6faefc7161
-size 18533123
+oid sha256:33e3e753188f83b403cf63de3bbcb0db7d9cd1a9a64ca5dadf8e9a781e843d4d
+size 17869057
diff --git a/masters/roboolympics-fun/inline_slalom/meta.json b/masters/roboolympics-fun/inline_slalom/meta.json
index 3f34046f16ef862642a2ab7d51c1d4ea783fc37d..879bf63b579452f58d73366e3b48593528e843f8 100644
--- a/masters/roboolympics-fun/inline_slalom/meta.json
+++ b/masters/roboolympics-fun/inline_slalom/meta.json
@@ -7,7 +7,7 @@
  "source": "scene-orbit",
  "why_no_video": "No evaluated agent has a success on this task version yet. Site rule (boards 0161 / 0252): show GPT-6 Astra's failed attempt with the most progress instead. This video is that attempt replayed exactly (not a solved run): the task's own referee at the end says it failed, as the verifier did.",
  "shows": "agent-attempt (failed):codex-gpt6_astra-high-embodied-olympics-fun-inline-slalom-i00-privileged/embodied-olympics-fun-inline-sla__azMgCQJ",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -63,10 +63,16 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.224,
-   3.65
+   3.41,
+   3.853
   ],
   "robot_root": "pelvis",
   "travel_m": 28.001,
@@ -87,7 +93,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.63,
+ "seconds_per_frame": 1.97,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -95,5 +101,5 @@
  "trial_choice": "no success on this task version; GPT-6 Astra's failed attempt (most progress)",
  "poster": "last frame (end of the failed attempt)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:43-0500"
+ "created": "2026-10-25T09:57:16-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/inline_slalom/poster.png b/masters/roboolympics-fun/inline_slalom/poster.png
index 8c321501c33b318b93b30028e6627184b04eaa0e..743d9ce060454236607dc9ef81094199e3a689ed 100644
--- a/masters/roboolympics-fun/inline_slalom/poster.png
+++ b/masters/roboolympics-fun/inline_slalom/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:169d61bcc477ec9533767a04bea54d2ca4ae16176ca2c1da2335277a7ec7c72e
-size 3123107
+oid sha256:f7f172015d0688d4434618513effb0573f388c55a70d5f3f4f65f184306cc589
+size 3109398
diff --git a/masters/roboolympics-fun/inline_slalom/video.mp4 b/masters/roboolympics-fun/inline_slalom/video.mp4
index f1535b6ae4d5a231f95878f566619c5cd9959a6e..431771261f23660ca882c5a71d93f5fd37857ade 100644
--- a/masters/roboolympics-fun/inline_slalom/video.mp4
+++ b/masters/roboolympics-fun/inline_slalom/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:713be793bc222e9ead1ea7d43784117078756e41311e11c5e1ea2729aafbf3a3
-size 10629146
+oid sha256:8d6b58d5a5200d43dea684224006c8ebf63a38ff6f2241b087a5fafbf23ffc19
+size 10444345
diff --git a/masters/roboolympics-fun/javelin/meta.json b/masters/roboolympics-fun/javelin/meta.json
index 1e22795bf793b72afbc5a4c6e1ceb017a4741cc5..260da02615e473308540bc4c8cd28db70db061b9 100644
--- a/masters/roboolympics-fun/javelin/meta.json
+++ b/masters/roboolympics-fun/javelin/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-javelin-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-javelin-i00-privileged/embodied-olympics-fun-javelin-i0__cCBqY3s",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -60,19 +60,25 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   4.484,
-   34.973
+   4.741,
+   34.997
   ],
   "robot_root": "pelvis",
   "travel_m": 2.686,
   "focus": {
+   "robot+object": 83,
    "robot": 188,
-   "object": 88,
    "robot+landed": 214,
-   "robot(incoming)": 5,
-   "robot+object": 83
+   "object": 88,
+   "robot(incoming)": 5
   },
   "objects": {
    "javelin": 21.81
@@ -86,13 +92,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.63,
+ "seconds_per_frame": 1.95,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
  "mode": "privileged",
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
- "poster": "frame 257 (same scene, camera and settings as the video)",
+ "poster": "frame 257 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:43-0500"
+ "created": "2026-10-25T09:57:16-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/javelin/poster.png b/masters/roboolympics-fun/javelin/poster.png
index 5448397b80e7bb28da2f172919c52a35ebf7639c..a3b6e4a354bdc839fa0095dfbf6bfc7d89a06e5e 100644
--- a/masters/roboolympics-fun/javelin/poster.png
+++ b/masters/roboolympics-fun/javelin/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:216d8e24d991d0e437318ffba019b65f88cfdfe1778dcad5eaf11d02f8365251
-size 3020923
+oid sha256:4b873328b8200e3739b362a806f01568b24dd3d33c37f06f97baf1b012bac301
+size 3022502
diff --git a/masters/roboolympics-fun/javelin/video.mp4 b/masters/roboolympics-fun/javelin/video.mp4
index c054709f3d3feac445d10011b0fe4a189f663043..0e0a8c284a4edc5bbd9c38c0d943c639103638ad 100644
--- a/masters/roboolympics-fun/javelin/video.mp4
+++ b/masters/roboolympics-fun/javelin/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:404301378155861053661384d4633f78284488efb274ae111bd205a62b3f767d
-size 10747105
+oid sha256:02dec00f1d68da6cb32aa4a0d1bddcab1096dde6dd3b438dc5d9107d363aa9b9
+size 10661133
diff --git a/masters/roboolympics-fun/juggle_three/meta.json b/masters/roboolympics-fun/juggle_three/meta.json
index feb558fbd966105f1b20676f23a5b47c55dbff7b..448316575b7e635f15f60c95c3b8ef57b1359dae 100644
--- a/masters/roboolympics-fun/juggle_three/meta.json
+++ b/masters/roboolympics-fun/juggle_three/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-juggle-three-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:claude_code-opus_5_5-high-embodied-olympics-fun-juggle-three-i00-privileged/embodied-olympics-fun-juggle-thr__E3xTcmn",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (tosser, tosser_gate, tosser_ram, tosser_wheel; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -51,20 +51,32 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -50.0,
-  "az_rule": "facing-50",
+  "az_deg": -103.5,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^tosser",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "tosser",
+   "tosser_gate",
+   "tosser_ram",
+   "tosser_wheel0",
+   "tosser_wheel1"
+  ],
   "distance_m": [
-   3.216,
-   4.722
+   3.933,
+   5.621
   ],
   "robot_root": "pelvis",
   "travel_m": 0.008,
   "focus": {
-   "robot+object": 144,
-   "robot": 65
+   "robot+object+equip": 144,
+   "robot+equip": 65
   },
   "objects": {
    "ball_a": 0.8,
@@ -80,7 +92,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.66,
+ "seconds_per_frame": 1.99,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "Claude Code + Opus 5.5 (high)",
  "run": "claude_code-opus_5_5-high",
@@ -88,5 +100,5 @@
  "trial_choice": "no GPT-6 Astra success; a counted success of the next run in order (Opus, Argon, ...)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:43-0500"
+ "created": "2026-10-25T09:57:16-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/juggle_three/poster.png b/masters/roboolympics-fun/juggle_three/poster.png
index 7ce5b20de6402ee06f2b092cd59b0e4a522d6a4c..c62dea94dca39e26dd3cc490bb463e627febc08a 100644
--- a/masters/roboolympics-fun/juggle_three/poster.png
+++ b/masters/roboolympics-fun/juggle_three/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4acbace4d334c2bff9f26c04740bb6cf8ea0a9b73ee824abac0d0f72baeb436d
-size 3089447
+oid sha256:5a92886a28cf4da85cb692bca8c87d62ae760b7b2f43e935a52deabd27dd8e8a
+size 3092238
diff --git a/masters/roboolympics-fun/juggle_three/video.mp4 b/masters/roboolympics-fun/juggle_three/video.mp4
index 1767944a6bb85010e49518495773fa25b764aced..734f7e72f1eaee8a009bbf3013a29bf094233fd9 100644
--- a/masters/roboolympics-fun/juggle_three/video.mp4
+++ b/masters/roboolympics-fun/juggle_three/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:250c69c742aa2d0e14f9cb217991a20edb196aedc4523f202fa5bd59bc852d11
-size 4391512
+oid sha256:06cd78b11831c59460f5525084b4a3ce2be8b6908653d74933c9c31f9ea4d4f4
+size 4457888
diff --git a/masters/roboolympics-fun/juggling/meta.json b/masters/roboolympics-fun/juggling/meta.json
index b1b87d19696bf48aeb3b019188da8e9d1c00a7c8..b4fca4775495b799af7d31551613a111b606f227 100644
--- a/masters/roboolympics-fun/juggling/meta.json
+++ b/masters/roboolympics-fun/juggling/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-juggling-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-juggling-i00-privileged/embodied-olympics-fun-juggling-i__d9MbrKA",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -63,16 +63,22 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.27,
-   5.265
+   3.464,
+   5.567
   ],
   "robot_root": "pelvis",
   "travel_m": 0.075,
   "focus": {
-   "robot+object": 200,
-   "robot": 140
+   "robot": 140,
+   "robot+object": 200
   },
   "objects": {
    "ball": 0.38
@@ -86,7 +92,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.56,
+ "seconds_per_frame": 1.98,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -94,5 +100,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:43-0500"
+ "created": "2026-10-25T09:57:16-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/juggling/poster.png b/masters/roboolympics-fun/juggling/poster.png
index 3cb36c73d491a03c9d1df5a730a221b1aebb82f2..212d1900149bdb6721c693cda11c5bc87555f79a 100644
--- a/masters/roboolympics-fun/juggling/poster.png
+++ b/masters/roboolympics-fun/juggling/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8fa82ecc99a44c6aa2afd07e016487556cdc136c9ae69a4c552ac7ea1fec8797
-size 3058387
+oid sha256:d5ef0a272b34c59c3eb33113fe3b1c656b53c4d69a35185c750bc9d9dac64557
+size 3049714
diff --git a/masters/roboolympics-fun/juggling/video.mp4 b/masters/roboolympics-fun/juggling/video.mp4
index e430626eb7de396d37b4180b3c026ed2525f1ce2..831a6dc76d93656147f4933f80c209e0a25fd38b 100644
--- a/masters/roboolympics-fun/juggling/video.mp4
+++ b/masters/roboolympics-fun/juggling/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:124ca24738e4db83b686db412254e40bb30933ef3c743a61aadf51d5137b4f64
-size 6213731
+oid sha256:a747a9001d87602eb785f0b2d96c9bfba642b505ade1ef97568bcc24db89dfe7
+size 6092442
diff --git a/masters/roboolympics-fun/juggling_feet/meta.json b/masters/roboolympics-fun/juggling_feet/meta.json
index 20c89c2916662fa94c6d14173e6d80cc50c57f98..cb2d5681aa3a74766062effd51de8818f7823ce4 100644
--- a/masters/roboolympics-fun/juggling_feet/meta.json
+++ b/masters/roboolympics-fun/juggling_feet/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-juggling-feet-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-juggling-feet-i00-privileged/embodied-olympics-fun-juggling-f__SnUL9zt",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -64,16 +64,22 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.098,
-   3.264
+   3.281,
+   3.457
   ],
   "robot_root": "pelvis",
   "travel_m": 0.172,
   "focus": {
-   "robot": 50,
-   "robot+object": 208
+   "robot+object": 208,
+   "robot": 50
   },
   "objects": {
    "ball": 0.19
@@ -87,7 +93,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.61,
+ "seconds_per_frame": 1.97,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -95,5 +101,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:43-0500"
+ "created": "2026-10-25T09:57:16-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/juggling_feet/poster.png b/masters/roboolympics-fun/juggling_feet/poster.png
index 747d1559d4cab1a2cc82390bcffb5bea166e3b93..32b0de04ae53a431a3c7da92efa86adca5bb3cf2 100644
--- a/masters/roboolympics-fun/juggling_feet/poster.png
+++ b/masters/roboolympics-fun/juggling_feet/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d96bee5bc4cbd96d3f4b1a4ba5ac00206966f73ee563fa858dd4eab334ba6129
-size 3135823
+oid sha256:8b455a287b90603cf30ea75e85c7e1e30b14926f4b5a094a0f6de1e9d8c6d677
+size 3126738
diff --git a/masters/roboolympics-fun/juggling_feet/video.mp4 b/masters/roboolympics-fun/juggling_feet/video.mp4
index 345a16f5dca498d2573677ce85d973fefc8eee99..f07a17d0b9c0914645399afd172cafc6e0adc388 100644
--- a/masters/roboolympics-fun/juggling_feet/video.mp4
+++ b/masters/roboolympics-fun/juggling_feet/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fa1edf3344b4e88e538bdba8ced7f0e0d12b4f6b790a032f2863adbf7e6372a3
-size 5523176
+oid sha256:a316ac6f19de546e5bd1c91872e5222636d20a858786539f755ec7adcb34c19f
+size 5385912
diff --git a/masters/roboolympics-fun/juggling_head/meta.json b/masters/roboolympics-fun/juggling_head/meta.json
index 34dfd74a7b05a210c907fa08011488b57ce52351..de1cb8867841c702cf524f74b14f130af817f277 100644
--- a/masters/roboolympics-fun/juggling_head/meta.json
+++ b/masters/roboolympics-fun/juggling_head/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-juggling-head-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-juggling-head-i00-privileged/embodied-olympics-fun-juggling-h__YtL3oAJ",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -63,10 +63,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   4.67,
-   6.438
+   4.935,
+   6.799
   ],
   "robot_root": "pelvis",
   "travel_m": 0.028,
@@ -87,7 +93,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.65,
+ "seconds_per_frame": 2.0,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -95,5 +101,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:43-0500"
+ "created": "2026-10-25T09:57:16-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/juggling_head/poster.png b/masters/roboolympics-fun/juggling_head/poster.png
index d3a4ceb0873d00fcb31e319fbebd6884447d2c39..c178d322435e56085ba9f4f6496f6e8e491aae50 100644
--- a/masters/roboolympics-fun/juggling_head/poster.png
+++ b/masters/roboolympics-fun/juggling_head/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6ef082de501b5d207e253890ef2ad3d5a384bec7c8421dc4c2151f356b33064e
-size 3132918
+oid sha256:a6f2b62b079832803d304b8bd41c42e2057f8905c6683864bf860d257cc965a4
+size 3126593
diff --git a/masters/roboolympics-fun/juggling_head/video.mp4 b/masters/roboolympics-fun/juggling_head/video.mp4
index 92519d1205e37d2865e8b6da12f5a9254ccb290b..8c573fbebb82e839ba824bf7b570d44edc0df433 100644
--- a/masters/roboolympics-fun/juggling_head/video.mp4
+++ b/masters/roboolympics-fun/juggling_head/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cb9241defef31711c935a277f6d6203de747abdb887559b4762b9e2904f0b814
-size 6180712
+oid sha256:ff44a1736ee74bb559574a976509be54a5093c6a259d0f42d723b19c2d66f357
+size 6115387
diff --git a/masters/roboolympics-fun/jump_rope/meta.json b/masters/roboolympics-fun/jump_rope/meta.json
index e5ed4f91942554cb64dc7cee11d4cd5f1403e9d0..a3eb10ff3b89a4a9f14d1060d90a06821297d0e1 100644
--- a/masters/roboolympics-fun/jump_rope/meta.json
+++ b/masters/roboolympics-fun/jump_rope/meta.json
@@ -7,7 +7,7 @@
  "source": "scene-orbit",
  "why_no_video": "No evaluated agent has a success on this task version yet. Site rule (boards 0161 / 0252): show GPT-6 Astra's failed attempt with the most progress instead. This video is that attempt replayed exactly (not a solved run): the task's own referee at the end says it failed, as the verifier did.",
