{
  "schema_version": "0.3.0",
  "annotation_revision": "public-main-2026-09-22",
  "dataset_id": "ego-atlas-public",
  "status": "assistant_draft_not_ground_truth",
  "semantic_method": "assistant_visual_review_of_timestamped_images",
  "local_semantic_model_run": false,
  "external_model_api_calls": 0,
  "time_convention": "seconds relative to first presented video frame; intervals [start,end)",
  "annotation_license": "CC BY-SA 4.0",
  "annotation_license_url": "https://creativecommons.org/licenses/by-sa/4.0/",
  "sampling_policy": "Full-frame overview about every 4s; chronological action pass about every 1s; selected bursts at 4fps. No exhaustive event coverage claim.",
  "hand_convention": "Anatomical hands inferred from first-person layout; absence of mirroring is not independently calibrated.",
  "featured_ids": [
    "M7",
    "E3",
    "E9",
    "E6",
    "M3",
    "M1"
  ],
  "clips": [
    {
      "id": "M7",
      "source": "https://media-demo.theroboticsstack.com/media/M7.mp4?v=86b014328911",
      "sha256": "86b01432891113e07a6c86d3faa695f76d2bc68d95292edc199d78efe46dc375",
      "duration_s": 74.633333,
      "width": 640,
      "height": 360,
      "fps_fraction": "30/1",
      "decoded_frame_count": 2239,
      "audio_stream_count": 0,
      "source_start_pts_s": 0.0,
      "last_frame_s": 74.6,
      "frame_interval_min_s": 0.03333299999999895,
      "frame_interval_max_s": 0.03333400000001063,
      "nonincreasing_pts_count": 0,
      "decode_diagnostics": "",
      "view_layout": "single_egocentric_full_frame",
      "sample_interval_target_s": 1.0,
      "extraction_settings": {
        "maximum_image_width": 1280,
        "crop": null,
        "frame_selection": "nearest_actual_PTS"
      },
      "title": "Chop sausage and fry an omelet",
      "observed_activity": "Cut a pale sausage into pieces on a wooden board at the counter, tilt the board to clear the pieces, stir a plate of beaten egg, carry it to the stove, pour the egg into a hot pan, return the emptied plate to the sink, and then work the setting omelet in the pan.",
      "inferred_goal": "Apparent preparation of an omelet with chopped sausage. Inferred from the visible ingredients and the cooking sequence, not from any supplied instruction.",
      "assigned_instruction": null,
      "playback_time_offset_s": 0.0,
      "visible_outcome": "A full yellow omelet occupies the pan at the final frame and is still being worked. No serving or plating is observed.",
      "task_success": "unknown",
      "unknowns": [
        "The published render is 640 by 360, so small-object identity and fine hand detail are limited.",
        "The item being cut is called sausage by the publisher. The imagery shows a pale block and reddish pieces, which is consistent but not independently conclusive.",
        "A white plate holding a flat brown item sits beside the hob from about 46 s. It is never identified.",
        "No task instruction was supplied; the goal is inferred from the visible sequence.",
        "Hand assignment follows the first-person layout, with the knife in the right hand throughout. No mirroring calibration was supplied.",
        "Cut cycles are grouped. No piece count is claimed.",
        "The clip ends with the omelet still in the pan, so task success is unknown."
      ],
      "selection_rationale": "Featured: four visually distinct stages with a workspace change mid-clip, two tool transitions, and a clear before-and-after on both the sausage and the egg.",
      "annotation_provenance": "assistant_visual_review",
      "review_status": "pending",
      "credit": {
        "attribution": "Mecka AI \u00b7 Egoverse \u00b7 recording 69bb080f92850385bd647356",
        "source_url": "https://partners.mecka.ai/api/egoverse/uploads/69bb080f92850385bd647356/video?redirect=1",
        "metadata_url": "https://partners.mecka.ai/egoverse?embodiment=human_bimanual&page=5",
        "license": "CC BY-SA 4.0",
        "license_url": "https://creativecommons.org/licenses/by-sa/4.0/",
        "changes": "Complete source published unchanged from object storage; no re-encoding. Evidence resized to at most 1280px; assistant temporal labels added."
      },
      "protocol_role": "featured",
      "objects": [
        {
          "id": "M7.board",
          "description": "Wooden cutting board",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M7.knife",
          "description": "Small kitchen knife",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M7.sausage",
          "description": "Pale sausage cut on the board; the publisher calls it sausage and the render does not resolve it independently",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M7.pieces",
          "description": "Group of chopped pieces on the board; count not established",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M7.plate",
          "description": "White plate holding the beaten egg",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M7.egg",
          "description": "Beaten egg mixture, bright orange",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M7.utensil",
          "description": "Spoon or fork resting in the egg plate; the two are not distinguishable at this resolution",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M7.pan",
          "description": "Dark square pan on the hob",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M7.omelet",
          "description": "Cooking egg mass in the pan",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M7.sink",
          "description": "Kitchen sink beside the board",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M7.yellow_item",
          "description": "Small yellow object handled beside the pan from about 65 s; identity not established",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        }
      ],
      "subtasks": [
        {
          "id": "M7.s01",
          "level": "subtask",
          "parent_id": null,
          "start_s": 0,
          "end_s": 2.0,
          "label": "Approach the counter and take up the knife",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M7.board",
            "M7.knife"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: approach the counter and take up the knife. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 0,
              "time_s": 0.0,
              "source_pts_s": 0.0,
              "image": "assets/frames/M7/f00000000.jpg"
            },
            {
              "frame_index": 30,
              "time_s": 1.0,
              "source_pts_s": 1.0,
              "image": "assets/frames/M7/f00000030.jpg"
            },
            {
              "frame_index": 60,
              "time_s": 2.0,
              "source_pts_s": 2.0,
              "image": "assets/frames/M7/f00000060.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.0,
                "image": "assets/frames/M7/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 2.0,
            "range_s": [
              1.0,
              3.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 30,
                "time_s": 1.0,
                "source_pts_s": 1.0,
                "image": "assets/frames/M7/f00000030.jpg"
              },
              {
                "frame_index": 90,
                "time_s": 3.0,
                "source_pts_s": 3.0,
                "image": "assets/frames/M7/f00000090.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.s02",
          "level": "subtask",
          "parent_id": null,
          "start_s": 2.0,
          "end_s": 32.367,
          "label": "Chop the sausage on the board",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M7.board",
            "M7.knife",
            "M7.sausage"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: chop the sausage on the board. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 60,
              "time_s": 2.0,
              "source_pts_s": 2.0,
              "image": "assets/frames/M7/f00000060.jpg"
            },
            {
              "frame_index": 510,
              "time_s": 17.0,
              "source_pts_s": 17.0,
              "image": "assets/frames/M7/f00000510.jpg"
            },
            {
              "frame_index": 960,
              "time_s": 32.0,
              "source_pts_s": 32.0,
              "image": "assets/frames/M7/f00000960.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 2.0,
            "range_s": [
              1.0,
              3.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 30,
                "time_s": 1.0,
                "source_pts_s": 1.0,
                "image": "assets/frames/M7/f00000030.jpg"
              },
              {
                "frame_index": 90,
                "time_s": 3.0,
                "source_pts_s": 3.0,
                "image": "assets/frames/M7/f00000090.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 32.367,
            "range_s": [
              32.0,
              33.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 960,
                "time_s": 32.0,
                "source_pts_s": 32.0,
                "image": "assets/frames/M7/f00000960.jpg"
              },
              {
                "frame_index": 990,
                "time_s": 33.0,
                "source_pts_s": 33.0,
                "image": "assets/frames/M7/f00000990.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.s03",
          "level": "subtask",
          "parent_id": null,
          "start_s": 32.367,
          "end_s": 37.867,
          "label": "Tilt the board and clear the chopped pieces",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M7.board",
            "M7.pieces",
            "M7.sink"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: tilt the board and clear the chopped pieces. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 960,
              "time_s": 32.0,
              "source_pts_s": 32.0,
              "image": "assets/frames/M7/f00000960.jpg"
            },
            {
              "frame_index": 1050,
              "time_s": 35.0,
              "source_pts_s": 35.0,
              "image": "assets/frames/M7/f00001050.jpg"
            },
            {
              "frame_index": 1140,
              "time_s": 38.0,
              "source_pts_s": 38.0,
              "image": "assets/frames/M7/f00001140.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 32.367,