  "shows": "agent-attempt (failed):codex-gpt6_astra-high-embodied-olympics-fun-jump-rope-i00-privileged/embodied-olympics-fun-jump-rope__VhNzNrn",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -59,16 +59,22 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.387,
-   3.499
+   3.584,
+   3.697
   ],
   "robot_root": "pelvis",
   "travel_m": 0.56,
   "focus": {
-   "robot+object": 286,
-   "robot": 1
+   "robot": 1,
+   "robot+object": 286
   },
   "objects": {
    "handle_left": 0.59,
@@ -111,7 +117,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.64,
+ "seconds_per_frame": 1.99,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -119,5 +125,5 @@
  "trial_choice": "no success on this task version; GPT-6 Astra's failed attempt (most progress)",
  "poster": "last frame (end of the failed attempt)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:44-0500"
+ "created": "2026-10-25T09:57:16-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/jump_rope/poster.png b/masters/roboolympics-fun/jump_rope/poster.png
index 91065dff3dfad494d62faf4f7495baf384c3510c..981aaef73b10988ecbfa7075551d2c611597c588 100644
--- a/masters/roboolympics-fun/jump_rope/poster.png
+++ b/masters/roboolympics-fun/jump_rope/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:52278e60056f74d6e8ab641efb393164feef98a1870539530c17e9b1daae5f5e
-size 3117959
+oid sha256:8312874c9993319306a497f13c6e664ec3a9285245200cdf4204501202fcf31f
+size 3104809
diff --git a/masters/roboolympics-fun/jump_rope/video.mp4 b/masters/roboolympics-fun/jump_rope/video.mp4
index 62581354f184605b87a9f7748d463701285bd970..ef9ed7a5bab000bd03bd876c2707777a9f421a29 100644
--- a/masters/roboolympics-fun/jump_rope/video.mp4
+++ b/masters/roboolympics-fun/jump_rope/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b0825c3e6b32989f8cfa709001b29b7f7c2cb7b9cd5df65dd5c0d7338a306a4f
-size 6450165
+oid sha256:d5318211c34793b84a2d201f2fe6a996908a11b53082a7cc32dc75c67bccbe1d
+size 6170290
diff --git a/masters/roboolympics-fun/kick_target/meta.json b/masters/roboolympics-fun/kick_target/meta.json
index 46cae2f74f6e29ef60cea1999dd7b74c39ab7904..82652bafc3958c9787e68052fa8f8b2dc36b5b30 100644
--- a/masters/roboolympics-fun/kick_target/meta.json
+++ b/masters/roboolympics-fun/kick_target/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-kick-target-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-kick-target-i00-privileged/embodied-olympics-fun-kick-targe__D9hQQeK",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (frame_arm, frame_post; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -53,20 +53,29 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -50.0,
-  "az_rule": "facing-50",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^frame_",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "frame_arm",
+   "frame_post"
+  ],
   "distance_m": [
-   5.142,
-   5.167
+   6.644,
+   6.67
   ],
   "robot_root": "pelvis",
   "travel_m": 0.111,
   "focus": {
-   "robot": 59,
-   "robot+object": 126
+   "robot+equip": 59,
+   "robot+object+equip": 126
   },
   "objects": {
    "bag": 0.28
@@ -80,13 +89,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.63,
+ "seconds_per_frame": 2.0,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 160 (same scene, camera and settings as the video)",
+ "poster": "frame 160 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:44-0500"
+ "created": "2026-10-25T09:57:16-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/kick_target/poster.png b/masters/roboolympics-fun/kick_target/poster.png
index 05a39a15394612ef46aab6554aade2cf9840be1a..2a2a172a4eacd1c79a7666ce1cf03189f31ef41e 100644
--- a/masters/roboolympics-fun/kick_target/poster.png
+++ b/masters/roboolympics-fun/kick_target/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b0210231dc9eba4d503d6ab34aa1d3205ff85324f908b8b14ec1e039398ffe09
-size 3156743
+oid sha256:b2c0bfaf11987bc8cd0f80b83486e2d977b5abda8e0ba904353a9f14c783e688
+size 3068583
diff --git a/masters/roboolympics-fun/kick_target/video.mp4 b/masters/roboolympics-fun/kick_target/video.mp4
index e0679f04a8415f8408033ca09d956ee4b0729b6b..12da31ae12e32d6d1209c5a964b367bf56d13cc9 100644
--- a/masters/roboolympics-fun/kick_target/video.mp4
+++ b/masters/roboolympics-fun/kick_target/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:92fc646c990e315858f102f0bfa28ea6e500aae6d542d3304b4380b04bf422f2
-size 3884908
+oid sha256:d0a57e843339d9700d6bbd1d7517154d66bed2b745860ebd424bc377d7d925d9
+size 3186371
diff --git a/masters/roboolympics-fun/layup/meta.json b/masters/roboolympics-fun/layup/meta.json
index b9522c0f7c888e2e3d5d00f7f1958b1ee752bf1c..c27b4e88c61fe3cf0d319a717f8792ba3674f242 100644
--- a/masters/roboolympics-fun/layup/meta.json
+++ b/masters/roboolympics-fun/layup/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-layup-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-layup-i00-privileged/embodied-olympics-fun-layup-i00__MG8gHYw",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (hoop; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -54,20 +54,28 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -50.0,
-  "az_rule": "facing-50",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^hoop_(rim|board|plate|net)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "hoop"
+  ],
   "distance_m": [
-   4.25,
-   8.211
+   11.045,
+   11.485
   ],
   "robot_root": "pelvis",
   "travel_m": 1.147,
   "focus": {
-   "robot+object": 107,
-   "robot": 69
+   "robot+object+equip": 107,
+   "robot+equip": 69
   },
   "objects": {
    "ball": 2.7
@@ -81,13 +89,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.9,
+ "seconds_per_frame": 2.07,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
  "mode": "privileged",
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
- "poster": "frame 133 (same scene, camera and settings as the video)",
+ "poster": "frame 133 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:44-0500"
+ "created": "2026-10-25T09:57:16-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/layup/poster.png b/masters/roboolympics-fun/layup/poster.png
index 9b0d0eb28c3309a5b9e738a382a790d60980a0de..2447c3b0c55022668ac01278692852763896b4af 100644
--- a/masters/roboolympics-fun/layup/poster.png
+++ b/masters/roboolympics-fun/layup/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e4774a2739f5745500d79e912003d3964cf717d6e102fc66e26b1d79927548d7
-size 3158165
+oid sha256:1fc093635eeab92121444a38e10539ca7799d13dd66c85c3d5672310c7e079db
+size 3201386
diff --git a/masters/roboolympics-fun/layup/video.mp4 b/masters/roboolympics-fun/layup/video.mp4
index adb20663fb2f265eabd1b7d8a9ddc318ce89e66d..45a8c009f42c2632c0feaea99e84b0cdebf2b883 100644
--- a/masters/roboolympics-fun/layup/video.mp4
+++ b/masters/roboolympics-fun/layup/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7c98b0f3d93beca87dcb619486efe9c613558f4e39e8a70a911ac92c291e7e3f
-size 4788944
+oid sha256:638f048ac7e9f6434e5940074882c6f8cdbb348f971d6090732a5a7b948f2966
+size 3609786
diff --git a/masters/roboolympics-fun/long_jump/meta.json b/masters/roboolympics-fun/long_jump/meta.json
index c350b4a57cef7c7cb2a5792733077440224fbf8f..092a541ef48d8fa5981f0dbc6725b46be8621c1c 100644
--- a/masters/roboolympics-fun/long_jump/meta.json
+++ b/masters/roboolympics-fun/long_jump/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-long-jump-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-long-jump-i00-privileged/embodied-olympics-fun-long-jump__mdnyE4R",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 0.0 s establishing shot framing the robot and the task's key equipment (sand; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,15 +57,24 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^sand",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 0.0,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "sand"
+  ],
   "distance_m": [
-   1.786,
-   3.889
+   3.337,
+   15.003
   ],
   "robot_root": "pelvis",
   "travel_m": 22.766,
   "focus": {
-   "robot": 309
+   "robot": 180,
+   "robot+equip": 129
   },
   "objects": {}
  },
@@ -77,7 +86,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.57,
+ "seconds_per_frame": 1.91,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -85,5 +94,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:44-0500"
+ "created": "2026-10-25T09:57:17-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/long_jump/poster.png b/masters/roboolympics-fun/long_jump/poster.png
index b7b49b57d025fceecbc0c89cfc6440c2fd4dc2dd..ff5233a83c58296f800352b298db6807a4e061cd 100644
--- a/masters/roboolympics-fun/long_jump/poster.png
+++ b/masters/roboolympics-fun/long_jump/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ad53969ea015869c7fb89491b65e67d100daf6a2d2a8a8019a23ba78bb01d572
-size 3164824
+oid sha256:875f2ffb008356e92c3514a889aad4b9e1ce0e0dccc4c7058465697512b5fe49
+size 2994502
diff --git a/masters/roboolympics-fun/long_jump/video.mp4 b/masters/roboolympics-fun/long_jump/video.mp4
index cf0e2ca4859c620b20afcfce3b098403898aa4a2..83cdab90224142ff2ea65a47596a92bec81906b1 100644
--- a/masters/roboolympics-fun/long_jump/video.mp4
+++ b/masters/roboolympics-fun/long_jump/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fda69ed651dd2c126f229cf6691c627b89654a8d228fb15ad4552862296e8a8e
-size 8150586
+oid sha256:0f3b62602273e8924bc004419206fe2a01735d8f28195ab4e6ff860a9607326e
+size 6598867
diff --git a/masters/roboolympics-fun/long_rope/meta.json b/masters/roboolympics-fun/long_rope/meta.json
index 9c58f4fa5b3a3f23b54db2c1d644819474fdd802..6efdbc6b142f82657ea56112c090336cc6923c29 100644
--- a/masters/roboolympics-fun/long_rope/meta.json
+++ b/masters/roboolympics-fun/long_rope/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-long-rope-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-long-rope-i00-privileged/embodied-olympics-fun-long-rope__JkqfFEE",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (machine+1_axle, machine+1_base, machine+1_motor, machine+1_post, machine-1_axle, machine-1_base, machine-1_motor, machine-1_post; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -60,20 +60,35 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -50.0,
-  "az_rule": "facing-50",
+  "az_deg": -25.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^machine",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "machine+1_axle",
+   "machine+1_base",
+   "machine+1_motor",
+   "machine+1_post",
+   "machine-1_axle",
+   "machine-1_base",
+   "machine-1_motor",
+   "machine-1_post"
+  ],
   "distance_m": [
-   4.706,
-   4.801
+   6.27,
+   6.391
   ],
   "robot_root": "pelvis",
   "travel_m": 0.041,
   "focus": {
-   "robot": 1,
-   "robot+object": 352
+   "robot+equip": 1,
+   "robot+object+equip": 352
   },
   "objects": {
    "crank_a": 0.3,
@@ -128,7 +143,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.66,
+ "seconds_per_frame": 1.97,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -136,5 +151,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:44-0500"
+ "created": "2026-10-25T09:57:17-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/long_rope/poster.png b/masters/roboolympics-fun/long_rope/poster.png
index 734330fbf1e702b082bbca27be386bae5c6c6943..e0c2ee24abc7d2dc1d76620fc36fe2f2e560edbf 100644
--- a/masters/roboolympics-fun/long_rope/poster.png
+++ b/masters/roboolympics-fun/long_rope/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:55aff98f35e85b94513bbfc72cbfd93f9b30a20d84ce86cc3a1c8049e43cab78
-size 3166255
+oid sha256:ead793c898664cdab2c09fe0a8cb93e0d47c28313190d1921b2337277a7ee58f
+size 3069877
diff --git a/masters/roboolympics-fun/long_rope/video.mp4 b/masters/roboolympics-fun/long_rope/video.mp4
index ae9eefb093de1badbc5d8574229a4963409fe9ab..d11c11cddb7121d1d628dbfa98c9762fe114e722 100644
--- a/masters/roboolympics-fun/long_rope/video.mp4
+++ b/masters/roboolympics-fun/long_rope/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bbdb63537828bb64ba2b744edd331c03fd7c80d504125989f7f6c86570cdf390
-size 7526302
+oid sha256:3ae42109ae3863b2d93984122ec8d071abbf941502a9b6536e8d6f5c2af6e709
+size 7097528
diff --git a/masters/roboolympics-fun/paddle_bounce/meta.json b/masters/roboolympics-fun/paddle_bounce/meta.json
index 95f197e2ed27f9276f62bfa72fe733e0d16eddc4..a6492b32000eea32afa85c0fe7def8dff972edf0 100644
--- a/masters/roboolympics-fun/paddle_bounce/meta.json
+++ b/masters/roboolympics-fun/paddle_bounce/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-paddle-bounce-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-paddle-bounce-i00-privileged/embodied-olympics-fun-paddle-bou__yr6NTKi",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -74,16 +74,22 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.714,
-   3.852
+   3.926,
+   4.068
   ],
   "robot_root": "pelvis",
   "travel_m": 0.002,
   "focus": {
-   "robot": 35,
-   "robot+object": 449
+   "robot+object": 449,
+   "robot": 35
   },
   "objects": {
    "dropper": 0.3,
@@ -99,7 +105,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.59,
+ "seconds_per_frame": 1.97,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -107,5 +113,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:44-0500"
+ "created": "2026-10-25T09:57:17-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/paddle_bounce/poster.png b/masters/roboolympics-fun/paddle_bounce/poster.png
index 1755306563ee8086f8f16af91ec905a22e73e9c6..a8fb3ef1055f0dba9717e1af0e69eb52877a8018 100644
--- a/masters/roboolympics-fun/paddle_bounce/poster.png
+++ b/masters/roboolympics-fun/paddle_bounce/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:24c5c4eecacfc4aa53ec05ffdcd9b1de94c4dbba517688818ab5d22d88445780
-size 3127825
+oid sha256:39e4d81872fed7afbf8a99ef8cc6b32e05ed33842d9a75954224b7cfea2bc0a8
+size 3120552
diff --git a/masters/roboolympics-fun/paddle_bounce/video.mp4 b/masters/roboolympics-fun/paddle_bounce/video.mp4
index 8c71f8f071a861b451ffb6504436092c2bb86a56..c767b3f4bd10fbd631763e3b4017706e84b9771b 100644
--- a/masters/roboolympics-fun/paddle_bounce/video.mp4
+++ b/masters/roboolympics-fun/paddle_bounce/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8ce6f2fc2ced5746916f377d59e33c1274b4eb02cd873b700ad77418665d7544
-size 6422969
+oid sha256:79a82fa285636ba49c4e48b757502e70f0aa822e9a0d2096670b2d95c5e071fb
+size 6280456
diff --git a/masters/roboolympics-fun/penalty_kick/meta.json b/masters/roboolympics-fun/penalty_kick/meta.json
index bce914507a369960e803d2a07070a090e9f8b46a..f13799717ecda1d637e9c2c8cd79ed1839f7170e 100644
--- a/masters/roboolympics-fun/penalty_kick/meta.json
+++ b/masters/roboolympics-fun/penalty_kick/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-penalty-kick-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-penalty-kick-i00-privileged/embodied-olympics-fun-penalty-ki__qWLCYv4",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (goal; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,17 +56,26 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^goal$",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "goal"
+  ],
   "distance_m": [
-   3.402,
-   12.31
+   3.602,
+   22.108
   ],
   "robot_root": "pelvis",
   "travel_m": 3.076,
   "focus": {
-   "object": 11,
-   "robot": 355,
-   "robot+object": 14
+   "robot+object": 11,
+   "object+equip": 11,
+   "robot+object+equip": 3,
+   "robot": 355
   },
   "objects": {
    "keeper": 0.94,
@@ -81,13 +90,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.6,
+ "seconds_per_frame": 2.1,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 357 (same scene, camera and settings as the video)",
+ "poster": "frame 357 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:45-0500"
+ "created": "2026-10-25T09:57:17-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/penalty_kick/poster.png b/masters/roboolympics-fun/penalty_kick/poster.png
index 2a1023cf4d0fea639dca30d16cee42823bcfa870..59d98e8f23cf7dca41f5c4ef67acd438e5901a4c 100644
--- a/masters/roboolympics-fun/penalty_kick/poster.png
+++ b/masters/roboolympics-fun/penalty_kick/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:476922ac37df2a30aa761c88954c672e177e4c0fd43f780ac3dd0469c5dbee3b
-size 3053194
+oid sha256:3836f03b7467c18b09e3b51d7d781fed6e6c714f6d4ef949bb9886ded07ec844
+size 3088045
diff --git a/masters/roboolympics-fun/penalty_kick/video.mp4 b/masters/roboolympics-fun/penalty_kick/video.mp4
index 41fc76244cfe31722ae4d2a1d011ac7f054535b2..cb70f12c26637b8ca0f85e8eec14355569f83bcc 100644
--- a/masters/roboolympics-fun/penalty_kick/video.mp4
+++ b/masters/roboolympics-fun/penalty_kick/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8369e7518f1be245677de10c4b92f4545a077aa97bf1d9145958804cbfed478b
-size 7597759
+oid sha256:7b167e45cf8c6c4bd8ffd6477aea6b71c6e858d8a18b3005b9a870880f259780
+size 7613235
diff --git a/masters/roboolympics-fun/pitch_target/meta.json b/masters/roboolympics-fun/pitch_target/meta.json
index ee420f87ce518cb1d451696b4723dd04ab7bf4b0..f8eb4cf5d7fa7cb79db0cd2ce7a853055efad592 100644
--- a/masters/roboolympics-fun/pitch_target/meta.json
+++ b/masters/roboolympics-fun/pitch_target/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-pitch-target-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-pitch-target-i00-privileged/embodied-olympics-fun-pitch-targ__C2UUpk9",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (zone_high, zone_low, zone_sideL, zone_sideR; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -54,21 +54,33 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -114.54,
-  "az_rule": "object-travel-115",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^zone_",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "zone_high",
+   "zone_low",
+   "zone_sideL",
+   "zone_sideR"
+  ],
   "distance_m": [
-   4.933,
-   17.033
+   12.863,
+   25.909
   ],
   "robot_root": "pelvis",
   "travel_m": 0.082,
   "focus": {
+   "robot": 11,
+   "object+equip": 40,
    "robot+object": 176,
-   "object": 77,
-   "robot": 11
+   "object": 37
   },
   "objects": {
    "ball": 17.72
@@ -82,13 +94,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.64,
+ "seconds_per_frame": 2.13,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
  "mode": "privileged",
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
- "poster": "frame 152 (same scene, camera and settings as the video)",
+ "poster": "frame 152 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:45-0500"
+ "created": "2026-10-25T09:57:17-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/pitch_target/poster.png b/masters/roboolympics-fun/pitch_target/poster.png
index 1547735db0199feea7705ecdfae6f25aa4fbd6ca..db44755ac7a2445fe9d43d68e567a9d35e51c530 100644
--- a/masters/roboolympics-fun/pitch_target/poster.png
+++ b/masters/roboolympics-fun/pitch_target/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:53628aa6a17b67e138e8e7f7319f1b862321140f2660a276ce616b05128605a0
-size 2966016
+oid sha256:0b71b8a44a9418933fad0675d74fe70ea8ac34b96608c1f984c128832fda29e4
+size 2983276
diff --git a/masters/roboolympics-fun/pitch_target/video.mp4 b/masters/roboolympics-fun/pitch_target/video.mp4
index e56639dc6f128e8caad7dd528c8f1d5e90315ead..687e0de1d9df892b2ba61a1ca5aea864cc81be2a 100644
--- a/masters/roboolympics-fun/pitch_target/video.mp4
+++ b/masters/roboolympics-fun/pitch_target/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8558e1c7402b5098d771b6305252c0b25f663176e7b2c40dc0f4ec60a6261979
-size 5190187
+oid sha256:35d59230be60cd492906d26dbf460c154d5defabe12534d1a211dd01856e9b7a
+size 4709952
diff --git a/masters/roboolympics-fun/platform_10m/meta.json b/masters/roboolympics-fun/platform_10m/meta.json
index 35c2a068ac715e06c37777feb1b422c8740edcb7..35ce5158aed0ab6896491a958ecef0860894ad6c 100644
--- a/masters/roboolympics-fun/platform_10m/meta.json
+++ b/masters/roboolympics-fun/platform_10m/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-platform-10m-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-platform-10m-i00-privileged/embodied-olympics-fun-platform-1__xbuUTG2",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; the pool's water drawn as refractive water (IOR 1.33, gentle wave bump, blue-green absorption); a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; the pool's water drawn as refractive water (IOR 1.33, gentle wave bump, blue-green absorption); a director camera: a 0.0 s establishing shot framing the robot and the task's key equipment (platform; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -66,15 +66,24 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^platform$",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 0.0,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "platform"
+  ],
   "distance_m": [
-   3.94,
-   13.034
+   6.838,
+   21.18
   ],
   "robot_root": "pelvis",
   "travel_m": 2.43,
   "focus": {
-   "robot": 252
+   "robot+equip": 212,
+   "robot": 40
   },
   "objects": {}
  },
@@ -86,13 +95,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 3.41,
+ "seconds_per_frame": 2.8,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 80 (same scene, camera and settings as the video)",
+ "poster": "frame 80 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:45-0500"
+ "created": "2026-10-25T09:57:17-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/platform_10m/poster.png b/masters/roboolympics-fun/platform_10m/poster.png
index da3a4ae52b22feba64f2b887c570e990c0f72d1f..cfc380ddfaf5fecd89c1a61c311cef9d9f0c5681 100644
--- a/masters/roboolympics-fun/platform_10m/poster.png
+++ b/masters/roboolympics-fun/platform_10m/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:feb8217ccb98aa406c9c3efeb9c1fabf08760d8f26d00bf5ffc48e25d5d9f920
-size 3155599
+oid sha256:37c0508809e92e7b0f4f8d7f520e9e67fd73b76978a5110fa6d2e7aa26754666
+size 3193906
diff --git a/masters/roboolympics-fun/platform_10m/video.mp4 b/masters/roboolympics-fun/platform_10m/video.mp4
index 18bc75d846a8c9f7a969d38eb71399e427fdd799..0f4fadc447d9d12aed977e0f12fd9420245c8178 100644
--- a/masters/roboolympics-fun/platform_10m/video.mp4
+++ b/masters/roboolympics-fun/platform_10m/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d02ddb2d28ec40fd8d5f12a17a319f6bab270835391c7a142ecdd34fb7297f61
-size 5868567
+oid sha256:46e93a0feb8a4e353be5515ce73073c59f96032b312019e3c8b285bfd5b93982
+size 5649218
diff --git a/masters/roboolympics-fun/pole_vault/meta.json b/masters/roboolympics-fun/pole_vault/meta.json
index 76e9077b7bb6150d28f38a62cf8595644aad77b6..676939882b05f60b5c97005cc9490ad2b99e8d05 100644
--- a/masters/roboolympics-fun/pole_vault/meta.json
+++ b/masters/roboolympics-fun/pole_vault/meta.json
@@ -7,7 +7,7 @@
  "source": "scene-orbit",
  "why_no_video": "No evaluated agent has a success on this task version yet. Site rule (boards 0161 / 0252): show GPT-6 Astra's failed attempt with the most progress instead. This video is that attempt replayed exactly (not a solved run): the task's own referee at the end says it failed, as the verifier did.",
  "shows": "agent-attempt (failed):codex-gpt6_astra-high-embodied-olympics-fun-pole-vault-i00-privileged/embodied-olympics-fun-pole-vault__C5Ja6o2",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 0.0 s establishing shot framing the robot and the task's key equipment (bar_pegL, bar_pegR, bar_uprightL, bar_uprightR, box_bottom, box_paint, box_sideL, box_sideR, box_stop; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -54,16 +54,33 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(bar_upright|bar_peg|box_)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 0.0,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "bar_pegL",
+   "bar_pegR",
+   "bar_uprightL",
+   "bar_uprightR",
+   "box_bottom",
+   "box_paint",
+   "box_sideL",
+   "box_sideR",
+   "box_stop"
+  ],
   "distance_m": [
-   5.38,
-   8.755
+   5.683,
+   15.001
   ],
   "robot_root": "pelvis",
   "travel_m": 20.285,
   "focus": {
-   "robot+object": 217,
-   "robot": 29
+   "robot": 29,
+   "robot+object+equip": 91,
+   "robot+object": 126
   },
   "objects": {
    "bar": 0.28,
@@ -78,7 +95,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.58,
+ "seconds_per_frame": 1.97,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -86,5 +103,5 @@
  "trial_choice": "no success on this task version; GPT-6 Astra's failed attempt (most progress)",
  "poster": "last frame (end of the failed attempt)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:45-0500"
+ "created": "2026-10-25T09:57:17-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/pole_vault/poster.png b/masters/roboolympics-fun/pole_vault/poster.png
index e733a99fb307b1d19b8b9e2a72ff38ce078f6dea..ff4d43f1c810974c12499df4a8a88470e0dcfc9e 100644
--- a/masters/roboolympics-fun/pole_vault/poster.png
+++ b/masters/roboolympics-fun/pole_vault/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8931a3c61c3e04ae7c0671d07ada051c3c8b38faf00eeae7c18da29a0f3d7773
-size 3078315
+oid sha256:cf37d130c967a90aa8bc8fef1bd59f2348fa19065fb59888a5d12756970e5b4a
+size 3020297
diff --git a/masters/roboolympics-fun/pole_vault/video.mp4 b/masters/roboolympics-fun/pole_vault/video.mp4
index d913d51e7e317aec831dd60e9ca4fe5b4a9be042..7a3382a01181f5a83a13b6834336d93685d34f5a 100644
--- a/masters/roboolympics-fun/pole_vault/video.mp4
+++ b/masters/roboolympics-fun/pole_vault/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e8a236cdc8bbfcd65239631c2bf41d996bcdf2eea4ea4ec5f02457de4c601a44
-size 6146674
+oid sha256:2dc78a1af08b2db4074212fea705295638d946214ae6c4af7448d5e393a0c09a
+size 5774839
diff --git a/masters/roboolympics-fun/round_off/meta.json b/masters/roboolympics-fun/round_off/meta.json
index 8ea6c8663e536682d663e556c7c201298ddaa77f..867f1c17a8bb3fcc98d6e44a4990558681891892 100644
--- a/masters/roboolympics-fun/round_off/meta.json
+++ b/masters/roboolympics-fun/round_off/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-round-off-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-round-off-i00-privileged/embodied-olympics-fun-round-off__9HYxiBi",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,10 +56,16 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.271,
-   3.704
+   3.462,
+   3.92
   ],
   "robot_root": "pelvis",
   "travel_m": 2.099,
@@ -76,7 +82,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.61,
+ "seconds_per_frame": 2.0,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
@@ -84,5 +90,5 @@
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:45-0500"
+ "created": "2026-10-25T09:57:17-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/round_off/poster.png b/masters/roboolympics-fun/round_off/poster.png
index 3aeabec65cec7fb3dbdd133b0014c9447df75731..ea8a64110a914b5d1c48c774e63b3e25572c5c1e 100644
--- a/masters/roboolympics-fun/round_off/poster.png
+++ b/masters/roboolympics-fun/round_off/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a2b7455f6b306b0f4bd32f1257b5775d7712551da2df94b41179c4c28f62524a
-size 3088622
+oid sha256:4e6b88868c9ac5e603b78055815fcae8b2795664c96a654c859ecda2bad8e10d
+size 3076867
diff --git a/masters/roboolympics-fun/round_off/video.mp4 b/masters/roboolympics-fun/round_off/video.mp4
index c377698b4fde15a6e06682dbb7a821b4a1e9b935..34331ac1ed44d9e8cadecf0fe15ccfb3e9233e3c 100644
--- a/masters/roboolympics-fun/round_off/video.mp4
+++ b/masters/roboolympics-fun/round_off/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8d905605fccdf82a13c007b6cf20a5f1148efd789daa80e960dba1c170cf863a
-size 5492038
+oid sha256:bc20f4b364413b376aa3374908dbb900079e3ff691e176beee61bcb24a2819f9
+size 5314507
diff --git a/masters/roboolympics-fun/rugby_conversion/meta.json b/masters/roboolympics-fun/rugby_conversion/meta.json
index 338ba4256f2bed9232584dca24b61c74429e9ccd..62f211ba67cf061b0f106093063250cf9823684c 100644
--- a/masters/roboolympics-fun/rugby_conversion/meta.json
+++ b/masters/roboolympics-fun/rugby_conversion/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-rugby-conversion-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-rugby-conversion-i00-privileged/embodied-olympics-fun-rugby-conv__WgWaLHT",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (crossbar, postL, postR; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -52,21 +52,31 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -115.41,
-  "az_rule": "object-travel-115",
+  "az_deg": -116.84,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(post[LR]|crossbar)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "crossbar",
+   "postL",
+   "postR"
+  ],
   "distance_m": [
-   13.458,
-   15.563
+   14.679,
+   17.463
   ],
   "robot_root": "pelvis",
   "travel_m": 0.002,
   "focus": {
-   "robot+object": 89,
-   "robot": 7,
-   "object": 29
+   "robot+object+equip": 89,
+   "object+equip": 29,
+   "robot+equip": 7
   },
   "objects": {
    "tee": 2.09,
@@ -81,13 +91,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.58,
+ "seconds_per_frame": 1.94,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 20 (same scene, camera and settings as the video)",
+ "poster": "frame 20 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:45-0500"
+ "created": "2026-10-25T09:57:17-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/rugby_conversion/poster.png b/masters/roboolympics-fun/rugby_conversion/poster.png
index bea1135b22195b8801fbca68b93333a9ed3c76cc..7faff1b0465120c54ea72e26b3d4495b0efe93e3 100644
--- a/masters/roboolympics-fun/rugby_conversion/poster.png
+++ b/masters/roboolympics-fun/rugby_conversion/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:85a6e306e658ab24711ef93b40f383b95b398b261e2f1a94bde273db0dc0b788
-size 3125207
+oid sha256:38e29bbe5d2704aef9288bd988bc974f29ae2a263f536c2347b7ea5fdcde14d9
+size 3117930
diff --git a/masters/roboolympics-fun/rugby_conversion/video.mp4 b/masters/roboolympics-fun/rugby_conversion/video.mp4
index 5bad9f554da6ac744793f32337c2e04ccf79e133..2fa6268844af4d24ccacd5c0a62261346804b743 100644
--- a/masters/roboolympics-fun/rugby_conversion/video.mp4