            "range_s": [
              32.0,
              33.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 960,
                "time_s": 32.0,
                "source_pts_s": 32.0,
                "image": "assets/frames/M7/f00000960.jpg"
              },
              {
                "frame_index": 990,
                "time_s": 33.0,
                "source_pts_s": 33.0,
                "image": "assets/frames/M7/f00000990.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 37.867,
            "range_s": [
              37.0,
              38.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1110,
                "time_s": 37.0,
                "source_pts_s": 37.0,
                "image": "assets/frames/M7/f00001110.jpg"
              },
              {
                "frame_index": 1140,
                "time_s": 38.0,
                "source_pts_s": 38.0,
                "image": "assets/frames/M7/f00001140.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.s04",
          "level": "subtask",
          "parent_id": null,
          "start_s": 37.867,
          "end_s": 39.867,
          "label": "Stir the beaten egg in its plate",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M7.egg",
            "M7.plate",
            "M7.utensil"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: stir the beaten egg in its plate. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1140,
              "time_s": 38.0,
              "source_pts_s": 38.0,
              "image": "assets/frames/M7/f00001140.jpg"
            },
            {
              "frame_index": 1170,
              "time_s": 39.0,
              "source_pts_s": 39.0,
              "image": "assets/frames/M7/f00001170.jpg"
            },
            {
              "frame_index": 1200,
              "time_s": 40.0,
              "source_pts_s": 40.0,
              "image": "assets/frames/M7/f00001200.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 37.867,
            "range_s": [
              37.0,
              38.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1110,
                "time_s": 37.0,
                "source_pts_s": 37.0,
                "image": "assets/frames/M7/f00001110.jpg"
              },
              {
                "frame_index": 1140,
                "time_s": 38.0,
                "source_pts_s": 38.0,
                "image": "assets/frames/M7/f00001140.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 39.867,
            "range_s": [
              39.0,
              40.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1170,
                "time_s": 39.0,
                "source_pts_s": 39.0,
                "image": "assets/frames/M7/f00001170.jpg"
              },
              {
                "frame_index": 1200,
                "time_s": 40.0,
                "source_pts_s": 40.0,
                "image": "assets/frames/M7/f00001200.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.s05",
          "level": "subtask",
          "parent_id": null,
          "start_s": 39.867,
          "end_s": 48.867,
          "label": "Carry the egg plate to the stove",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M7.egg",
            "M7.plate"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: carry the egg plate to the stove. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1200,
              "time_s": 40.0,
              "source_pts_s": 40.0,
              "image": "assets/frames/M7/f00001200.jpg"
            },
            {
              "frame_index": 1320,
              "time_s": 44.0,
              "source_pts_s": 44.0,
              "image": "assets/frames/M7/f00001320.jpg"
            },
            {
              "frame_index": 1470,
              "time_s": 49.0,
              "source_pts_s": 49.0,
              "image": "assets/frames/M7/f00001470.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 39.867,
            "range_s": [
              39.0,
              40.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1170,
                "time_s": 39.0,
                "source_pts_s": 39.0,
                "image": "assets/frames/M7/f00001170.jpg"
              },
              {
                "frame_index": 1200,
                "time_s": 40.0,
                "source_pts_s": 40.0,
                "image": "assets/frames/M7/f00001200.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 48.867,
            "range_s": [
              48.0,
              49.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1440,
                "time_s": 48.0,
                "source_pts_s": 48.0,
                "image": "assets/frames/M7/f00001440.jpg"
              },
              {
                "frame_index": 1470,
                "time_s": 49.0,
                "source_pts_s": 49.0,
                "image": "assets/frames/M7/f00001470.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.s06",
          "level": "subtask",
          "parent_id": null,
          "start_s": 48.867,
          "end_s": 54.867,
          "label": "Pour the beaten egg into the pan",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M7.egg",
            "M7.pan",
            "M7.plate"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: pour the beaten egg into the pan. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1470,
              "time_s": 49.0,
              "source_pts_s": 49.0,
              "image": "assets/frames/M7/f00001470.jpg"
            },
            {
              "frame_index": 1560,
              "time_s": 52.0,
              "source_pts_s": 52.0,
              "image": "assets/frames/M7/f00001560.jpg"
            },
            {
              "frame_index": 1650,
              "time_s": 55.0,
              "source_pts_s": 55.0,
              "image": "assets/frames/M7/f00001650.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 48.867,
            "range_s": [
              48.0,
              49.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1440,
                "time_s": 48.0,
                "source_pts_s": 48.0,
                "image": "assets/frames/M7/f00001440.jpg"
              },
              {
                "frame_index": 1470,
                "time_s": 49.0,
                "source_pts_s": 49.0,
                "image": "assets/frames/M7/f00001470.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 54.867,
            "range_s": [
              54.0,
              55.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1620,
                "time_s": 54.0,
                "source_pts_s": 54.0,
                "image": "assets/frames/M7/f00001620.jpg"
              },
              {
                "frame_index": 1650,
                "time_s": 55.0,
                "source_pts_s": 55.0,
                "image": "assets/frames/M7/f00001650.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.s07",
          "level": "subtask",
          "parent_id": null,
          "start_s": 54.867,
          "end_s": 65.367,
          "label": "Return the emptied plate and tend the setting omelet",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M7.omelet",
            "M7.pan",
            "M7.plate",
            "M7.sink"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: return the emptied plate and tend the setting omelet. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1650,
              "time_s": 55.0,
              "source_pts_s": 55.0,
              "image": "assets/frames/M7/f00001650.jpg"
            },
            {
              "frame_index": 1800,
              "time_s": 60.0,
              "source_pts_s": 60.0,
              "image": "assets/frames/M7/f00001800.jpg"
            },
            {
              "frame_index": 1950,
              "time_s": 65.0,
              "source_pts_s": 65.0,
              "image": "assets/frames/M7/f00001950.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 54.867,
            "range_s": [
              54.0,
              55.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1620,
                "time_s": 54.0,
                "source_pts_s": 54.0,
                "image": "assets/frames/M7/f00001620.jpg"
              },
              {
                "frame_index": 1650,
                "time_s": 55.0,
                "source_pts_s": 55.0,
                "image": "assets/frames/M7/f00001650.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 65.367,
            "range_s": [
              65.0,
              66.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1950,
                "time_s": 65.0,
                "source_pts_s": 65.0,
                "image": "assets/frames/M7/f00001950.jpg"
              },
              {
                "frame_index": 1980,
                "time_s": 66.0,
                "source_pts_s": 66.0,
                "image": "assets/frames/M7/f00001980.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.s08",
          "level": "subtask",
          "parent_id": null,
          "start_s": 65.367,
          "end_s": 74.633333,
          "label": "Work the omelet in the pan",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M7.omelet",
            "M7.pan",
            "M7.yellow_item"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: work the omelet in the pan. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1950,
              "time_s": 65.0,
              "source_pts_s": 65.0,
              "image": "assets/frames/M7/f00001950.jpg"
            },
            {
              "frame_index": 2100,
              "time_s": 70.0,
              "source_pts_s": 70.0,
              "image": "assets/frames/M7/f00002100.jpg"
            },
            {
              "frame_index": 2238,
              "time_s": 74.6,
              "source_pts_s": 74.6,
              "image": "assets/frames/M7/f00002238.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 65.367,
            "range_s": [
              65.0,
              66.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1950,
                "time_s": 65.0,
                "source_pts_s": 65.0,
                "image": "assets/frames/M7/f00001950.jpg"
              },
              {
                "frame_index": 1980,
                "time_s": 66.0,
                "source_pts_s": 66.0,
                "image": "assets/frames/M7/f00001980.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 74.633333,