+++ b/masters/roboolympics-fun/rugby_conversion/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0999289140897d28f7892f59fb3d1dc0e9f9b8f6f14cdf1cdb3e59db15915a44
-size 3019427
+oid sha256:ad6c2dd123ca484dc1d9aa87235af8a0c997723ad94584c58debaafbe87680b2
+size 2826247
diff --git a/masters/roboolympics-fun/shot_put/meta.json b/masters/roboolympics-fun/shot_put/meta.json
index 6d0a5b50461d471d7275e0ba922334675033a05a..d18046c2c75469210ac68ed218cbf6c832fad096 100644
--- a/masters/roboolympics-fun/shot_put/meta.json
+++ b/masters/roboolympics-fun/shot_put/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-shot-put-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-shot-put-i00-privileged/embodied-olympics-fun-shot-put-i__WRpuCTL",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule object-travel-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -60,18 +60,24 @@
   "az_rule": "object-travel-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   4.377,
-   19.644
+   4.632,
+   19.715
   ],
   "robot_root": "pelvis",
   "travel_m": 0.816,
   "focus": {
-   "robot+landed": 61,
-   "robot+object": 41,
    "object": 101,
-   "robot": 335
+   "robot+object": 41,
+   "robot": 335,
+   "robot+landed": 61
   },
   "objects": {
    "shot": 12.58
@@ -85,13 +91,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.73,
+ "seconds_per_frame": 2.03,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 344 (same scene, camera and settings as the video)",
+ "poster": "frame 344 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:46-0500"
+ "created": "2026-10-25T09:57:17-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/shot_put/poster.png b/masters/roboolympics-fun/shot_put/poster.png
index b594dad4edf79001733740c52071e5a91047ed82..05fb8a4ac4aaf6c6d10bd614f2d31f909093b3da 100644
--- a/masters/roboolympics-fun/shot_put/poster.png
+++ b/masters/roboolympics-fun/shot_put/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3429c0b215f0a28896c66c06e465605c4e3c586a0be8dda3cca22cc374dd9072
-size 3105325
+oid sha256:d56a8c6a365d6b7e8393d5cf2d8387f1de98f4b5dde5258d5ea4338efc818660
+size 3105953
diff --git a/masters/roboolympics-fun/shot_put/video.mp4 b/masters/roboolympics-fun/shot_put/video.mp4
index b04868f275a814c525a9dfbc8759a3618a3bb0dc..c3d89205bf69981f0fe5345efab0f26efe5f7ecb 100644
--- a/masters/roboolympics-fun/shot_put/video.mp4
+++ b/masters/roboolympics-fun/shot_put/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1e4f1af2a11520aaa13df22a4cdc103b929bc21fa7f2ecd39d0631c10fe1ea58
-size 9773666
+oid sha256:d3e1a9c980a1d5b2bc2df63c208684ca7b4bbc31b65fe35ee86fc6a386c059a0
+size 9584096
diff --git a/masters/roboolympics-fun/shuttle_run/meta.json b/masters/roboolympics-fun/shuttle_run/meta.json
index 062683655791f85fdb12ff07b5dd4ea58d15d245..79cacbded7d0515710a381de71922630bdf4738f 100644
--- a/masters/roboolympics-fun/shuttle_run/meta.json
+++ b/masters/roboolympics-fun/shuttle_run/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-shuttle-run-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-shuttle-run-i00-privileged/embodied-olympics-fun-shuttle-ru__HTxEsWc",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,10 +57,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.107,
-   3.939
+   3.287,
+   3.93
   ],
   "robot_root": "pelvis",
   "travel_m": 0.812,
@@ -77,7 +83,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.62,
+ "seconds_per_frame": 1.97,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
@@ -85,5 +91,5 @@
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:46-0500"
+ "created": "2026-10-25T09:57:18-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/shuttle_run/poster.png b/masters/roboolympics-fun/shuttle_run/poster.png
index fba4775e779c641e96153ab1b512210a3b2b6888..f81d6aa2abb2da6e4a41f2f0d9bc669e09d4f1ba 100644
--- a/masters/roboolympics-fun/shuttle_run/poster.png
+++ b/masters/roboolympics-fun/shuttle_run/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f10d3a22768383984c44e13fa432d4f9d6690cd6c4d6eef879b55cfe4a249be4
-size 3078484
+oid sha256:3313796b4ffe1c6b5e2922b1c412259b2f5e6840155320f1625f10f7c0da0274
+size 3069689
diff --git a/masters/roboolympics-fun/shuttle_run/video.mp4 b/masters/roboolympics-fun/shuttle_run/video.mp4
index b2ecdb6658daeae05bf8668d5d45ecf79c34ec3f..d381a76849439d4ffc6ab18cb9770f636ba6a200 100644
--- a/masters/roboolympics-fun/shuttle_run/video.mp4
+++ b/masters/roboolympics-fun/shuttle_run/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9e7c752ada66caa2040cd2ba73a6b855e3c2a008f820d2653d9601bdb2075390
-size 15003552
+oid sha256:3277436370cc51503430140642f6258d33d07ac2e52d18ee7bb4d452b1831e9d
+size 14356704
diff --git a/masters/roboolympics-fun/skateboard_ollie/meta.json b/masters/roboolympics-fun/skateboard_ollie/meta.json
index 3086bdad7c32099d4407fade318c148d460d5527..5a4c583e4a1a83fa74f142548795a19a235b4ccd 100644
--- a/masters/roboolympics-fun/skateboard_ollie/meta.json
+++ b/masters/roboolympics-fun/skateboard_ollie/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-skateboard-ollie-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-skateboard-ollie-i00-privileged/embodied-olympics-fun-skateboard__chkrGqp",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,16 +56,22 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.29,
-   3.578
+   3.474,
+   3.784
   ],
   "robot_root": "pelvis",
   "travel_m": 0.048,
   "focus": {
-   "robot+object": 134,
-   "robot": 105
+   "robot": 105,
+   "robot+object": 134
   },
   "objects": {
    "board": 0.05
@@ -79,7 +85,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.57,
+ "seconds_per_frame": 1.97,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -87,5 +93,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:46-0500"
+ "created": "2026-10-25T09:57:18-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/skateboard_ollie/poster.png b/masters/roboolympics-fun/skateboard_ollie/poster.png
index a8f3135e3709572774f9ff4f38b5bf5ef72a17fe..a9204f4e3001a08cc891ce186aed4274b280fd8a 100644
--- a/masters/roboolympics-fun/skateboard_ollie/poster.png
+++ b/masters/roboolympics-fun/skateboard_ollie/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bbfe0fb8eaeaf8b1b9ac7ec9956a60b0c91a85eeffa0d6c053902b0fc92aea2d
-size 3075737
+oid sha256:ed4d7009ad267140032d4a8252cef0d60a1a9f0624df874a886111c6140ae198
+size 3067191
diff --git a/masters/roboolympics-fun/skateboard_ollie/video.mp4 b/masters/roboolympics-fun/skateboard_ollie/video.mp4
index 2e2b442eac494c42d56dced713632bc96d00bc7c..f61e1240e218dade674c98ab7f8f3423c3071786 100644
--- a/masters/roboolympics-fun/skateboard_ollie/video.mp4
+++ b/masters/roboolympics-fun/skateboard_ollie/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:be688d49ab0ae7a7733f5fc15801aa52958cc8626f15d36a5480eb885afe336a
-size 4132561
+oid sha256:4d564e83267e480710d876d5d8ce89414d05b3e4984f55650fb1ad7741e589a6
+size 4078050
diff --git a/masters/roboolympics-fun/slalom/meta.json b/masters/roboolympics-fun/slalom/meta.json
index 340fdd14c333cbad3d1ae63b9e38413536dbc37d..41dbe55f43eb1cd298962711131b2326b9f8a796 100644
--- a/masters/roboolympics-fun/slalom/meta.json
+++ b/masters/roboolympics-fun/slalom/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-slalom-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-slalom-i00-privileged/embodied-olympics-fun-slalom-i00__mvdtZvN",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (cone_centre, cone_far_left, cone_far_right, cone_finish, cone_start; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -65,15 +65,30 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^cone",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "cone_centre1",
+   "cone_centre2",
+   "cone_centre3",
+   "cone_centre4",
+   "cone_far_left",
+   "cone_far_right",
+   "cone_finish",
+   "cone_start"
+  ],
   "distance_m": [
-   3.154,
-   3.935
+   7.364,
+   12.577
   ],
   "robot_root": "pelvis",
   "travel_m": 3.697,
   "focus": {
-   "robot": 710
+   "robot+equip": 710
   },
   "objects": {}
  },
@@ -85,7 +100,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.6,
+ "seconds_per_frame": 1.93,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -93,5 +108,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:46-0500"
+ "created": "2026-10-25T09:57:18-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/slalom/poster.png b/masters/roboolympics-fun/slalom/poster.png
index a60fbb08d9642040afaf8c238be7fe185b4a3bb8..d296759db639824e480733b41ab23712bcd002e8 100644
--- a/masters/roboolympics-fun/slalom/poster.png
+++ b/masters/roboolympics-fun/slalom/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5337fa82e50cb9f23c3445baab2c8f819d81d1451b3b2401bb5ade71c449df83
-size 3204974
+oid sha256:c06f95efa68238382737261227770843543c7abdbc6120474b3ab1f4f9555d14
+size 3088828
diff --git a/masters/roboolympics-fun/slalom/video.mp4 b/masters/roboolympics-fun/slalom/video.mp4
index b7f53b9572889551c52533180384ff13caa84fbf..377efea0cecef41dc18b07d64d2da6b9563b8271 100644
--- a/masters/roboolympics-fun/slalom/video.mp4
+++ b/masters/roboolympics-fun/slalom/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dff2a4ca8745baa3be8269db96133531d774948a44f12e85dd8e2cbc3a34fd09
-size 17859480
+oid sha256:f243d9690f9eaf12dd818672be2889ff4515c28228be160574096c7142554c04
+size 11960447
diff --git a/masters/roboolympics-fun/sled_push/meta.json b/masters/roboolympics-fun/sled_push/meta.json
index 0b944a5059978307cce4c771eb7892f82004a143..e6a185fe87b4c5567ef3c2499b3e282e7415dc0c 100644
--- a/masters/roboolympics-fun/sled_push/meta.json
+++ b/masters/roboolympics-fun/sled_push/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-sled-push-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-sled-push-i00-privileged/embodied-olympics-fun-sled-push__6xkVG2A",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -55,10 +55,16 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   5.449,
-   5.666
+   5.763,
+   5.99
   ],
   "robot_root": "pelvis",
   "travel_m": 13.438,
@@ -78,7 +84,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.59,
+ "seconds_per_frame": 1.95,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
@@ -86,5 +92,5 @@
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:46-0500"
+ "created": "2026-10-25T09:57:18-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/sled_push/poster.png b/masters/roboolympics-fun/sled_push/poster.png
index 250521a62d0e95653c853310b021efbc2c1af705..d81e70e13a7b4d8e67a48cb3bf26714c421cbd9b 100644
--- a/masters/roboolympics-fun/sled_push/poster.png
+++ b/masters/roboolympics-fun/sled_push/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:202457efc08658f2f8ef98031170a30aa94cf5904e3e7203332ab39e7a653d9e
-size 3169197
+oid sha256:fca8af48bf16955296cda35a4d204c03b407019e0ab5aa9e35961be2710f6dce
+size 3163398
diff --git a/masters/roboolympics-fun/sled_push/video.mp4 b/masters/roboolympics-fun/sled_push/video.mp4
index d0f8af494a7b83214c5324f277495aa784d2619f..8925969ca60a6b1e7d86bed0eae0e64cca222d3b 100644
--- a/masters/roboolympics-fun/sled_push/video.mp4
+++ b/masters/roboolympics-fun/sled_push/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cc5683530a63507634ae196ed859fd3ed8ef6d3726ee74dc7c4e02033f00cbf8
-size 7355645
+oid sha256:3c86594d65f00f64475423aa3c347d00783c47063183a6a164f6d67113c63a97
+size 7195587
diff --git a/masters/roboolympics-fun/snowboard/meta.json b/masters/roboolympics-fun/snowboard/meta.json
index 76d324bf42cbae1f6d1daef8b7d11f752b95b09e..9372b5352f6dfd1f32e03ed3dde6bcba669fffa3 100644
--- a/masters/roboolympics-fun/snowboard/meta.json
+++ b/masters/roboolympics-fun/snowboard/meta.json
@@ -7,7 +7,7 @@
  "source": "scene-orbit",
  "why_no_video": "No evaluated agent has a success on this task version yet. Site rule (boards 0161 / 0252): show GPT-6 Astra's failed attempt with the most progress instead. This video is that attempt replayed exactly (not a solved run): the task's own referee at the end says it failed, as the verifier did.",
  "shows": "agent-attempt (failed):codex-gpt6_astra_quick-high-embodied-olympics-fun-snowboard-i00-privileged/embodied-olympics-fun-snowboard__sKGCCXG",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; snow venue: a snow shader (bright blue-white, rough, slight subsurface, fine bump) on the course's snow geoms, snow-white valley floor; a chase camera: per-frame fit of the robot (height included, so jumps and spins stay framed), at least 6 m away, 24 degree elevation, smoothed in time",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; snow venue: a snow shader (bright blue-white, rough, slight subsurface, fine bump) on the course's snow geoms, snow-white valley floor; a director camera: a 0.0 s establishing shot framing the robot and the task's key equipment (gate0_turn, gate10_turn, gate11_turn, gate12_turn, gate13_turn, gate14_turn, gate15_turn, gate16_turn, gate17_turn, gate1_turn, gate2_turn, gate3_turn, gate4_turn, gate5_turn, gate6_turn, gate7_turn, gate8_turn, gate9_turn; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 24.0-degree elevation, a slow eased 0.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -71,16 +71,42 @@
   "az_rule": "robot-travel-70",
   "el_deg": 24.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^gate\\d+_turn",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 0.0,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "gate0_turn",
+   "gate10_turn",
+   "gate11_turn",
+   "gate12_turn",
+   "gate13_turn",
+   "gate14_turn",
+   "gate15_turn",
+   "gate16_turn",
+   "gate17_turn",
+   "gate1_turn",
+   "gate2_turn",
+   "gate3_turn",
+   "gate4_turn",
+   "gate5_turn",
+   "gate6_turn",
+   "gate7_turn",
+   "gate8_turn",
+   "gate9_turn"
+  ],
   "distance_m": [
    6.24,
-   11.451
+   15.004
   ],
   "robot_root": "pelvis",
   "travel_m": 172.758,
   "focus": {
    "robot": 234,
-   "robot+object": 545
+   "robot+object+equip": 417,
+   "robot+object": 128
   },
   "objects": {
    "gate0_turn": 0.0,
@@ -102,7 +128,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 4.56,
+ "seconds_per_frame": 3.78,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
@@ -110,5 +136,5 @@
  "trial_choice": "no success on this task version; GPT-6 Astra's failed attempt (most progress)",
  "poster": "last frame (end of the failed attempt)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:46-0500"
+ "created": "2026-10-25T09:57:18-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/snowboard/poster.png b/masters/roboolympics-fun/snowboard/poster.png
index 75a04c471c513c2260f5cc0bdd913cb830abcbb1..6eb68651316672f6a2138c4d832b174a34c77a3e 100644
--- a/masters/roboolympics-fun/snowboard/poster.png
+++ b/masters/roboolympics-fun/snowboard/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:89496f810fbaa964d2dd8f6465d27b034c5eac0d43db4bab4f234b8fe5459248
-size 2914692
+oid sha256:25a297444738fa723fb0a9981910fd5b1a29b28bb53e736954ea37531b0c7991
+size 2909150
diff --git a/masters/roboolympics-fun/snowboard/video.mp4 b/masters/roboolympics-fun/snowboard/video.mp4
index 2f100c84df9652cbedfc7cb5ae97fa5a87fab08e..ef551eae56cdeffa1709799114430ee2f032fdab 100644
--- a/masters/roboolympics-fun/snowboard/video.mp4
+++ b/masters/roboolympics-fun/snowboard/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6b15d7e1c6e3ec18df853c2890048e719685aa4d05945e06f28377c1818fa63e
-size 9803920
+oid sha256:f551cc67fdfeb9e35a0e6789c67211a2438cf488028c9776b36791343625cf90
+size 9136888
diff --git a/masters/roboolympics-fun/springboard_3m/meta.json b/masters/roboolympics-fun/springboard_3m/meta.json
index b8ceb21f688189e3611452d6d37b2f6fc24ec110..c3d04b668916a2628524491ad3ff32681b8a452d 100644
--- a/masters/roboolympics-fun/springboard_3m/meta.json
+++ b/masters/roboolympics-fun/springboard_3m/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-springboard-3m-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-springboard-3m-i00-privileged/embodied-olympics-fun-springboar__vjAXTWo",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; the pool's water drawn as refractive water (IOR 1.33, gentle wave bump, blue-green absorption); a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; the pool's water drawn as refractive water (IOR 1.33, gentle wave bump, blue-green absorption); a director camera: a 0.0 s establishing shot framing the robot and the task's key equipment (board_fulcrum, board_hinge, board_rear, board_seg; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -67,16 +67,34 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^board_",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 0.0,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "board_fulcrum",
+   "board_hinge",
+   "board_rear",
+   "board_seg0",
+   "board_seg1",
+   "board_seg2",
+   "board_seg3",
+   "board_seg4",
+   "board_seg5",
+   "board_seg6",
+   "board_seg7"
+  ],
   "distance_m": [
-   5.09,
-   12.505
+   7.465,
+   13.367
   ],
   "robot_root": "pelvis",
   "travel_m": 1.919,
   "focus": {
-   "robot+object": 130,
-   "robot": 55
+   "robot+equip": 55,
+   "robot+object+equip": 130
   },
   "objects": {
    "board_seg0": 0.0
@@ -90,13 +108,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 4.27,
+ "seconds_per_frame": 3.44,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 59 (same scene, camera and settings as the video)",
+ "poster": "frame 59 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:47-0500"
+ "created": "2026-10-25T09:57:18-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/springboard_3m/poster.png b/masters/roboolympics-fun/springboard_3m/poster.png
index a3cdce502ea9cd53c2c7601199c7751f92a2fb16..21b9cf7bd273e261697258f1f2637129e57292dd 100644
--- a/masters/roboolympics-fun/springboard_3m/poster.png
+++ b/masters/roboolympics-fun/springboard_3m/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:be93182847eeba94e1eda052d5b69b2ec023a8acdfb20d7dacf0a798a2cfc9f6
-size 3189888
+oid sha256:73a2b4588cd35dcac4053dda8e74e33ad10871afe5d90dc4e1494c1546262b21
+size 3159520
diff --git a/masters/roboolympics-fun/springboard_3m/video.mp4 b/masters/roboolympics-fun/springboard_3m/video.mp4
index ef180af93e051d84426fd5fa5ce6aa65eafbb67f..990ca821073eea5f2fa124d476af83a91857d378 100644
--- a/masters/roboolympics-fun/springboard_3m/video.mp4
+++ b/masters/roboolympics-fun/springboard_3m/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1dd678d92e38406bef3872bc9b48a766f3d12cd9fe843de44a69c67cb534f594
-size 4296160
+oid sha256:0da89f96ca668d78540d1ddc085ba8218ce42e4fa76695719fd030d082d4446b
+size 4173044
diff --git a/masters/roboolympics-fun/sprint/meta.json b/masters/roboolympics-fun/sprint/meta.json
index c37526783239b711d9b59ebce1df0b7c66f532ef..d2966aa2f84f773bd2605845ee158e083be49593 100644
--- a/masters/roboolympics-fun/sprint/meta.json
+++ b/masters/roboolympics-fun/sprint/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-sprint-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-sprint-i00-privileged/embodied-olympics-fun-sprint-i00__pp2ccSJ",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,10 +57,16 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   2.999,
-   4.902
+   3.17,
+   4.91
   ],
   "robot_root": "pelvis",
   "travel_m": 49.883,
@@ -77,7 +83,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.6,
+ "seconds_per_frame": 1.98,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -85,5 +91,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:47-0500"
+ "created": "2026-10-25T09:57:18-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/sprint/poster.png b/masters/roboolympics-fun/sprint/poster.png
index 621c0f7b3fbd52cd6d0c4c981545b47aeda3ad4d..6ea310552a34757bab7b4edbf7ea323855bc66b8 100644
--- a/masters/roboolympics-fun/sprint/poster.png
+++ b/masters/roboolympics-fun/sprint/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0b98587029f7346f203eb54691181c3c1f7643c3d001a27d619cd0adeb05c52a
-size 3131054
+oid sha256:0feab059d4f9362a97e0868b96218a9c366e454549a2b848a8760357f65c9740
+size 3132507
diff --git a/masters/roboolympics-fun/sprint/video.mp4 b/masters/roboolympics-fun/sprint/video.mp4
index cbb4e5ca49a2050c3558a3217ead75eb4417f5ba..6720a8a1f25c849aefb897369ee4508563a3d194 100644
--- a/masters/roboolympics-fun/sprint/video.mp4
+++ b/masters/roboolympics-fun/sprint/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:61c652651b88052ff0191047a0e978786dc8bc9ca7fd2c08d712599a0e602933
-size 14024444
+oid sha256:af8a9c87cabcd8f530a042614dcfe4139116bc1b174b5f07a8d1a8440ec3627c
+size 13514551
diff --git a/masters/roboolympics-fun/standing_long_jump/meta.json b/masters/roboolympics-fun/standing_long_jump/meta.json
index d9232e4811cf0c7c68ca68465a565551f9803112..4562f7a35277c8c5f981c4faa0323a7f0b18c46e 100644
--- a/masters/roboolympics-fun/standing_long_jump/meta.json
+++ b/masters/roboolympics-fun/standing_long_jump/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-standing-long-jump-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-standing-long-jump-i00-privileged/embodied-olympics-fun-standing-l__6XUKDGd",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,10 +57,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.293,
-   4.363
+   3.484,
+   4.371
   ],
   "robot_root": "pelvis",
   "travel_m": 1.865,
@@ -77,7 +83,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.61,
+ "seconds_per_frame": 2.01,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -85,5 +91,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:47-0500"
+ "created": "2026-10-25T09:57:18-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/standing_long_jump/poster.png b/masters/roboolympics-fun/standing_long_jump/poster.png
index f5dc48cb7de8183bc9d29f9e812c986d74037a92..bd76d8443b7f09e19e42b97c77f47f82d5344e6e 100644
--- a/masters/roboolympics-fun/standing_long_jump/poster.png
+++ b/masters/roboolympics-fun/standing_long_jump/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:998a3cb8be611b87a9aed94afeca6cb3512ee8d8a3ed0028177d77a6e1683daf
-size 3174821
+oid sha256:ab566f3c85b44a09c9b7bfbb00c8baeb15a1e5cbb1daaf9f88e4bdcf2728cae1
+size 3158472
diff --git a/masters/roboolympics-fun/standing_long_jump/video.mp4 b/masters/roboolympics-fun/standing_long_jump/video.mp4
index d04723bc1f31884a120258b28720c9c72855fb2f..e188d152e468f317e68538d25c4734a6e2be1103 100644
--- a/masters/roboolympics-fun/standing_long_jump/video.mp4
+++ b/masters/roboolympics-fun/standing_long_jump/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5ad7fd202cc50cb5c2c2edce20080ab8cd9aa2861769cf6b664dbaec522f3a9f
-size 6523515
+oid sha256:ef7aaaa21ed70034b995d537e996180a8be794552b6433952d61e3a91d5909ba
+size 6381078
diff --git a/masters/roboolympics-fun/steeplechase/meta.json b/masters/roboolympics-fun/steeplechase/meta.json
index b2f79caffe2ac76a3c75a779cf752433c230b473..4ec2acc6bc49067cc5654fb212a5410dac4cb5a7 100644
--- a/masters/roboolympics-fun/steeplechase/meta.json
+++ b/masters/roboolympics-fun/steeplechase/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-steeplechase-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-steeplechase-i00-privileged/embodied-olympics-fun-steeplecha__q4qT2kF",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 0.0 s establishing shot framing the robot and the task's key equipment (hurdle_bar, hurdle_postL, hurdle_postR, water; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,15 +56,27 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(hurdle_|water)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 0.0,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "hurdle_bar",
+   "hurdle_postL",
+   "hurdle_postR",
+   "water"
+  ],
   "distance_m": [
-   3.886,
-   4.41
+   4.604,
+   14.484
   ],
   "robot_root": "pelvis",
   "travel_m": 14.413,
   "focus": {
-   "robot": 675
+   "robot": 174,
+   "robot+equip": 501
   },
   "objects": {}
  },
@@ -76,7 +88,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 3.14,
+ "seconds_per_frame": 2.23,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -84,5 +96,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:47-0500"
+ "created": "2026-10-25T09:57:18-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/steeplechase/poster.png b/masters/roboolympics-fun/steeplechase/poster.png
index 62df563901ef63133c03b79764ded61ab414ed1a..32d515e8fc765141e640317da1f67913b9be0abf 100644
--- a/masters/roboolympics-fun/steeplechase/poster.png
+++ b/masters/roboolympics-fun/steeplechase/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:48e604b05083fccd63db1247f9ca48b418e4437a1626f25f8011362b4be1c65a
-size 3170891
+oid sha256:502f06b3e794fbb69856a314ed4b18a75dc6fc175982d9a3b57f5ee9543307f6
+size 3146179
diff --git a/masters/roboolympics-fun/steeplechase/video.mp4 b/masters/roboolympics-fun/steeplechase/video.mp4
index 52d3ced4b5e07681114bbf39a7d0b0951d1b3b3a..2787f47a4912e263746442e990f967d53a2f33df 100644
--- a/masters/roboolympics-fun/steeplechase/video.mp4
+++ b/masters/roboolympics-fun/steeplechase/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0fc5a1e8e9e5eaded93214ab4c7bdf5ead064013275421bf96f251f903764cc2
-size 14309105
+oid sha256:13c5bdde2eed355db60b05f60453d279b73baff0a62fe5343a4b5e870dedba67
+size 11060661
diff --git a/masters/roboolympics-fun/sumo/meta.json b/masters/roboolympics-fun/sumo/meta.json
index 4e5d7027c83c85e4832baed92e2331785e5a27ad..a889a4114538be85b4b0212fb3607255dbc0acab 100644
--- a/masters/roboolympics-fun/sumo/meta.json
+++ b/masters/roboolympics-fun/sumo/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-sumo-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-sumo-i00-privileged/embodied-olympics-fun-sumo-i00-p__HgxJJxC",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -53,16 +53,22 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.741,
-   4.117
+   3.954,
+   4.354
   ],
   "robot_root": "pelvis",
   "travel_m": 1.88,
   "focus": {
-   "robot": 67,
-   "robot+object": 138
+   "robot+object": 138,
+   "robot": 67
   },
   "objects": {
    "opp_pelvis": 0.55
@@ -76,7 +82,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.73,
+ "seconds_per_frame": 2.06,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -84,5 +90,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:47-0500"
+ "created": "2026-10-25T09:57:18-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/sumo/poster.png b/masters/roboolympics-fun/sumo/poster.png
index dcdffa83d9c312b0b7013ba9ca976920ecd010c7..f9e7546fe745d605fd3e771710fc662e2b418976 100644
--- a/masters/roboolympics-fun/sumo/poster.png
+++ b/masters/roboolympics-fun/sumo/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2d6f3051374550973ce50bb9e16f4d392b150c5c3bab435f74a4a48632fa6a69
-size 3260155
+oid sha256:3a68e1bf1fff8d1a70d620d1aa4f94bba3b73fa84d27adc83a33e916317853bd
+size 3244475
diff --git a/masters/roboolympics-fun/sumo/video.mp4 b/masters/roboolympics-fun/sumo/video.mp4
index 8ded366e13ac1b30f3d2bbafb18eb2065f5836d3..65b0c049d6306f1dd30e9d2fda3d4bd1af074a3d 100644
--- a/masters/roboolympics-fun/sumo/video.mp4
+++ b/masters/roboolympics-fun/sumo/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e4474d8179a8a7a14360dd29bfc3c625db305c23b9f0c98cc483c3a3da41a13d
-size 5550747
+oid sha256:3dfedd60368c35f1b6f6dd041175962b5a95511bec9fb7a6d163d5a9abd5f314
+size 5376003
diff --git a/masters/roboolympics-fun/sweeper_jump/meta.json b/masters/roboolympics-fun/sweeper_jump/meta.json
index 9a955d8e3b11f101e2d5868fdb30a1d1bbb6b608..a2f30a931928e043cf84a177d794bafbc7fd15d9 100644
--- a/masters/roboolympics-fun/sweeper_jump/meta.json
+++ b/masters/roboolympics-fun/sweeper_jump/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-sweeper-jump-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-sweeper-jump-i00-privileged/embodied-olympics-fun-sweeper-ju__TqwYdsR",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,10 +57,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.316,
-   8.914
+   3.511,
+   8.904
   ],
   "robot_root": "pelvis",
   "travel_m": 0.126,
@@ -79,7 +85,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.71,
+ "seconds_per_frame": 2.06,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -87,5 +93,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:47-0500"
+ "created": "2026-10-25T09:57:19-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/sweeper_jump/poster.png b/masters/roboolympics-fun/sweeper_jump/poster.png
index 5c9709449f4fd5efa6ce0c8cb59241fbefd15b90..8d2f4dd6a1389ec166c757d1b86207bb341cf535 100644
--- a/masters/roboolympics-fun/sweeper_jump/poster.png
+++ b/masters/roboolympics-fun/sweeper_jump/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:76528b2d6d70ca162949c0efa6d4b7ffd32fbcb72d266b0bd0ff68555f007d38
-size 3132496
+oid sha256:8ce4f156ee14624865ae56a96e3479dd9ed2cd651dc23ad0013cde3f10197d64
+size 3124625
diff --git a/masters/roboolympics-fun/sweeper_jump/video.mp4 b/masters/roboolympics-fun/sweeper_jump/video.mp4
index 7305917592238be367c609050eb9f06ca65b8a04..2052f787d63f09e6ef0bc0e39b96f421208770d0 100644
--- a/masters/roboolympics-fun/sweeper_jump/video.mp4
+++ b/masters/roboolympics-fun/sweeper_jump/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e0c172ebe278846669a774dda5f13c3315f6636a9762e54c5a4d63b13259cba1
-size 14891831
+oid sha256:235361cfa4b1d2e57d11febab1d77a5859106a89bf211cc35e1b84a8580a4bf6
+size 14396643
diff --git a/masters/roboolympics-fun/table_tennis_return/meta.json b/masters/roboolympics-fun/table_tennis_return/meta.json
index 2227516f1f45dcfbbef1d6fbecea314aaea785fd..6fcff9e6f6b31e54a29dc2cf032654ba169608d8 100644
--- a/masters/roboolympics-fun/table_tennis_return/meta.json
+++ b/masters/roboolympics-fun/table_tennis_return/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-table-tennis-return-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-table-tennis-return-i00-privileged/embodied-olympics-fun-table-tenn__uMUjtpN",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (net_p, table; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -54,20 +54,34 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -50.0,