            "range_s": [
              74.633333,
              74.633333
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 2238,
                "time_s": 74.6,
                "source_pts_s": 74.6,
                "image": "assets/frames/M7/f00002238.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "actions": [
        {
          "id": "M7.a01",
          "level": "action",
          "parent_id": "M7.s01",
          "start_s": 0,
          "end_s": 2.0,
          "label": "Reach across the counter and settle at the board",
          "verb": "reach",
          "hands": "right",
          "hand_roles": {
            "right": "Reach across the counter and settle at the board"
          },
          "object_ids": [
            "M7.board",
            "M7.knife"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: reach across the counter and settle at the board. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 0,
              "time_s": 0.0,
              "source_pts_s": 0.0,
              "image": "assets/frames/M7/f00000000.jpg"
            },
            {
              "frame_index": 30,
              "time_s": 1.0,
              "source_pts_s": 1.0,
              "image": "assets/frames/M7/f00000030.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.0,
                "image": "assets/frames/M7/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 2.0,
            "range_s": [
              1.0,
              3.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 30,
                "time_s": 1.0,
                "source_pts_s": 1.0,
                "image": "assets/frames/M7/f00000030.jpg"
              },
              {
                "frame_index": 90,
                "time_s": 3.0,
                "source_pts_s": 3.0,
                "image": "assets/frames/M7/f00000090.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a02",
          "level": "action",
          "parent_id": "M7.s02",
          "start_s": 2.0,
          "end_s": 32.367,
          "label": "Cut the sausage into pieces with the knife",
          "verb": "cut",
          "hands": "right",
          "hand_roles": {
            "right": "Cut the sausage into pieces with the knife"
          },
          "object_ids": [
            "M7.knife",
            "M7.sausage",
            "M7.board"
          ],
          "state_before": "A pale sausage lies whole on the board beside the knife.",
          "state_after": "Chopped pieces are spread across the board and the sausage is gone.",
          "evidence_description": "Ordered sampled images support: cut the sausage into pieces with the knife. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 150,
              "time_s": 5.0,
              "source_pts_s": 5.0,
              "image": "assets/frames/M7/f00000150.jpg"
            },
            {
              "frame_index": 450,
              "time_s": 15.0,
              "source_pts_s": 15.0,
              "image": "assets/frames/M7/f00000450.jpg"
            },
            {
              "frame_index": 750,
              "time_s": 25.0,
              "source_pts_s": 25.0,
              "image": "assets/frames/M7/f00000750.jpg"
            },
            {
              "frame_index": 930,
              "time_s": 31.0,
              "source_pts_s": 31.0,
              "image": "assets/frames/M7/f00000930.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 2.0,
            "range_s": [
              1.0,
              3.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 30,
                "time_s": 1.0,
                "source_pts_s": 1.0,
                "image": "assets/frames/M7/f00000030.jpg"
              },
              {
                "frame_index": 90,
                "time_s": 3.0,
                "source_pts_s": 3.0,
                "image": "assets/frames/M7/f00000090.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 32.367,
            "range_s": [
              32.0,
              33.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 960,
                "time_s": 32.0,
                "source_pts_s": 32.0,
                "image": "assets/frames/M7/f00000960.jpg"
              },
              {
                "frame_index": 990,
                "time_s": 33.0,
                "source_pts_s": 33.0,
                "image": "assets/frames/M7/f00000990.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a03",
          "level": "action",
          "parent_id": "M7.s02",
          "start_s": 2.0,
          "end_s": 32.367,
          "label": "Hold and feed the sausage under the knife",
          "verb": "hold",
          "hands": "left",
          "hand_roles": {
            "left": "Hold and feed the sausage under the knife"
          },
          "object_ids": [
            "M7.sausage",
            "M7.board"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: hold and feed the sausage under the knife. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 240,
              "time_s": 8.0,
              "source_pts_s": 8.0,
              "image": "assets/frames/M7/f00000240.jpg"
            },
            {
              "frame_index": 540,
              "time_s": 18.0,
              "source_pts_s": 18.0,
              "image": "assets/frames/M7/f00000540.jpg"
            },
            {
              "frame_index": 840,
              "time_s": 28.0,
              "source_pts_s": 28.0,
              "image": "assets/frames/M7/f00000840.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 2.0,
            "range_s": [
              1.0,
              3.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 30,
                "time_s": 1.0,
                "source_pts_s": 1.0,
                "image": "assets/frames/M7/f00000030.jpg"
              },
              {
                "frame_index": 90,
                "time_s": 3.0,
                "source_pts_s": 3.0,
                "image": "assets/frames/M7/f00000090.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 32.367,
            "range_s": [
              32.0,
              33.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 960,
                "time_s": 32.0,
                "source_pts_s": 32.0,
                "image": "assets/frames/M7/f00000960.jpg"
              },
              {
                "frame_index": 990,
                "time_s": 33.0,
                "source_pts_s": 33.0,
                "image": "assets/frames/M7/f00000990.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a04",
          "level": "action",
          "parent_id": "M7.s03",
          "start_s": 32.367,
          "end_s": 35.0,
          "label": "Lift and tilt the cutting board",
          "verb": "lift",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M7.board",
            "M7.pieces"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: lift and tilt the cutting board. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 990,
              "time_s": 33.0,
              "source_pts_s": 33.0,
              "image": "assets/frames/M7/f00000990.jpg"
            },
            {
              "frame_index": 1020,
              "time_s": 34.0,
              "source_pts_s": 34.0,
              "image": "assets/frames/M7/f00001020.jpg"
            },
            {
              "frame_index": 1050,
              "time_s": 35.0,
              "source_pts_s": 35.0,
              "image": "assets/frames/M7/f00001050.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 32.367,
            "range_s": [
              32.0,
              33.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 960,
                "time_s": 32.0,
                "source_pts_s": 32.0,
                "image": "assets/frames/M7/f00000960.jpg"
              },
              {
                "frame_index": 990,
                "time_s": 33.0,
                "source_pts_s": 33.0,
                "image": "assets/frames/M7/f00000990.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 35.0,
            "range_s": [
              34.0,
              36.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1020,
                "time_s": 34.0,
                "source_pts_s": 34.0,
                "image": "assets/frames/M7/f00001020.jpg"
              },
              {
                "frame_index": 1080,
                "time_s": 36.0,
                "source_pts_s": 36.0,
                "image": "assets/frames/M7/f00001080.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a05",
          "level": "action",
          "parent_id": "M7.s03",
          "start_s": 35.0,
          "end_s": 37.867,
          "label": "Clear the chopped pieces from the board",
          "verb": "clear",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M7.board",
            "M7.pieces",
            "M7.sink"
          ],
          "state_before": "Chopped pieces rest on the tilted board.",
          "state_after": "The board is bare and returns to the counter.",
          "evidence_description": "Ordered sampled images support: clear the chopped pieces from the board. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1050,
              "time_s": 35.0,
              "source_pts_s": 35.0,
              "image": "assets/frames/M7/f00001050.jpg"
            },
            {
              "frame_index": 1080,
              "time_s": 36.0,
              "source_pts_s": 36.0,
              "image": "assets/frames/M7/f00001080.jpg"
            },
            {
              "frame_index": 1110,
              "time_s": 37.0,
              "source_pts_s": 37.0,
              "image": "assets/frames/M7/f00001110.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 35.0,
            "range_s": [
              34.0,
              36.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1020,
                "time_s": 34.0,
                "source_pts_s": 34.0,
                "image": "assets/frames/M7/f00001020.jpg"
              },
              {
                "frame_index": 1080,
                "time_s": 36.0,
                "source_pts_s": 36.0,
                "image": "assets/frames/M7/f00001080.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 37.867,
            "range_s": [
              37.0,
              38.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1110,
                "time_s": 37.0,
                "source_pts_s": 37.0,
                "image": "assets/frames/M7/f00001110.jpg"
              },
              {
                "frame_index": 1140,
                "time_s": 38.0,
                "source_pts_s": 38.0,
                "image": "assets/frames/M7/f00001140.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a06",
          "level": "action",
          "parent_id": "M7.s04",
          "start_s": 37.867,
          "end_s": 39.867,
          "label": "Stir the beaten egg with the utensil",
          "verb": "stir",
          "hands": "right",
          "hand_roles": {
            "right": "Stir the beaten egg with the utensil"
          },
          "object_ids": [
            "M7.egg",
            "M7.plate",
            "M7.utensil"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: stir the beaten egg with the utensil. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1140,
              "time_s": 38.0,
              "source_pts_s": 38.0,
              "image": "assets/frames/M7/f00001140.jpg"
            },
            {
              "frame_index": 1170,
              "time_s": 39.0,
              "source_pts_s": 39.0,
              "image": "assets/frames/M7/f00001170.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 37.867,
            "range_s": [
              37.0,
              38.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1110,
                "time_s": 37.0,
                "source_pts_s": 37.0,
                "image": "assets/frames/M7/f00001110.jpg"
              },
              {
                "frame_index": 1140,
                "time_s": 38.0,
                "source_pts_s": 38.0,
                "image": "assets/frames/M7/f00001140.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 39.867,
            "range_s": [
              39.0,
              40.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1170,
                "time_s": 39.0,
                "source_pts_s": 39.0,
                "image": "assets/frames/M7/f00001170.jpg"
              },
              {
                "frame_index": 1200,
                "time_s": 40.0,
                "source_pts_s": 40.0,
                "image": "assets/frames/M7/f00001200.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a07",
          "level": "action",
          "parent_id": "M7.s05",
          "start_s": 39.867,
          "end_s": 48.867,
          "label": "Carry the plate of beaten egg to the stove",
          "verb": "carry",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M7.plate",
            "M7.egg"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: carry the plate of beaten egg to the stove. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1230,
              "time_s": 41.0,
              "source_pts_s": 41.0,
              "image": "assets/frames/M7/f00001230.jpg"
            },
            {
              "frame_index": 1320,
              "time_s": 44.0,
              "source_pts_s": 44.0,
              "image": "assets/frames/M7/f00001320.jpg"
            },
            {
              "frame_index": 1410,
              "time_s": 47.0,
              "source_pts_s": 47.0,
              "image": "assets/frames/M7/f00001410.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 39.867,
            "range_s": [
              39.0,
              40.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1170,
                "time_s": 39.0,
                "source_pts_s": 39.0,
                "image": "assets/frames/M7/f00001170.jpg"
              },
              {
                "frame_index": 1200,
                "time_s": 40.0,
                "source_pts_s": 40.0,
                "image": "assets/frames/M7/f00001200.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 48.867,
            "range_s": [
              48.0,
              49.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1440,
                "time_s": 48.0,
                "source_pts_s": 48.0,
                "image": "assets/frames/M7/f00001440.jpg"
              },
              {
                "frame_index": 1470,
                "time_s": 49.0,
                "source_pts_s": 49.0,
                "image": "assets/frames/M7/f00001470.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a08",
          "level": "action",
          "parent_id": "M7.s06",
          "start_s": 48.867,
          "end_s": 52.0,
          "label": "Pour the beaten egg into the pan",
          "verb": "pour",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M7.plate",
            "M7.egg",
            "M7.pan"
          ],
          "state_before": "Bright orange beaten egg fills the white plate and the pan is dark and empty.",
          "state_after": "Egg covers the base of the pan and the plate is nearly empty.",
          "evidence_description": "Ordered sampled images support: pour the beaten egg into the pan. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1470,
              "time_s": 49.0,
              "source_pts_s": 49.0,
              "image": "assets/frames/M7/f00001470.jpg"
            },
            {
              "frame_index": 1500,
              "time_s": 50.0,
              "source_pts_s": 50.0,
              "image": "assets/frames/M7/f00001500.jpg"
            },
            {
              "frame_index": 1530,
              "time_s": 51.0,
              "source_pts_s": 51.0,
              "image": "assets/frames/M7/f00001530.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 48.867,
            "range_s": [
              48.0,
              49.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1440,
                "time_s": 48.0,
                "source_pts_s": 48.0,
                "image": "assets/frames/M7/f00001440.jpg"
              },
              {
                "frame_index": 1470,
                "time_s": 49.0,
                "source_pts_s": 49.0,
                "image": "assets/frames/M7/f00001470.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 52.0,
            "range_s": [
              51.0,
              53.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1530,
                "time_s": 51.0,
                "source_pts_s": 51.0,
                "image": "assets/frames/M7/f00001530.jpg"
              },
              {
                "frame_index": 1590,
                "time_s": 53.0,
                "source_pts_s": 53.0,
                "image": "assets/frames/M7/f00001590.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a09",
          "level": "action",
          "parent_id": "M7.s06",
          "start_s": 52.0,
          "end_s": 54.867,
          "label": "Lift the emptied plate clear of the pan",
          "verb": "lift",
          "hands": "left",
          "hand_roles": {
            "left": "Lift the emptied plate clear of the pan"
          },
          "object_ids": [
            "M7.plate",
            "M7.pan"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: lift the emptied plate clear of the pan. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1590,
              "time_s": 53.0,
              "source_pts_s": 53.0,
              "image": "assets/frames/M7/f00001590.jpg"
            },
            {
              "frame_index": 1620,
              "time_s": 54.0,
              "source_pts_s": 54.0,
              "image": "assets/frames/M7/f00001620.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 52.0,
            "range_s": [
              51.0,
              53.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1530,
                "time_s": 51.0,
                "source_pts_s": 51.0,
                "image": "assets/frames/M7/f00001530.jpg"
              },
              {
                "frame_index": 1590,
                "time_s": 53.0,
                "source_pts_s": 53.0,
                "image": "assets/frames/M7/f00001590.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 54.867,
            "range_s": [
              54.0,
              55.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1620,
                "time_s": 54.0,
                "source_pts_s": 54.0,
                "image": "assets/frames/M7/f00001620.jpg"
              },
              {
                "frame_index": 1650,
                "time_s": 55.0,
                "source_pts_s": 55.0,
                "image": "assets/frames/M7/f00001650.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a10",
          "level": "action",
          "parent_id": "M7.s07",
          "start_s": 54.867,
          "end_s": 57.0,
          "label": "Carry the emptied plate to the sink",
          "verb": "carry",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M7.plate",
            "M7.sink"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: carry the emptied plate to the sink. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1650,
              "time_s": 55.0,
              "source_pts_s": 55.0,
              "image": "assets/frames/M7/f00001650.jpg"
            },
            {
              "frame_index": 1680,
              "time_s": 56.0,
              "source_pts_s": 56.0,
              "image": "assets/frames/M7/f00001680.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 54.867,
            "range_s": [
              54.0,
              55.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1620,
                "time_s": 54.0,
                "source_pts_s": 54.0,
                "image": "assets/frames/M7/f00001620.jpg"
              },
              {
                "frame_index": 1650,
                "time_s": 55.0,
                "source_pts_s": 55.0,
                "image": "assets/frames/M7/f00001650.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 57.0,
            "range_s": [
              56.0,
              58.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1680,
                "time_s": 56.0,
                "source_pts_s": 56.0,
                "image": "assets/frames/M7/f00001680.jpg"
              },
              {
                "frame_index": 1740,
                "time_s": 58.0,
                "source_pts_s": 58.0,
                "image": "assets/frames/M7/f00001740.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a11",
          "level": "action",
          "parent_id": "M7.s07",
          "start_s": 57.0,
          "end_s": 63.0,
          "label": "Tend the omelet as it sets",