-  "az_rule": "facing-50",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(table$|net_p)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "net_p0",
+   "net_p1",
+   "net_p2",
+   "net_p3",
+   "net_p4",
+   "net_p5",
+   "table"
+  ],
   "distance_m": [
-   3.033,
-   5.167
+   4.764,
+   5.164
   ],
   "robot_root": "pelvis",
   "travel_m": 0.453,
   "focus": {
-   "robot+object": 184,
-   "robot": 57
+   "robot+equip": 57,
+   "robot+object+equip": 184
   },
   "objects": {
    "ball": 0.76,
@@ -82,13 +96,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.79,
+ "seconds_per_frame": 2.18,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 240 (same scene, camera and settings as the video)",
+ "poster": "frame 240 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:48-0500"
+ "created": "2026-10-25T09:57:19-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/table_tennis_return/poster.png b/masters/roboolympics-fun/table_tennis_return/poster.png
index 2cfe545a2e7bc0c152f16e86e6bea9ebf6ed5aa3..ca837f2b42b9657c5162203a85d62ae5f476fc64 100644
--- a/masters/roboolympics-fun/table_tennis_return/poster.png
+++ b/masters/roboolympics-fun/table_tennis_return/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:10d3d560bd278ad69ebfc68886b525007542c192136e18dbf2c5964a581a6143
-size 3142889
+oid sha256:67f531e6ed1ee3361efe27b7e86ac228f7263fb00a8ba6732012186381506841
+size 3162413
diff --git a/masters/roboolympics-fun/table_tennis_return/video.mp4 b/masters/roboolympics-fun/table_tennis_return/video.mp4
index e003339e3cbe8f578691e92ba47893ae38b4adcb..f83a9ffc1e510d7681d24bc590a351016a4e2397 100644
--- a/masters/roboolympics-fun/table_tennis_return/video.mp4
+++ b/masters/roboolympics-fun/table_tennis_return/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:864120cc60f42b8b8b1afabc01178e5c2297944aedd72b10802b98a00023856d
-size 5719878
+oid sha256:8bb994ae4dd5b061625e7b2f1903e353e053c1019eb4279e1a1a17db365eb5dd
+size 5036178
diff --git a/masters/roboolympics-fun/taekwondo_kick/meta.json b/masters/roboolympics-fun/taekwondo_kick/meta.json
index fff5dc6445994836a10afa0bdda424510842bcf2..c84f3594934dc65026e994de8d66c9e44eac0b71 100644
--- a/masters/roboolympics-fun/taekwondo_kick/meta.json
+++ b/masters/roboolympics-fun/taekwondo_kick/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-taekwondo-kick-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-taekwondo-kick-i00-privileged/embodied-olympics-fun-taekwondo__Sqd8zUs",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,10 +56,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   3.196,
-   3.341
+   3.386,
+   3.534
   ],
   "robot_root": "pelvis",
   "travel_m": 0.099,
@@ -76,7 +82,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.64,
+ "seconds_per_frame": 2.04,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -84,5 +90,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:48-0500"
+ "created": "2026-10-25T09:57:19-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/taekwondo_kick/poster.png b/masters/roboolympics-fun/taekwondo_kick/poster.png
index 907d191e2f6662221386f77635304ebb0eedafdc..0abcea72d168b2ddf0cf21278663dddeecdf5fba 100644
--- a/masters/roboolympics-fun/taekwondo_kick/poster.png
+++ b/masters/roboolympics-fun/taekwondo_kick/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9797f4f48b3aeb2633b30dd9923846b7035a0da4167313603af8a3b2c18844f5
-size 3238496
+oid sha256:223ce8b0e81b00afcb1caef5bd932c0b8881db1e2bcfa6a2791cf05583aed22b
+size 3228182
diff --git a/masters/roboolympics-fun/taekwondo_kick/video.mp4 b/masters/roboolympics-fun/taekwondo_kick/video.mp4
index 9e0f3f956842796469cf1f9a6db01f8d9f22607e..abbe8b97fa450df60b479a9997a4556cf11f761c 100644
--- a/masters/roboolympics-fun/taekwondo_kick/video.mp4
+++ b/masters/roboolympics-fun/taekwondo_kick/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9b5be1df24548ecfe0ea77394f460c6ef04b3bec9c2b6b28e46ca4ba23d59e1b
-size 5085340
+oid sha256:296f0ea082156b25dce9c3de4ab1011b7cb9a2d42311aafe04575b858fbe9eda
+size 5002451
diff --git a/masters/roboolympics-fun/teaser.mp4 b/masters/roboolympics-fun/teaser.mp4
index 35ecaae0f22b7f0aef36a300162817d38cee6ad1..61a1b07e5e8753a870ea70e3522476aeae397d41 100644
--- a/masters/roboolympics-fun/teaser.mp4
+++ b/masters/roboolympics-fun/teaser.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ad7cdfce402d206c89478423922feb7322dbd479f5e3da34ebceb3b96caa2b52
-size 5720370
+oid sha256:9777c53972d3cc65c57fd585affc725bef712f31b29e3972b6b5e7e63b7598fe
+size 5591664
diff --git a/masters/roboolympics-fun/teaser_poster.png b/masters/roboolympics-fun/teaser_poster.png
index 2a3c11f8da1edddffeaa73e07b3abaed4a7872f0..861e0920ae157bc1f1a3194662597c4c2618c810 100644
--- a/masters/roboolympics-fun/teaser_poster.png
+++ b/masters/roboolympics-fun/teaser_poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:18d98d68879eb9777501b8e346b1036397a08e7f540fcf2717d351b6996f041b
-size 3088548
+oid sha256:5ce61c85f020ece3e59206812880ffe906e609b519f77e54dcea4deb7baa611c
+size 3082184
diff --git a/masters/roboolympics-fun/tennis_return/meta.json b/masters/roboolympics-fun/tennis_return/meta.json
index 3e5847b7e223d3289e0700ba566cdcad57f0ecdf..949ee81e183270bd62ee405c5396699535c64501 100644
--- a/masters/roboolympics-fun/tennis_return/meta.json
+++ b/masters/roboolympics-fun/tennis_return/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-tennis-return-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-tennis-return-i00-privileged/embodied-olympics-fun-tennis-ret__hbqEGUK",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (net, net_L, net_R; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -53,22 +53,40 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -287.7,
-  "az_rule": "object-travel-115",
+  "az_deg": -116.83,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^net",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "net",
+   "net_L0",
+   "net_L1",
+   "net_L2",
+   "net_L3",
+   "net_R0",
+   "net_R1",
+   "net_R2",
+   "net_R3"
+  ],
   "distance_m": [
-   4.318,
-   13.752
+   15.542,
+   22.923
   ],
   "robot_root": "pelvis",
   "travel_m": 1.507,
   "focus": {
    "robot": 59,
-   "object": 38,
-   "robot(incoming)": 24,
-   "robot+object": 69
+   "robot+object+equip": 2,
+   "robot(incoming)": 6,
+   "robot(incoming)+equip": 18,
+   "robot+object": 67,
+   "object+equip": 38
   },
   "objects": {
    "server": 0.01,
@@ -84,13 +102,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.71,
+ "seconds_per_frame": 2.02,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
  "mode": "privileged",
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
- "poster": "frame 135 (same scene, camera and settings as the video)",
+ "poster": "frame 135 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:48-0500"
+ "created": "2026-10-25T09:57:19-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/tennis_return/poster.png b/masters/roboolympics-fun/tennis_return/poster.png
index 7e362343337d1068679162522e79336db5dc9451..9c3f34086c5848f00e7fda7265f4e7129d44fb31 100644
--- a/masters/roboolympics-fun/tennis_return/poster.png
+++ b/masters/roboolympics-fun/tennis_return/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8042632030f9903ef86aca8bded922d9f7f1a9df2be88142138cb5cb26bc3fef
-size 3052457
+oid sha256:74271c7805ad78667df4e1ecfb24a1e42a7001423f14393fa619c764de60d758
+size 3054530
diff --git a/masters/roboolympics-fun/tennis_return/video.mp4 b/masters/roboolympics-fun/tennis_return/video.mp4
index 8eac598d60cd0bfa6f1e182870fe29b8ac3420a3..604704e5afad416c2147d5ad8150fbebddcb7e45 100644
--- a/masters/roboolympics-fun/tennis_return/video.mp4
+++ b/masters/roboolympics-fun/tennis_return/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8dcfc162125af8db457fde0ef3781ac41bec42483eae981c8c0170087b7257eb
-size 4012382
+oid sha256:ac8532ec5f0c9673f1b7ad214daab5ba869e7d7de3fcdc276f5f4dbf64e0743e
+size 4126397
diff --git a/masters/roboolympics-fun/tennis_serve/meta.json b/masters/roboolympics-fun/tennis_serve/meta.json
index 3358d8f84b3b308b86bc391499de5772950c2b84..ad2d902c784630b30734e3dc8691c05a15e7ab74 100644
--- a/masters/roboolympics-fun/tennis_serve/meta.json
+++ b/masters/roboolympics-fun/tennis_serve/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-tennis-serve-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra_quick-high-embodied-olympics-fun-tennis-serve-i00-privileged/embodied-olympics-fun-tennis-ser__mTzrBbF",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (net, net_cord, net_seg0, net_seg1, net_seg10, net_seg11, net_seg12, net_seg13, net_seg14, net_seg15, net_seg16, net_seg17, net_seg18, net_seg19, net_seg2, net_seg3, net_seg4, net_seg5, net_seg6, net_seg7, net_seg8, net_seg9; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -54,21 +54,110 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -104.21,
-  "az_rule": "object-travel-115",
+  "az_deg": -113.86,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^net",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "net",
+   "net_cord0",
+   "net_cord1",
+   "net_cord10",
+   "net_cord11",
+   "net_cord12",
+   "net_cord13",
+   "net_cord14",
+   "net_cord15",
+   "net_cord16",
+   "net_cord17",
+   "net_cord18",
+   "net_cord19",
+   "net_cord2",
+   "net_cord3",
+   "net_cord4",
+   "net_cord5",
+   "net_cord6",
+   "net_cord7",
+   "net_cord8",
+   "net_cord9",
+   "net_seg0_0",
+   "net_seg0_1",
+   "net_seg0_2",
+   "net_seg10_0",
+   "net_seg10_1",
+   "net_seg10_2",
+   "net_seg11_0",
+   "net_seg11_1",
+   "net_seg11_2",
+   "net_seg12_0",
+   "net_seg12_1",
+   "net_seg12_2",
+   "net_seg13_0",
+   "net_seg13_1",
+   "net_seg13_2",
+   "net_seg14_0",
+   "net_seg14_1",
+   "net_seg14_2",
+   "net_seg15_0",
+   "net_seg15_1",
+   "net_seg15_2",
+   "net_seg16_0",
+   "net_seg16_1",
+   "net_seg16_2",
+   "net_seg17_0",
+   "net_seg17_1",
+   "net_seg17_2",
+   "net_seg18_0",
+   "net_seg18_1",
+   "net_seg18_2",
+   "net_seg19_0",
+   "net_seg19_1",
+   "net_seg19_2",
+   "net_seg1_0",
+   "net_seg1_1",
+   "net_seg1_2",
+   "net_seg2_0",
+   "net_seg2_1",
+   "net_seg2_2",
+   "net_seg3_0",
+   "net_seg3_1",
+   "net_seg3_2",
+   "net_seg4_0",
+   "net_seg4_1",
+   "net_seg4_2",
+   "net_seg5_0",
+   "net_seg5_1",
+   "net_seg5_2",
+   "net_seg6_0",
+   "net_seg6_1",
+   "net_seg6_2",
+   "net_seg7_0",
+   "net_seg7_1",
+   "net_seg7_2",
+   "net_seg8_0",
+   "net_seg8_1",
+   "net_seg8_2",
+   "net_seg9_0",
+   "net_seg9_1",
+   "net_seg9_2"
+  ],
   "distance_m": [
-   5.664,
-   15.68
+   18.553,
+   23.896
   ],
   "robot_root": "pelvis",
   "travel_m": 0.012,
   "focus": {
-   "robot+object": 131,
+   "robot+object+equip": 2,
    "robot": 8,
-   "object": 77
+   "object+equip": 77,
+   "robot+object": 129
   },
   "objects": {
    "ball": 15.46,
@@ -83,13 +172,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.87,
+ "seconds_per_frame": 2.07,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high), quick-check run",
  "run": "codex-gpt6_astra_quick-high",
  "mode": "privileged",
  "trial_choice": "GPT-6 Astra's privileged success in the quick check of this task version (no formal run on it yet)",
- "poster": "frame 96 (same scene, camera and settings as the video)",
+ "poster": "frame 96 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:48-0500"
+ "created": "2026-10-25T09:57:19-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/tennis_serve/poster.png b/masters/roboolympics-fun/tennis_serve/poster.png
index 9b5ca7db77c560241782772bc46f585ec1487dac..79cd044fe3ef1966bf10bd1faa01cd6e19e2882c 100644
--- a/masters/roboolympics-fun/tennis_serve/poster.png
+++ b/masters/roboolympics-fun/tennis_serve/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9651838eed909bda6bb4f26b121525d0d04fbb9133aa6df16fec335647e5fe38
-size 3134247
+oid sha256:605e2b1900b04e7d6aaf9dc8c2987abbd9078b51cfdc46df35ff22b19d9efd12
+size 3129443
diff --git a/masters/roboolympics-fun/tennis_serve/video.mp4 b/masters/roboolympics-fun/tennis_serve/video.mp4
index 908d1799f910c8588640bf72d5ae9692a84df8fd..db97151115f3f65007866de1d491dbfc706d2683 100644
--- a/masters/roboolympics-fun/tennis_serve/video.mp4
+++ b/masters/roboolympics-fun/tennis_serve/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ab327c1ea09b9ec906b39d0a3ca0224a0b6657ef9f44af83aa4b08b14a10594
-size 5328967
+oid sha256:b0ecb9d72e7df789377de0538229ce87b6e5fc6555a8ae23891e53538a8bc207
+size 4728603
diff --git a/masters/roboolympics-fun/toes_to_bar/meta.json b/masters/roboolympics-fun/toes_to_bar/meta.json
index 6d6c6fdc6f76a0524b2839bd1a1c034879d52233..3eba73b47a2ed11c6ee9e733f081872506ba0c14 100644
--- a/masters/roboolympics-fun/toes_to_bar/meta.json
+++ b/masters/roboolympics-fun/toes_to_bar/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-toes-to-bar-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-toes-to-bar-i00-privileged/embodied-olympics-fun-toes-to-ba__x8iZgcy",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (bar; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -59,15 +59,23 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^bar$",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "bar"
+  ],
   "distance_m": [
-   2.267,
-   3.92
+   2.596,
+   4.338
   ],
   "robot_root": "pelvis",
   "travel_m": 0.049,
   "focus": {
-   "robot": 737
+   "robot+equip": 737
   },
   "objects": {}
  },
@@ -79,7 +87,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.6,
+ "seconds_per_frame": 1.99,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -87,5 +95,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:48-0500"
+ "created": "2026-10-25T09:57:19-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/toes_to_bar/poster.png b/masters/roboolympics-fun/toes_to_bar/poster.png
index fe3d87465c9161bb85d2f3ffde56dcafb5e32446..ad78b31b79e879e0521fcd1283102df627335529 100644
--- a/masters/roboolympics-fun/toes_to_bar/poster.png
+++ b/masters/roboolympics-fun/toes_to_bar/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b37209e0517bdbf22e36cc4cdba9d1323359f43062492987e1f352bd3082291d
-size 3160660
+oid sha256:339e467560e3e6bf54906a7c1025efbdbfc059690df43f66f86a204c19bd27d7
+size 3140277
diff --git a/masters/roboolympics-fun/toes_to_bar/video.mp4 b/masters/roboolympics-fun/toes_to_bar/video.mp4
index 75d3d190705dc5f84b192805d159265865528fc7..3c2bedb6b7378525c912eb1c75fc94976624ef0c 100644
--- a/masters/roboolympics-fun/toes_to_bar/video.mp4
+++ b/masters/roboolympics-fun/toes_to_bar/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0ac10aff752e22d8e22bff26d20b054deeb61f2a18093965df9f96f18fefbaae
-size 13523955
+oid sha256:bca7ed63aed4f497b0cf2e155b397fb2d2ba5b9ac423d3975e56d920b64ede85
+size 12823211
diff --git a/masters/roboolympics-fun/trampoline/meta.json b/masters/roboolympics-fun/trampoline/meta.json