          "verb": "tend",
          "hands": "right",
          "hand_roles": {
            "right": "Tend the omelet as it sets"
          },
          "object_ids": [
            "M7.omelet",
            "M7.pan"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: tend the omelet as it sets. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1740,
              "time_s": 58.0,
              "source_pts_s": 58.0,
              "image": "assets/frames/M7/f00001740.jpg"
            },
            {
              "frame_index": 1800,
              "time_s": 60.0,
              "source_pts_s": 60.0,
              "image": "assets/frames/M7/f00001800.jpg"
            },
            {
              "frame_index": 1860,
              "time_s": 62.0,
              "source_pts_s": 62.0,
              "image": "assets/frames/M7/f00001860.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 57.0,
            "range_s": [
              56.0,
              58.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1680,
                "time_s": 56.0,
                "source_pts_s": 56.0,
                "image": "assets/frames/M7/f00001680.jpg"
              },
              {
                "frame_index": 1740,
                "time_s": 58.0,
                "source_pts_s": 58.0,
                "image": "assets/frames/M7/f00001740.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 63.0,
            "range_s": [
              62.0,
              64.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1860,
                "time_s": 62.0,
                "source_pts_s": 62.0,
                "image": "assets/frames/M7/f00001860.jpg"
              },
              {
                "frame_index": 1920,
                "time_s": 64.0,
                "source_pts_s": 64.0,
                "image": "assets/frames/M7/f00001920.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a12",
          "level": "action",
          "parent_id": "M7.s07",
          "start_s": 63.0,
          "end_s": 65.367,
          "label": "Turn away from the stove and return",
          "verb": "turn",
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: turn away from the stove and return. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1890,
              "time_s": 63.0,
              "source_pts_s": 63.0,
              "image": "assets/frames/M7/f00001890.jpg"
            },
            {
              "frame_index": 1920,
              "time_s": 64.0,
              "source_pts_s": 64.0,
              "image": "assets/frames/M7/f00001920.jpg"
            },
            {
              "frame_index": 1950,
              "time_s": 65.0,
              "source_pts_s": 65.0,
              "image": "assets/frames/M7/f00001950.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 63.0,
            "range_s": [
              62.0,
              64.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1860,
                "time_s": 62.0,
                "source_pts_s": 62.0,
                "image": "assets/frames/M7/f00001860.jpg"
              },
              {
                "frame_index": 1920,
                "time_s": 64.0,
                "source_pts_s": 64.0,
                "image": "assets/frames/M7/f00001920.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 65.367,
            "range_s": [
              65.0,
              66.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1950,
                "time_s": 65.0,
                "source_pts_s": 65.0,
                "image": "assets/frames/M7/f00001950.jpg"
              },
              {
                "frame_index": 1980,
                "time_s": 66.0,
                "source_pts_s": 66.0,
                "image": "assets/frames/M7/f00001980.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a13",
          "level": "action",
          "parent_id": "M7.s08",
          "start_s": 65.367,
          "end_s": 74.633333,
          "label": "Work and turn the omelet in the pan",
          "verb": "work",
          "hands": "right",
          "hand_roles": {
            "right": "Work and turn the omelet in the pan"
          },
          "object_ids": [
            "M7.omelet",
            "M7.pan"
          ],
          "state_before": "A set yellow omelet fills the pan.",
          "state_after": "The omelet is still being worked at the final frame.",
          "evidence_description": "Ordered sampled images support: work and turn the omelet in the pan. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2010,
              "time_s": 67.0,
              "source_pts_s": 67.0,
              "image": "assets/frames/M7/f00002010.jpg"
            },
            {
              "frame_index": 2100,
              "time_s": 70.0,
              "source_pts_s": 70.0,
              "image": "assets/frames/M7/f00002100.jpg"
            },
            {
              "frame_index": 2190,
              "time_s": 73.0,
              "source_pts_s": 73.0,
              "image": "assets/frames/M7/f00002190.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 65.367,
            "range_s": [
              65.0,
              66.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1950,
                "time_s": 65.0,
                "source_pts_s": 65.0,
                "image": "assets/frames/M7/f00001950.jpg"
              },
              {
                "frame_index": 1980,
                "time_s": 66.0,
                "source_pts_s": 66.0,
                "image": "assets/frames/M7/f00001980.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 74.633333,
            "range_s": [
              74.633333,
              74.633333
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 2238,
                "time_s": 74.6,
                "source_pts_s": 74.6,
                "image": "assets/frames/M7/f00002238.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M7.a14",
          "level": "action",
          "parent_id": "M7.s08",
          "start_s": 65.367,
          "end_s": 74.633333,
          "label": "Hold a small yellow object beside the pan",
          "verb": "hold",
          "hands": "left",
          "hand_roles": {
            "left": "Hold a small yellow object beside the pan"
          },
          "object_ids": [
            "M7.yellow_item"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: hold a small yellow object beside the pan. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1980,
              "time_s": 66.0,
              "source_pts_s": 66.0,
              "image": "assets/frames/M7/f00001980.jpg"
            },
            {
              "frame_index": 2070,
              "time_s": 69.0,
              "source_pts_s": 69.0,
              "image": "assets/frames/M7/f00002070.jpg"
            },
            {
              "frame_index": 2160,
              "time_s": 72.0,
              "source_pts_s": 72.0,
              "image": "assets/frames/M7/f00002160.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 65.367,
            "range_s": [
              65.0,
              66.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1950,
                "time_s": 65.0,
                "source_pts_s": 65.0,
                "image": "assets/frames/M7/f00001950.jpg"
              },
              {
                "frame_index": 1980,
                "time_s": 66.0,
                "source_pts_s": 66.0,
                "image": "assets/frames/M7/f00001980.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 74.633333,
            "range_s": [
              74.633333,
              74.633333
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 2238,
                "time_s": 74.6,
                "source_pts_s": 74.6,
                "image": "assets/frames/M7/f00002238.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "sampling": {
        "base_frames_extracted": 76,
        "base_frames_visually_reviewed": 76,
        "dense_frames_extracted": 94,
        "dense_frames_visually_reviewed": 0,
        "unique_frames_visually_reviewed": 76,
        "source_frame_count": 2239
      },
      "playback": {
        "sha256": "86b01432891113e07a6c86d3faa695f76d2bc68d95292edc199d78efe46dc375",
        "bytes": 5084768,
        "playback_width": 640,
        "playback_height": 360,
        "source_width": 640,
        "source_height": 360,
        "frame_count": 2239,
        "max_timestamp_error_s": 0.0,
        "duration_s": 74.633333,
        "audio_stream_count": 0,
        "playback_time_offset_s": 0.0,
        "kind": "original_served_from_r2",
        "source_sha256": "86b01432891113e07a6c86d3faa695f76d2bc68d95292edc199d78efe46dc375",
        "url": "https://media-demo.theroboticsstack.com/media/M7.mp4?v=86b014328911"
      }
    },
    {
      "id": "E3",
      "source": "https://media-demo.theroboticsstack.com/media/E3.mp4?v=ca49c6d1c776",
      "sha256": "8d4cea41035c88328e014b18bbc1f8cd2f30312ea823656e7a12b6bc1d010ab2",
      "duration_s": 81.3146,
      "width": 1920,
      "height": 1080,
      "fps_fraction": "9885000/406573",
      "decoded_frame_count": 1977,
      "audio_stream_count": 1,
      "source_start_pts_s": 0.4094,
      "last_frame_s": 81.233689,
      "frame_interval_min_s": 0.030388999999999555,
      "frame_interval_max_s": 0.12113399999999785,
      "nonincreasing_pts_count": 0,
      "decode_diagnostics": "",
      "view_layout": "single_egocentric_full_frame",
      "sample_interval_target_s": 1.0,
      "extraction_settings": {
        "maximum_image_width": 1280,
        "crop": null,
        "frame_selection": "nearest_actual_PTS"
      },
      "title": "Rinse and peel a potato",
      "observed_activity": "Retrieve a potato from a bag, rinse it at the sink, peel it with a small knife over the drainboard, then rinse it again.",
      "inferred_goal": "Remove the potato skin before further food preparation.",
      "assigned_instruction": null,
      "playback_time_offset_s": 0,
      "visible_outcome": "A largely peeled potato is held over the drainboard beside the accumulated peelings.",
      "task_success": "unknown",
      "unknowns": [
        "No cutting into portions or grating is visible in this clip.",
        "Peeling strokes and regrips repeat faster than the baseline sampling; no stroke count is claimed.",
        "The grater in the sink is a background object, not evidence of grating."