index daa2b0bd631c8b0d1ddb982f77df1cc93e4d7f6f..3f01aae563f913593d5a9cac2d5ee2db012dc34a 100644
--- a/masters/roboolympics-fun/trampoline/meta.json
+++ b/masters/roboolympics-fun/trampoline/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-trampoline-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-trampoline-i00-privileged/embodied-olympics-fun-trampoline__THfW4gu",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,16 +57,22 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   6.415,
-   6.541
+   6.781,
+   6.914
   ],
   "robot_root": "pelvis",
   "travel_m": 0.059,
   "focus": {
-   "robot+object": 61,
-   "robot": 147
+   "robot": 147,
+   "robot+object": 61
   },
   "objects": {
    "bed": 0.0
@@ -80,7 +86,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.89,
+ "seconds_per_frame": 2.26,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -88,5 +94,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:48-0500"
+ "created": "2026-10-25T09:57:19-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/trampoline/poster.png b/masters/roboolympics-fun/trampoline/poster.png
index b4107e841a0b3b710c94be233c0cdf455269e163..52137fe4f83847a84cdc396b1d18bdaaf7f38756 100644
--- a/masters/roboolympics-fun/trampoline/poster.png
+++ b/masters/roboolympics-fun/trampoline/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d6624f5e7e3b5f41dd5f333c110b631a29137322998693041eb03773aea02ad3
-size 3266836
+oid sha256:b6e2c78aa0bddce8efc5e37bb907d2e7a7a448ab35344284a2d40ba20a29c701
+size 3250390
diff --git a/masters/roboolympics-fun/trampoline/video.mp4 b/masters/roboolympics-fun/trampoline/video.mp4
index afd1921c68251d4864559dda8986cbfa7330af69..68a6d96ca5fa3521277fae968fe241a06485b82e 100644
--- a/masters/roboolympics-fun/trampoline/video.mp4
+++ b/masters/roboolympics-fun/trampoline/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a148f335f3ec351586f8916f1339c328a2244a008587949c500a33b467a3a33
-size 5188501
+oid sha256:bb00274118a242dc3702faf0d07e60333a916b08b2e7162cef54b22d2bdc75e1
+size 5053202
diff --git a/masters/roboolympics-fun/trap_ball/meta.json b/masters/roboolympics-fun/trap_ball/meta.json
index fc9f4ac6e7a3213f7d733a962476d11b7b67a315..44f0785eb025e14f7ab844bda418b587c4dbd783 100644
--- a/masters/roboolympics-fun/trap_ball/meta.json
+++ b/masters/roboolympics-fun/trap_ball/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-trap-ball-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-trap-ball-i00-privileged/embodied-olympics-fun-trap-ball__PAXYgiG",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule object-travel-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,17 +57,23 @@
   "az_rule": "object-travel-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   10.927,
-   32.692
+   10.934,
+   32.696
   ],
   "robot_root": "pelvis",
   "travel_m": 0.328,
   "focus": {
-   "object": 61,
    "robot(incoming)": 99,
    "robot": 3,
+   "object": 61,
    "robot+object": 38
   },
   "objects": {
@@ -83,13 +89,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.56,
+ "seconds_per_frame": 1.92,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 169 (same scene, camera and settings as the video)",
+ "poster": "frame 169 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:49-0500"
+ "created": "2026-10-25T09:57:19-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/trap_ball/poster.png b/masters/roboolympics-fun/trap_ball/poster.png
index a131d079c2abe02ce1b22aed3880449e777fc390..780a713a0d4ee9fbc19fac2a96fade7cea3d3643 100644
--- a/masters/roboolympics-fun/trap_ball/poster.png
+++ b/masters/roboolympics-fun/trap_ball/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:575cadfc3dd00ec1fab8f73f9b7c7ddc9a5a8f1de9fe5ea4ba43b5f1b43c35d8
-size 2957243
+oid sha256:09134a31747397e6dc312fdea4b4fc5f6cc0e76f6e674d000ca221d1dcdac3b4
+size 2957454
diff --git a/masters/roboolympics-fun/trap_ball/video.mp4 b/masters/roboolympics-fun/trap_ball/video.mp4
index 94934475eed8f429b8fc5a751a30883ac008db4a..6f15688f9db7044c724564b7867bc57e70bec32e 100644
--- a/masters/roboolympics-fun/trap_ball/video.mp4
+++ b/masters/roboolympics-fun/trap_ball/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:160e790a2f9f93d1184247ebf2a946a55db4fb64fe1177968ed10a7b510194e3
-size 3942959
+oid sha256:f8b22d6527a6e9e3d12c8a4e76106cafa561822350528287bc0fe6f24d89584e
+size 3897449
diff --git a/masters/roboolympics-fun/triple_jump/meta.json b/masters/roboolympics-fun/triple_jump/meta.json
index e6bd1b2ffb5592adc7ea43cbd0fe0d1b80ba92ea..68df65e40e8bdea1d0566c4714819c87239a5550 100644
--- a/masters/roboolympics-fun/triple_jump/meta.json
+++ b/masters/roboolympics-fun/triple_jump/meta.json
@@ -7,7 +7,7 @@
  "source": "scene-orbit",
  "why_no_video": "No evaluated agent has a success on this task version yet. Site rule (boards 0161 / 0252): show GPT-6 Astra's failed attempt with the most progress instead. This video is that attempt replayed exactly (not a solved run): the task's own referee at the end says it failed, as the verifier did.",
  "shows": "agent-attempt (failed):codex-gpt6_astra-high-embodied-olympics-fun-triple-jump-i00-privileged/embodied-olympics-fun-triple-jum__QzLruGy",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 0.0 s establishing shot framing the robot and the task's key equipment (sand; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -61,15 +61,24 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^sand",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 0.0,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "sand"
+  ],
   "distance_m": [
-   2.97,
-   3.883
+   3.282,
+   15.017
   ],
   "robot_root": "pelvis",
   "travel_m": 26.423,
   "focus": {
-   "robot": 270
+   "robot": 227,
+   "robot+equip": 43
   },
   "objects": {}
  },
@@ -81,7 +90,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.54,
+ "seconds_per_frame": 1.94,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -89,5 +98,5 @@
  "trial_choice": "no success on this task version; GPT-6 Astra's failed attempt (most progress)",
  "poster": "last frame (end of the failed attempt)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:49-0500"
+ "created": "2026-10-25T09:57:19-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/triple_jump/poster.png b/masters/roboolympics-fun/triple_jump/poster.png
index 07cc5eccce5e0708781baf325064451fc5e06b06..e4c679fa78debe6c8b5037dcae314e34c48fa647 100644
--- a/masters/roboolympics-fun/triple_jump/poster.png
+++ b/masters/roboolympics-fun/triple_jump/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b6e486cd631be3ea688a3189559dc16d02e8698c26daf3cac31c1854092264ae
-size 3037777
+oid sha256:05a1c1ce65512547bff314e582fb56019fbd5ceb8a228a2dce49f887c0a2777a
+size 2960191
diff --git a/masters/roboolympics-fun/triple_jump/video.mp4 b/masters/roboolympics-fun/triple_jump/video.mp4
index 5a8dad6c43780de838251370e92c1253034f900c..1b8f560ad0e242b46cfb559f0f6683571fd6af14 100644
--- a/masters/roboolympics-fun/triple_jump/video.mp4
+++ b/masters/roboolympics-fun/triple_jump/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:88df6da3d9ad1cac9591c3374ca1e2abe73e5c3cca1b29d6973e2cd0ab1df95d
-size 7828365
+oid sha256:3f738b303a32f30de373ba2adcbf39c582fbb4628e601e5f14692117490519a7
+size 6715527
diff --git a/masters/roboolympics-fun/vault/meta.json b/masters/roboolympics-fun/vault/meta.json
index 323f4b818cc742424f86619e97d7b663c9e56123..44037a8dac2b185e56175a37f4d76b26f71b7528 100644
--- a/masters/roboolympics-fun/vault/meta.json
+++ b/masters/roboolympics-fun/vault/meta.json
@@ -7,7 +7,7 @@
  "source": "scene-orbit",
  "why_no_video": "No evaluated agent has a success on this task version yet. Site rule (boards 0161 / 0252): show GPT-6 Astra's failed attempt with the most progress instead. This video is that attempt replayed exactly (not a solved run): the task's own referee at the end says it failed, as the verifier did.",
  "shows": "agent-attempt (failed):codex-gpt6_astra-high-embodied-olympics-fun-vault-i00-privileged/embodied-olympics-fun-vault-i00__ojeVhV4",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 0.0 s establishing shot framing the robot and the task's key equipment (springboard, springboard_base, table_top; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -55,15 +55,26 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(table_top|springboard)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 0.0,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "springboard",
+   "springboard_base",
+   "table_top"
+  ],
   "distance_m": [
-   3.782,
-   4.862
+   4.166,
+   12.117
   ],
   "robot_root": "pelvis",
   "travel_m": 27.132,
   "focus": {
-   "robot": 701
+   "robot": 326,
+   "robot+equip": 375
   },
   "objects": {}
  },
@@ -75,7 +86,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.63,
+ "seconds_per_frame": 2.04,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -83,5 +94,5 @@
  "trial_choice": "no success on this task version; GPT-6 Astra's failed attempt (most progress)",
  "poster": "last frame (end of the failed attempt)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:49-0500"
+ "created": "2026-10-25T09:57:19-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/vault/poster.png b/masters/roboolympics-fun/vault/poster.png
index eac3c1017b7caa6dc47e65ca890409a2c269dc58..7e71186c5dbf75d1917d11af6b9a9f3fc5015fa3 100644
--- a/masters/roboolympics-fun/vault/poster.png
+++ b/masters/roboolympics-fun/vault/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8909a4f9413d26e3e9e3ce74b19d6850f8f3c50db146ce3a86bfb422c73a65d9
-size 3152122
+oid sha256:a73af7b6ae03503cea73b29e2d797e34b570efbcd927979d43148a7f4d92bdd1
+size 3146904
diff --git a/masters/roboolympics-fun/vault/video.mp4 b/masters/roboolympics-fun/vault/video.mp4
index 150830f88e95577f71efbc3ef40ac9066f4888da..d30e640efbe4bcd20e8d6942f49fb98d3ba54762 100644
--- a/masters/roboolympics-fun/vault/video.mp4
+++ b/masters/roboolympics-fun/vault/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d81ab3ece855a80e5f5d1c7dea1dfbfebac3951a23f976736c48736e251676ac
-size 14721898
+oid sha256:c7cf93bc54ec4e635cf2a30d0f8fc30f0038e7645d0b5cefda862101edff1f70
+size 14083153
diff --git a/masters/roboolympics-fun/vertical_jump/meta.json b/masters/roboolympics-fun/vertical_jump/meta.json
index 64df8e8ab9fc0ab6b3129bc7069bd91c2fdc7e6e..a19e08a7beef4db7a610835e98e5851ce9ca8e25 100644
--- a/masters/roboolympics-fun/vertical_jump/meta.json
+++ b/masters/roboolympics-fun/vertical_jump/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-vertical-jump-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-vertical-jump-i00-privileged/embodied-olympics-fun-vertical-j__kjbpYBB",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early); azimuth rule facing-50, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,10 +57,16 @@
   "az_rule": "facing-50",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [],
   "distance_m": [
-   5.819,
-   5.952
+   6.149,
+   6.309
   ],
   "robot_root": "pelvis",
   "travel_m": 0.215,
@@ -79,7 +85,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.61,
+ "seconds_per_frame": 1.99,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -87,5 +93,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:49-0500"
+ "created": "2026-10-25T09:57:19-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/vertical_jump/poster.png b/masters/roboolympics-fun/vertical_jump/poster.png
index 39897e7e415f1be87a396c856ac0f6686d2f17b6..8e8675850a9b4d115f015dfbb0b49cab120581c6 100644
--- a/masters/roboolympics-fun/vertical_jump/poster.png
+++ b/masters/roboolympics-fun/vertical_jump/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d3485234a85dc901902cc5f1d852aeb9c4c507ac44064a84f6a374b33d4abd8b
-size 3046555
+oid sha256:7f6c9664fde520845042e9b76adea2872e124b41cbc5e03e4a35bc665d575ab8
+size 3037622
diff --git a/masters/roboolympics-fun/vertical_jump/video.mp4 b/masters/roboolympics-fun/vertical_jump/video.mp4
index 19874f056fa6a86a359d7cc5e90aa82721f063c0..c6d2b5bf10f77bd9ae89f783684a9fa4ab32e55e 100644
--- a/masters/roboolympics-fun/vertical_jump/video.mp4
+++ b/masters/roboolympics-fun/vertical_jump/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e098f1d1c355554a05e71aaef83344f608b5f8ffc7b30c93079d93e55f64ada3
-size 3723221
+oid sha256:1eaf04cc19a9c3099a091bfc0a019613eb930bf9624fa319bda0124e1aee0878
+size 3678865
diff --git a/masters/roboolympics-fun/volley_shot/meta.json b/masters/roboolympics-fun/volley_shot/meta.json
index cc7164ca4d4db176c5cc9d19834c09c9321319e8..ace806c852205112c48b071a36a4f8791a3d5a16 100644
--- a/masters/roboolympics-fun/volley_shot/meta.json
+++ b/masters/roboolympics-fun/volley_shot/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-volley-shot-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-volley-shot-i00-privileged/embodied-olympics-fun-volley-sho__7Kg3EQH",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (goal; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -52,21 +52,31 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -190.83,
-  "az_rule": "object-travel-115",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^goal$",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "goal"
+  ],
   "distance_m": [
-   10.23,
-   32.454
+   11.499,
+   23.371
   ],
   "robot_root": "pelvis",
   "travel_m": 0.102,
   "focus": {
-   "object": 57,
    "robot": 1,
-   "robot+object": 35,
+   "robot+object+equip": 4,
+   "object": 49,
+   "object+equip": 8,
+   "robot+object": 31,
    "robot(incoming)": 93
   },
   "objects": {
@@ -82,13 +92,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.82,
+ "seconds_per_frame": 2.24,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 163 (same scene, camera and settings as the video)",
+ "poster": "frame 163 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:49-0500"
+ "created": "2026-10-25T09:57:20-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/volley_shot/poster.png b/masters/roboolympics-fun/volley_shot/poster.png
index e5724bb2a94da5c9bdf12973d2bea85671d22f11..b6075df33f70e42ceab73afeec008cf3a627e526 100644
--- a/masters/roboolympics-fun/volley_shot/poster.png
+++ b/masters/roboolympics-fun/volley_shot/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:62db62c2088f999e0c47f2412143f62ae51f62d6326e45d45c7439f3795a10cc
-size 3071094
+oid sha256:b2eab562416a97634c3dbe75dcef7a701246b0fc7b483cb02a0dd2012627a1d1
+size 3065223
diff --git a/masters/roboolympics-fun/volley_shot/video.mp4 b/masters/roboolympics-fun/volley_shot/video.mp4
index d72e6a3144eaa747245824d8c2c2b3cbc3c08dfd..ed55cad7d802c6c97bc543363ab74894e95c6f63 100644
--- a/masters/roboolympics-fun/volley_shot/video.mp4
+++ b/masters/roboolympics-fun/volley_shot/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2236829e9dbfbb0f94321ac6ec97e7cd78eea0c1480fee6f899e0d76c1aec2cf
-size 4782639
+oid sha256:d89636755ef7f440d7c2b152f73b182ea9e7807f535a47c8b84455049d851a3f
+size 3917842