      ],
      "selection_rationale": "Strong: clear retrieval, rinse, peel and rinse sequence, with visible skin removal.",
      "annotation_provenance": "assistant_visual_review",
      "review_status": "pending",
      "credit": {
        "attribution": "Eidon AI \u00b7 Tracker POV \u00b7 contributor 6 \u00b7 recording 6335",
        "source_url": "https://huggingface.co/datasets/eidon-ai/tracker-pov/resolve/main/recordings/data/006/6335_video.mp4",
        "metadata_url": "https://huggingface.co/datasets/eidon-ai/tracker-pov/resolve/main/recordings/metadata.parquet",
        "license": "CC BY 4.0",
        "license_url": "https://creativecommons.org/licenses/by/4.0/",
        "changes": "Complete video re-encoded as H.264 at 8 Mbps for delivery, native resolution, audio omitted, first video frame normalised to zero; every source frame time verified. Evidence resized to at most 1280px; assistant temporal labels added."
      },
      "protocol_role": "featured",
      "objects": [
        {
          "id": "E3.potato",
          "description": "One brown-skinned potato",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E3.bag",
          "description": "Plastic bag containing potatoes",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E3.knife",
          "description": "Small white-handled knife",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E3.tap",
          "description": "Sink faucet",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E3.peels",
          "description": "Group of potato peelings on the drainboard",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        }
      ],
      "subtasks": [
        {
          "id": "E3.s01",
          "level": "subtask",
          "parent_id": null,
          "start_s": 0,
          "end_s": 5.5,
          "label": "Retrieve a potato",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E3.bag",
            "E3.potato"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: retrieve a potato. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 0,
              "time_s": 0.0,
              "source_pts_s": 0.4094,
              "image": "assets/frames/E3/f00000000.jpg"
            },
            {
              "frame_index": 72,
              "time_s": 2.987922,
              "source_pts_s": 3.397322,
              "image": "assets/frames/E3/f00000072.jpg"
            },
            {
              "frame_index": 120,
              "time_s": 5.008256,
              "source_pts_s": 5.417656,
              "image": "assets/frames/E3/f00000120.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.4094,
                "image": "assets/frames/E3/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 5.5,
            "range_s": [
              5.008256,
              6.018533
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 120,
                "time_s": 5.008256,
                "source_pts_s": 5.417656,
                "image": "assets/frames/E3/f00000120.jpg"
              },
              {
                "frame_index": 145,
                "time_s": 6.018533,
                "source_pts_s": 6.427933,
                "image": "assets/frames/E3/f00000145.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.s02",
          "level": "subtask",
          "parent_id": null,
          "start_s": 5.5,
          "end_s": 14.5,
          "label": "Rinse the unpeeled potato",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E3.potato",
            "E3.tap"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: rinse the unpeeled potato. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 120,
              "time_s": 5.008256,
              "source_pts_s": 5.417656,
              "image": "assets/frames/E3/f00000120.jpg"
            },
            {
              "frame_index": 244,
              "time_s": 10.018856,
              "source_pts_s": 10.428256,
              "image": "assets/frames/E3/f00000244.jpg"
            },
            {
              "frame_index": 342,
              "time_s": 14.019222,
              "source_pts_s": 14.428622,
              "image": "assets/frames/E3/f00000342.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 5.5,
            "range_s": [
              5.008256,
              6.018533
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 120,
                "time_s": 5.008256,
                "source_pts_s": 5.417656,
                "image": "assets/frames/E3/f00000120.jpg"
              },
              {
                "frame_index": 145,
                "time_s": 6.018533,
                "source_pts_s": 6.427933,
                "image": "assets/frames/E3/f00000145.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 14.5,
            "range_s": [
              14.019222,
              14.989089
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 342,
                "time_s": 14.019222,
                "source_pts_s": 14.428622,
                "image": "assets/frames/E3/f00000342.jpg"
              },
              {
                "frame_index": 366,
                "time_s": 14.989089,
                "source_pts_s": 15.398489,
                "image": "assets/frames/E3/f00000366.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.s03",
          "level": "subtask",
          "parent_id": null,
          "start_s": 14.5,
          "end_s": 20.1209,
          "label": "Position the potato and knife",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E3.knife",
            "E3.potato"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: position the potato and knife. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 342,
              "time_s": 14.019222,
              "source_pts_s": 14.428622,
              "image": "assets/frames/E3/f00000342.jpg"
            },
            {
              "frame_index": 416,
              "time_s": 17.009444,
              "source_pts_s": 17.418844,
              "image": "assets/frames/E3/f00000416.jpg"
            },
            {
              "frame_index": 490,
              "time_s": 19.999667,
              "source_pts_s": 20.409067,
              "image": "assets/frames/E3/f00000490.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 14.5,
            "range_s": [
              14.019222,
              14.989089
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 342,
                "time_s": 14.019222,
                "source_pts_s": 14.428622,
                "image": "assets/frames/E3/f00000342.jpg"
              },
              {
                "frame_index": 366,
                "time_s": 14.989089,
                "source_pts_s": 15.398489,
                "image": "assets/frames/E3/f00000366.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 20.1209,
            "range_s": [
              19.999667,
              20.242133
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Knife moves from visibly separate to the potato surface for the first peeling stroke.",
            "evidence_frames": [
              {
                "frame_index": 490,
                "time_s": 19.999667,
                "source_pts_s": 20.409067,
                "image": "assets/frames/E3/f00000490.jpg"
              },
              {
                "frame_index": 496,
                "time_s": 20.242133,
                "source_pts_s": 20.651533,
                "image": "assets/frames/E3/f00000496.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.s04",
          "level": "subtask",
          "parent_id": null,
          "start_s": 20.1209,
          "end_s": 69,
          "label": "Peel the potato",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E3.knife",
            "E3.peels",
            "E3.potato"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: peel the potato. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 490,
              "time_s": 19.999667,
              "source_pts_s": 20.409067,
              "image": "assets/frames/E3/f00000490.jpg"
            },
            {
              "frame_index": 1098,
              "time_s": 45.012056,
              "source_pts_s": 45.421456,
              "image": "assets/frames/E3/f00001098.jpg"
            },
            {
              "frame_index": 1686,
              "time_s": 69.014322,
              "source_pts_s": 69.423722,
              "image": "assets/frames/E3/f00001686.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 20.1209,
            "range_s": [
              19.999667,
              20.242133
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Knife moves from visibly separate to the potato surface for the first peeling stroke.",
            "evidence_frames": [
              {
                "frame_index": 490,
                "time_s": 19.999667,
                "source_pts_s": 20.409067,
                "image": "assets/frames/E3/f00000490.jpg"
              },
              {
                "frame_index": 496,
                "time_s": 20.242133,
                "source_pts_s": 20.651533,
                "image": "assets/frames/E3/f00000496.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 69,
            "range_s": [
              68.004267,
              69.014322
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1661,
                "time_s": 68.004267,
                "source_pts_s": 68.413667,
                "image": "assets/frames/E3/f00001661.jpg"
              },
              {
                "frame_index": 1686,
                "time_s": 69.014322,
                "source_pts_s": 69.423722,
                "image": "assets/frames/E3/f00001686.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.s05",
          "level": "subtask",
          "parent_id": null,
          "start_s": 69,
          "end_s": 74.5,
          "label": "Rinse the peeled potato",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E3.knife",
            "E3.potato",
            "E3.tap"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: rinse the peeled potato. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1686,
              "time_s": 69.014322,
              "source_pts_s": 69.423722,
              "image": "assets/frames/E3/f00001686.jpg"
            },
            {
              "frame_index": 1759,
              "time_s": 72.004633,
              "source_pts_s": 72.414033,
              "image": "assets/frames/E3/f00001759.jpg"
            },
            {
              "frame_index": 1825,
              "time_s": 74.9948,
              "source_pts_s": 75.4042,
              "image": "assets/frames/E3/f00001825.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 69,
            "range_s": [
              68.004267,
              69.014322
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1661,
                "time_s": 68.004267,
                "source_pts_s": 68.413667,
                "image": "assets/frames/E3/f00001661.jpg"
              },
              {
                "frame_index": 1686,
                "time_s": 69.014322,
                "source_pts_s": 69.423722,
                "image": "assets/frames/E3/f00001686.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 74.5,
            "range_s": [
              73.984544,
              74.9948
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1802,
                "time_s": 73.984544,
                "source_pts_s": 74.393944,
                "image": "assets/frames/E3/f00001802.jpg"
              },
              {
                "frame_index": 1825,
                "time_s": 74.9948,
                "source_pts_s": 75.4042,
                "image": "assets/frames/E3/f00001825.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.s06",
          "level": "subtask",
          "parent_id": null,
          "start_s": 74.5,
          "end_s": 81.3146,
          "label": "Hold the peeled potato over the drainboard",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E3.peels",
            "E3.potato"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: hold the peeled potato over the drainboard. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1825,
              "time_s": 74.9948,
              "source_pts_s": 75.4042,
              "image": "assets/frames/E3/f00001825.jpg"
            },
            {
              "frame_index": 1899,
              "time_s": 77.993,
              "source_pts_s": 78.4024,
              "image": "assets/frames/E3/f00001899.jpg"
            },
            {
              "frame_index": 1976,
              "time_s": 81.233689,
              "source_pts_s": 81.643089,
              "image": "assets/frames/E3/f00001976.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 74.5,
            "range_s": [
              73.984544,
              74.9948
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1802,
                "time_s": 73.984544,
                "source_pts_s": 74.393944,
                "image": "assets/frames/E3/f00001802.jpg"
              },
              {
                "frame_index": 1825,
                "time_s": 74.9948,
                "source_pts_s": 75.4042,
                "image": "assets/frames/E3/f00001825.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 81.3146,
            "range_s": [
              81.3146,
              81.3146
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 1976,
                "time_s": 81.233689,
                "source_pts_s": 81.643089,
                "image": "assets/frames/E3/f00001976.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "actions": [
        {
          "id": "E3.a01",
          "level": "action",
          "parent_id": "E3.s01",
          "start_s": 0,
          "end_s": 3.5,
          "label": "Open the bag and lift out a potato",
          "verb": "open",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E3.bag",
            "E3.potato"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: open the bag and lift out a potato. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 0,
              "time_s": 0.0,
              "source_pts_s": 0.4094,
              "image": "assets/frames/E3/f00000000.jpg"
            },
            {
              "frame_index": 48,
              "time_s": 2.018156,
              "source_pts_s": 2.427556,
              "image": "assets/frames/E3/f00000048.jpg"
            },
            {
              "frame_index": 72,
              "time_s": 2.987922,
              "source_pts_s": 3.397322,
              "image": "assets/frames/E3/f00000072.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.4094,
                "image": "assets/frames/E3/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 3.5,
            "range_s": [
              2.987922,
              4.038511
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 72,
                "time_s": 2.987922,
                "source_pts_s": 3.397322,
                "image": "assets/frames/E3/f00000072.jpg"
              },
              {
                "frame_index": 97,
                "time_s": 4.038511,
                "source_pts_s": 4.447911,
                "image": "assets/frames/E3/f00000097.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.a02",
          "level": "action",
          "parent_id": "E3.s01",
          "start_s": 3.5,
          "end_s": 5.5,
          "label": "Carry the potato toward the sink",
          "verb": "carry",
          "hands": "right",
          "hand_roles": {
            "right": "Carry the potato toward the sink"
          },
          "object_ids": [
            "E3.potato"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: carry the potato toward the sink. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 97,
              "time_s": 4.038511,
              "source_pts_s": 4.447911,
              "image": "assets/frames/E3/f00000097.jpg"
            },
            {
              "frame_index": 120,
              "time_s": 5.008256,
              "source_pts_s": 5.417656,
              "image": "assets/frames/E3/f00000120.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 3.5,
            "range_s": [
              2.987922,
              4.038511
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 72,
                "time_s": 2.987922,
                "source_pts_s": 3.397322,
                "image": "assets/frames/E3/f00000072.jpg"
              },
              {
                "frame_index": 97,
                "time_s": 4.038511,
                "source_pts_s": 4.447911,
                "image": "assets/frames/E3/f00000097.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 5.5,
            "range_s": [
              5.008256,
              6.018533
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 120,
                "time_s": 5.008256,
                "source_pts_s": 5.417656,
                "image": "assets/frames/E3/f00000120.jpg"
              },
              {
                "frame_index": 145,
                "time_s": 6.018533,
                "source_pts_s": 6.427933,
                "image": "assets/frames/E3/f00000145.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.a03",
          "level": "action",
          "parent_id": "E3.s02",
          "start_s": 5.5,
          "end_s": 14.5,
          "label": "Hold and turn the potato beneath running water",
          "verb": "hold",
          "hands": "left",
          "hand_roles": {
            "left": "Hold and turn the potato beneath running water"
          },
          "object_ids": [
            "E3.potato",
            "E3.tap"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: hold and turn the potato beneath running water. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 145,
              "time_s": 6.018533,
              "source_pts_s": 6.427933,
              "image": "assets/frames/E3/f00000145.jpg"
            },
            {
              "frame_index": 244,
              "time_s": 10.018856,
              "source_pts_s": 10.428256,
              "image": "assets/frames/E3/f00000244.jpg"
            },
            {
              "frame_index": 317,
              "time_s": 13.009044,
              "source_pts_s": 13.418444,
              "image": "assets/frames/E3/f00000317.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 5.5,
            "range_s": [
              5.008256,
              6.018533
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 120,
                "time_s": 5.008256,
                "source_pts_s": 5.417656,
                "image": "assets/frames/E3/f00000120.jpg"
              },
              {
                "frame_index": 145,
                "time_s": 6.018533,
                "source_pts_s": 6.427933,
                "image": "assets/frames/E3/f00000145.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 14.5,
            "range_s": [
              14.019222,
              14.989089
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 342,
                "time_s": 14.019222,
                "source_pts_s": 14.428622,
                "image": "assets/frames/E3/f00000342.jpg"
              },
              {
                "frame_index": 366,
                "time_s": 14.989089,
                "source_pts_s": 15.398489,
                "image": "assets/frames/E3/f00000366.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.a04",
          "level": "action",
          "parent_id": "E3.s03",
          "start_s": 14.5,
          "end_s": 18.5,
          "label": "Set the potato on the drainboard and reach into the sink",
          "verb": "set",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E3.potato",
            "E3.knife"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: set the potato on the drainboard and reach into the sink. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 342,
              "time_s": 14.019222,
              "source_pts_s": 14.428622,
              "image": "assets/frames/E3/f00000342.jpg"
            },
            {
              "frame_index": 391,
              "time_s": 15.999267,
              "source_pts_s": 16.408667,
              "image": "assets/frames/E3/f00000391.jpg"
            },
            {
              "frame_index": 441,
              "time_s": 18.019622,
              "source_pts_s": 18.429022,
              "image": "assets/frames/E3/f00000441.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 14.5,
            "range_s": [
              14.019222,
              14.989089
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 342,
                "time_s": 14.019222,
                "source_pts_s": 14.428622,
                "image": "assets/frames/E3/f00000342.jpg"
              },
              {
                "frame_index": 366,
                "time_s": 14.989089,
                "source_pts_s": 15.398489,
                "image": "assets/frames/E3/f00000366.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 18.5,
            "range_s": [
              18.019622,
              18.989489
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 441,
                "time_s": 18.019622,
                "source_pts_s": 18.429022,
                "image": "assets/frames/E3/f00000441.jpg"
              },
              {
                "frame_index": 465,
                "time_s": 18.989489,
                "source_pts_s": 19.398889,
                "image": "assets/frames/E3/f00000465.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.a05",
          "level": "action",
          "parent_id": "E3.s03",
          "start_s": 18.5,
          "end_s": 20.1209,