diff --git a/masters/roboolympics-fun/volleyball_bump/meta.json b/masters/roboolympics-fun/volleyball_bump/meta.json
index 6619e1c044bf63ffed6335ccb49b3906b8addc68..22f74381960ebc9a47996f480b13c8085e1b96cf 100644
--- a/masters/roboolympics-fun/volleyball_bump/meta.json
+++ b/masters/roboolympics-fun/volleyball_bump/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-volleyball-bump-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-volleyball-bump-i00-privileged/embodied-olympics-fun-volleyball__hL2NK3F",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (net, target; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,22 +57,31 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -279.46,
-  "az_rule": "object-travel-115",
+  "az_deg": -131.7,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(net|target)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "net",
+   "target"
+  ],
   "distance_m": [
-   8.373,
-   33.383
+   12.381,
+   19.057
   ],
   "robot_root": "pelvis",
   "travel_m": 0.932,
   "focus": {
-   "robot+object": 75,
-   "robot(incoming)": 91,
-   "robot": 1,
-   "object": 32
+   "robot+equip": 1,
+   "object+equip": 32,
+   "robot+object+equip": 75,
+   "robot(incoming)+equip": 91
   },
   "objects": {
    "machine": 0.03,
@@ -87,13 +96,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.7,
+ "seconds_per_frame": 2.12,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 145 (same scene, camera and settings as the video)",
+ "poster": "frame 145 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:49-0500"
+ "created": "2026-10-25T09:57:20-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/volleyball_bump/poster.png b/masters/roboolympics-fun/volleyball_bump/poster.png
index 990914a66645b25fe7b0b3c3a2733c6a54a179b1..cc3957f640bd9527c21eef52693e71098cb76cac 100644
--- a/masters/roboolympics-fun/volleyball_bump/poster.png
+++ b/masters/roboolympics-fun/volleyball_bump/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ff59bf2f925452b1c397e8b514b1a95cd0ce9c95e088068fa5998c5086384da5
-size 3098821
+oid sha256:68f99fcadf0601478dd0a2d725ae3a0bb260a8f04dea2c2e4b1f2a82f9c19aff
+size 3195752
diff --git a/masters/roboolympics-fun/volleyball_bump/video.mp4 b/masters/roboolympics-fun/volleyball_bump/video.mp4
index b343b322581e14ed3c7dbdd47420bb4641d62d68..2c13ad46a8d1634a862f6a03ca03ab61c8d070d7 100644
--- a/masters/roboolympics-fun/volleyball_bump/video.mp4
+++ b/masters/roboolympics-fun/volleyball_bump/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dfc5b2b3488e33ef48ca92bd27fd9a1b4761629b6ede1d5e740a304cfcd11d8b
-size 4871630
+oid sha256:00966b295e844532ebf9231619bcf3cf64a93434e1084b5aabdf5377e5c0702e
+size 4538986
diff --git a/masters/roboolympics-fun/volleyball_serve/meta.json b/masters/roboolympics-fun/volleyball_serve/meta.json
index 2a8fa8a3628aaac7e890a750bc8d465ec9c10284..d3a417981dc939acde03e64098b3e7197d6b3733 100644
--- a/masters/roboolympics-fun/volleyball_serve/meta.json
+++ b/masters/roboolympics-fun/volleyball_serve/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-volleyball-serve-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-volleyball-serve-i00-privileged/embodied-olympics-fun-volleyball__kiujwRr",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (net; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -56,21 +56,30 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -115.42,
-  "az_rule": "object-travel-115",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^net",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "net"
+  ],
   "distance_m": [
-   10.736,
-   15.145
+   14.399,
+   16.574
   ],
   "robot_root": "pelvis",
   "travel_m": 0.009,
   "focus": {
-   "robot+object": 75,
    "robot": 15,
-   "object": 120
+   "object+equip": 120,
+   "robot+object+equip": 31,
+   "robot+object": 44
   },
   "objects": {
    "ball": 13.84
@@ -84,13 +93,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.75,
+ "seconds_per_frame": 2.07,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success",
- "poster": "frame 31 (same scene, camera and settings as the video)",
+ "poster": "frame 31 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:50-0500"
+ "created": "2026-10-25T09:57:20-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/volleyball_serve/poster.png b/masters/roboolympics-fun/volleyball_serve/poster.png
index 64e5bb0bff3d6176b92eb0db4582a51a36a4c2b0..06ac0d45c25490db6d6abb689e93f181055bf226 100644
--- a/masters/roboolympics-fun/volleyball_serve/poster.png
+++ b/masters/roboolympics-fun/volleyball_serve/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d2c69e65fd70e70fb6e8ff6dca2fcd0b36435e4cdf8c9e186d135fa77b1c231e
-size 3062444
+oid sha256:2e14779bb3747027123b7fd2e8230d5cd9184510de3cbb0a51e6a00eed7eb6eb
+size 3117990
diff --git a/masters/roboolympics-fun/volleyball_serve/video.mp4 b/masters/roboolympics-fun/volleyball_serve/video.mp4
index 45ae12216168cf5390d20e698c5f0557150ebcc3..65ffb3f7db0d921aefd2974e7d8b9428e933c6ac 100644
--- a/masters/roboolympics-fun/volleyball_serve/video.mp4
+++ b/masters/roboolympics-fun/volleyball_serve/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bc8ef0132f2a45b07fd25144a3b135b28b105105c41558ab5554b3a7b3c3a526
-size 5267889
+oid sha256:5db0944a203dbb5e3603b5f756daa207ce6d4f2d549caedc523047d8da350062
+size 4870736
diff --git a/masters/roboolympics-fun/volleyball_set/meta.json b/masters/roboolympics-fun/volleyball_set/meta.json
index b0323b0d523b41ede96bd50deec8da9c0abd20bb..dd39ce2fdd7cb5da715b16b843a1df60b2b63f72 100644
--- a/masters/roboolympics-fun/volleyball_set/meta.json
+++ b/masters/roboolympics-fun/volleyball_set/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-volleyball-set-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-volleyball-set-i00-privileged/embodied-olympics-fun-volleyball__X8pPtPq",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (net, target; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -57,20 +57,29 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -173.32,
-  "az_rule": "object-travel-115",
+  "az_deg": -148.69,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^(net|target)",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "net",
+   "target"
+  ],
   "distance_m": [
-   4.442,
-   12.387
+   11.32,
+   17.331
   ],
   "robot_root": "pelvis",
   "travel_m": 0.014,
   "focus": {
-   "robot": 96,
-   "robot+object": 108
+   "robot+object+equip": 108,
+   "robot+equip": 96
   },
   "objects": {
    "machine": 0.02,
@@ -85,7 +94,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.96,
+ "seconds_per_frame": 2.11,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -93,5 +102,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:50-0500"
+ "created": "2026-10-25T09:57:20-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/volleyball_set/poster.png b/masters/roboolympics-fun/volleyball_set/poster.png
index e545239f6a3a04bb62fce475d983bb31a7fd6738..41519ae0f7f2a18f5f4abbd51886ff215131c511 100644
--- a/masters/roboolympics-fun/volleyball_set/poster.png
+++ b/masters/roboolympics-fun/volleyball_set/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:af67f6d0d426cd79d5e34ac6f8ff400808d351b3bbbab9556ec3e6d9700d94c4
-size 3217910
+oid sha256:bfea70a1baa7c2e043407e967533da0de8b19f1d8875580ace07d0b5629b1020
+size 3151354
diff --git a/masters/roboolympics-fun/volleyball_set/video.mp4 b/masters/roboolympics-fun/volleyball_set/video.mp4
index a8a5c75b56ed8ee1603f88f906d0f3d4b8ac0292..82a6c32e9c28839f3d2eeee4b093a256cd0d9efc 100644
--- a/masters/roboolympics-fun/volleyball_set/video.mp4
+++ b/masters/roboolympics-fun/volleyball_set/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:207095f1fe554fabd983516eac52eed8fc6c151c39159b6246d110cc326d5738
-size 5237453
+oid sha256:85b016c1d2c101ddc7edf7fb51c1d625d027b02ab14d5d7d52065e3ca57b5d54
+size 4724010
diff --git a/masters/roboolympics-fun/volleyball_spike/meta.json b/masters/roboolympics-fun/volleyball_spike/meta.json
index 4f27e74f9e044413696859299759992d086a3f58..3460696532290d53efc9748ee09bd704ea341959 100644
--- a/masters/roboolympics-fun/volleyball_spike/meta.json
+++ b/masters/roboolympics-fun/volleyball_spike/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-volleyball-spike-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@4fde6ba91c",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-volleyball-spike-i00-privileged/embodied-olympics-fun-volleyball__gAT4uFP",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (net; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule robot-travel-70, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -63,17 +63,25 @@
   "az_rule": "robot-travel-70",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^net",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "net"
+  ],
   "distance_m": [
-   4.548,
-   10.605
+   8.02,
+   13.764
   ],
   "robot_root": "pelvis",
   "travel_m": 2.43,
   "focus": {
-   "object": 4,
-   "robot+object": 122,
-   "robot": 97
+   "robot+object+equip": 122,
+   "object+equip": 4,
+   "robot+equip": 97
   },
   "objects": {
    "machine": 0.02,
@@ -88,7 +96,7 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.87,
+ "seconds_per_frame": 2.14,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
@@ -96,5 +104,5 @@
  "trial_choice": "a counted GPT-6 Astra success",
  "poster": "last frame (solved state)",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T13:13:50-0500"
+ "created": "2026-10-25T09:57:20-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/volleyball_spike/poster.png b/masters/roboolympics-fun/volleyball_spike/poster.png
index 07874d378e7a53e09b20ce3fee0b458bcc911b64..4bda544cf3c395aaedefe2fcb076ee881a7d2292 100644
--- a/masters/roboolympics-fun/volleyball_spike/poster.png
+++ b/masters/roboolympics-fun/volleyball_spike/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:855e26a6e35eb7419b872952e70698920e39f55e98a26ea78d6edd867456dbc8
-size 3113351
+oid sha256:dcb28a236b94022d2a02cc5009e0359ab180998d6463d8df75110800f69bc105
+size 3105272
diff --git a/masters/roboolympics-fun/volleyball_spike/video.mp4 b/masters/roboolympics-fun/volleyball_spike/video.mp4
index feb92f4f9f9e8b9cc80b53b22aafec52c3eda6bf..4561559c35f0e58b37605d799c0f2e4b903cd450 100644
--- a/masters/roboolympics-fun/volleyball_spike/video.mp4
+++ b/masters/roboolympics-fun/volleyball_spike/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fb5b98c53f39748d3f5051935e7e9c4ab674df6ea209f61a7b3825e67f347b7e
-size 6214779
+oid sha256:d8cc1676466afbf8c15d6991d97d24948547f404abfa2619b06b03d4681cf2ea
+size 5363773
diff --git a/masters/roboolympics-fun/wall_ball/meta.json b/masters/roboolympics-fun/wall_ball/meta.json
index b76f7ccf01ba854363a7f777d4143d921210bc5a..1eff39047ca364f28bb8ca6f2a7d25a8f7d5664f 100644
--- a/masters/roboolympics-fun/wall_ball/meta.json
+++ b/masters/roboolympics-fun/wall_ball/meta.json
@@ -5,7 +5,7 @@
  "task_dir": "embodied-olympics-fun-wall-ball-i00-privileged",
  "task_version": "JamesKrW/embodied-first-exam@aef3b33b59",
  "source": "agent-success:codex-gpt6_astra-high-embodied-olympics-fun-wall-ball-i00-privileged/embodied-olympics-fun-wall-ball__is3AtBU",
- "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a tracking camera (per-frame fit of the robot and the object in play, smoothed in time, zooming out early) with a slow eased 10.0-degree orbit",
+ "renderer": "Blender 4.5.3 LTS Cycles path tracing (GPU OptiX), adaptive sampling (threshold 0.01, 64-128 spp), OpenImageDenoise with albedo + normal passes; kinematic playback of the exact MuJoCo replay poses (MuJoCo's own compiled meshes, primitives and colours); daylight: a sun lamp (2 degree disc) and a Nishita sky (no sun disc), the venue's own ground colour on a large ground plane with a faint world-space variation; AgX view transform, Medium High Contrast look; a director camera: a 1.5 s establishing shot framing the robot and the task's key equipment (target, target_centre; extents from the task's own scene geometry), eased over 1.5 s into a tracking camera (per-frame fit of the robot, the object in play, and the equipment while the action is within 6.0 m of it, of long equipment the 3.5 m nearest the action; smoothed in time, zooming out early); azimuth rule equip-115, fill 0.68, 34.0-degree vertical field of view, 22.0-degree elevation, a slow eased 10.0-degree orbit",
  "rendering": "path-tracing",
  "width": 2560,
  "height": 1440,
@@ -55,20 +55,29 @@
  ],
  "camera": {
   "mode": "director",
-  "az_deg": -50.0,
-  "az_rule": "facing-50",
+  "az_deg": -115.0,
+  "az_rule": "equip-115",
   "el_deg": 22.0,
   "fovy_deg": 34.0,
-  "fill": 0.72,
+  "fill": 0.68,
+  "equip": "^target",
+  "equip_r": 15.0,
+  "equip_near": 6.0,
+  "open_s": 1.5,
+  "ease_s": 1.5,
+  "equip_parts": [
+   "target",
+   "target_centre"
+  ],
   "distance_m": [
-   4.339,
-   10.56
+   7.565,
+   7.641
   ],
   "robot_root": "pelvis",
   "travel_m": 0.09,
   "focus": {
-   "robot": 52,
-   "robot+object": 273
+   "robot+object+equip": 273,
+   "robot+equip": 52
   },
   "objects": {
    "ball": 0.43
@@ -82,13 +91,13 @@
   "end_s": 1.5,
   "note": "first/last frame held, no extra simulation"
  },
- "seconds_per_frame": 2.68,
+ "seconds_per_frame": 2.29,
  "image": "ghcr.io/mll-lab-nu/efe:embodied-olympics-fun-0.2.1 + embodied-olympics-fun-files-0.3.3 (the task's tests/ verifier image)",
  "agent": "codex + GPT-6 Astra (high)",
  "run": "codex-gpt6_astra-high",
  "mode": "privileged",
  "trial_choice": "a counted GPT-6 Astra success: its formal privileged result on this task version (the quick check of aef3b33b59, reviewed and promoted)",
- "poster": "frame 175 (same scene, camera and settings as the video)",
+ "poster": "frame 175 of the video's own render",
  "encode": "libx264 preset slow CRF 16, yuv420p, bt709, +faststart",
- "created": "2026-10-24T21:09:41-0500"
+ "created": "2026-10-25T09:57:20-0500"
 }
\ No newline at end of file
diff --git a/masters/roboolympics-fun/wall_ball/poster.png b/masters/roboolympics-fun/wall_ball/poster.png
index 6caa093f894a595790fdfc176c8323b69b239bea..88dbbb37104ec16a07ecc4435dbf6d1c90763ce6 100644
--- a/masters/roboolympics-fun/wall_ball/poster.png
+++ b/masters/roboolympics-fun/wall_ball/poster.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:257cb5638e1d50831c3c75356016e4e99856b1adbb8be55a95652584e9ce8cfa
-size 3025004
+oid sha256:4f322be305512640c1217d9e99f0c54513aba221e4f2d1b37b321f380c93b474
+size 3045600
diff --git a/masters/roboolympics-fun/wall_ball/video.mp4 b/masters/roboolympics-fun/wall_ball/video.mp4
index 6b46d905c1a427225cfbfefdcdb1daa2a8523023..cd87259734a7dd82d164da5c87b981bbe7b3ae87 100644
--- a/masters/roboolympics-fun/wall_ball/video.mp4
+++ b/masters/roboolympics-fun/wall_ball/video.mp4
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8baaf1ddc443b1c727367af9c89ca8de57d8ab2a4cacf69fc63bf8a51e41c81b
-size 5859350
+oid sha256:3dbb45a730d3e3a4eb6e5c3d621606b3200dcf5cb7c107d9d8cd8ef077196a1c
+size 4497535