          "label": "Take the knife and pick up the potato",
          "verb": "take",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E3.knife",
            "E3.potato"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: take the knife and pick up the potato. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 465,
              "time_s": 18.989489,
              "source_pts_s": 19.398889,
              "image": "assets/frames/E3/f00000465.jpg"
            },
            {
              "frame_index": 490,
              "time_s": 19.999667,
              "source_pts_s": 20.409067,
              "image": "assets/frames/E3/f00000490.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 18.5,
            "range_s": [
              18.019622,
              18.989489
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 441,
                "time_s": 18.019622,
                "source_pts_s": 18.429022,
                "image": "assets/frames/E3/f00000441.jpg"
              },
              {
                "frame_index": 465,
                "time_s": 18.989489,
                "source_pts_s": 19.398889,
                "image": "assets/frames/E3/f00000465.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 20.1209,
            "range_s": [
              19.999667,
              20.242133
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Knife moves from visibly separate to the potato surface for the first peeling stroke.",
            "evidence_frames": [
              {
                "frame_index": 490,
                "time_s": 19.999667,
                "source_pts_s": 20.409067,
                "image": "assets/frames/E3/f00000490.jpg"
              },
              {
                "frame_index": 496,
                "time_s": 20.242133,
                "source_pts_s": 20.651533,
                "image": "assets/frames/E3/f00000496.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.a06",
          "level": "action",
          "parent_id": "E3.s04",
          "start_s": 20.1209,
          "end_s": 69,
          "label": "Scrape strips of skin from the potato with the knife",
          "verb": "scrape",
          "hands": "right",
          "hand_roles": {
            "right": "Scrape strips of skin from the potato with the knife"
          },
          "object_ids": [
            "E3.potato",
            "E3.knife",
            "E3.peels"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: scrape strips of skin from the potato with the knife. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 540,
              "time_s": 22.020011,
              "source_pts_s": 22.429411,
              "image": "assets/frames/E3/f00000540.jpg"
            },
            {
              "frame_index": 1050,
              "time_s": 42.991633,
              "source_pts_s": 43.401033,
              "image": "assets/frames/E3/f00001050.jpg"
            },
            {
              "frame_index": 1661,
              "time_s": 68.004267,
              "source_pts_s": 68.413667,
              "image": "assets/frames/E3/f00001661.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 20.1209,
            "range_s": [
              19.999667,
              20.242133
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Knife moves from visibly separate to the potato surface for the first peeling stroke.",
            "evidence_frames": [
              {
                "frame_index": 490,
                "time_s": 19.999667,
                "source_pts_s": 20.409067,
                "image": "assets/frames/E3/f00000490.jpg"
              },
              {
                "frame_index": 496,
                "time_s": 20.242133,
                "source_pts_s": 20.651533,
                "image": "assets/frames/E3/f00000496.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 69,
            "range_s": [
              68.004267,
              69.014322
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1661,
                "time_s": 68.004267,
                "source_pts_s": 68.413667,
                "image": "assets/frames/E3/f00001661.jpg"
              },
              {
                "frame_index": 1686,
                "time_s": 69.014322,
                "source_pts_s": 69.423722,
                "image": "assets/frames/E3/f00001686.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.a07",
          "level": "action",
          "parent_id": "E3.s04",
          "start_s": 20.1209,
          "end_s": 69,
          "label": "Hold and rotate the potato between peeling strokes",
          "verb": "hold",
          "hands": "left",
          "hand_roles": {
            "left": "Hold and rotate the potato between peeling strokes"
          },
          "object_ids": [
            "E3.potato"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: hold and rotate the potato between peeling strokes. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 614,
              "time_s": 25.010233,
              "source_pts_s": 25.419633,
              "image": "assets/frames/E3/f00000614.jpg"
            },
            {
              "frame_index": 1172,
              "time_s": 48.002156,
              "source_pts_s": 48.411556,
              "image": "assets/frames/E3/f00001172.jpg"
            },
            {
              "frame_index": 1613,
              "time_s": 65.9839,
              "source_pts_s": 66.3933,
              "image": "assets/frames/E3/f00001613.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 20.1209,
            "range_s": [
              19.999667,
              20.242133
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Knife moves from visibly separate to the potato surface for the first peeling stroke.",
            "evidence_frames": [
              {
                "frame_index": 490,
                "time_s": 19.999667,
                "source_pts_s": 20.409067,
                "image": "assets/frames/E3/f00000490.jpg"
              },
              {
                "frame_index": 496,
                "time_s": 20.242133,
                "source_pts_s": 20.651533,
                "image": "assets/frames/E3/f00000496.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 69,
            "range_s": [
              68.004267,
              69.014322
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1661,
                "time_s": 68.004267,
                "source_pts_s": 68.413667,
                "image": "assets/frames/E3/f00001661.jpg"
              },
              {
                "frame_index": 1686,
                "time_s": 69.014322,
                "source_pts_s": 69.423722,
                "image": "assets/frames/E3/f00001686.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.a08",
          "level": "action",
          "parent_id": "E3.s05",
          "start_s": 69,
          "end_s": 74.5,
          "label": "Move the peeled potato and knife under the faucet",
          "verb": "move",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E3.potato",
            "E3.knife",
            "E3.tap"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: move the peeled potato and knife under the faucet. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1710,
              "time_s": 69.984144,
              "source_pts_s": 70.393544,
              "image": "assets/frames/E3/f00001710.jpg"
            },
            {
              "frame_index": 1759,
              "time_s": 72.004633,
              "source_pts_s": 72.414033,
              "image": "assets/frames/E3/f00001759.jpg"
            },
            {
              "frame_index": 1802,
              "time_s": 73.984544,
              "source_pts_s": 74.393944,
              "image": "assets/frames/E3/f00001802.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 69,
            "range_s": [
              68.004267,
              69.014322
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1661,
                "time_s": 68.004267,
                "source_pts_s": 68.413667,
                "image": "assets/frames/E3/f00001661.jpg"
              },
              {
                "frame_index": 1686,
                "time_s": 69.014322,
                "source_pts_s": 69.423722,
                "image": "assets/frames/E3/f00001686.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 74.5,
            "range_s": [
              73.984544,
              74.9948
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1802,
                "time_s": 73.984544,
                "source_pts_s": 74.393944,
                "image": "assets/frames/E3/f00001802.jpg"
              },
              {
                "frame_index": 1825,
                "time_s": 74.9948,
                "source_pts_s": 75.4042,
                "image": "assets/frames/E3/f00001825.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E3.a09",
          "level": "action",
          "parent_id": "E3.s06",
          "start_s": 74.5,
          "end_s": 81.3146,
          "label": "Hold the peeled potato above the discarded skins",
          "verb": "hold",
          "hands": "left",
          "hand_roles": {
            "left": "Hold the peeled potato above the discarded skins"
          },
          "object_ids": [
            "E3.potato",
            "E3.peels"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: hold the peeled potato above the discarded skins. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1825,
              "time_s": 74.9948,
              "source_pts_s": 75.4042,
              "image": "assets/frames/E3/f00001825.jpg"
            },
            {
              "frame_index": 1899,
              "time_s": 77.993,
              "source_pts_s": 78.4024,
              "image": "assets/frames/E3/f00001899.jpg"
            },
            {
              "frame_index": 1972,
              "time_s": 80.991067,
              "source_pts_s": 81.400467,
              "image": "assets/frames/E3/f00001972.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 74.5,
            "range_s": [
              73.984544,
              74.9948
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1802,
                "time_s": 73.984544,
                "source_pts_s": 74.393944,
                "image": "assets/frames/E3/f00001802.jpg"
              },
              {
                "frame_index": 1825,
                "time_s": 74.9948,
                "source_pts_s": 75.4042,
                "image": "assets/frames/E3/f00001825.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 81.3146,
            "range_s": [
              81.3146,
              81.3146
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 1976,
                "time_s": 81.233689,
                "source_pts_s": 81.643089,
                "image": "assets/frames/E3/f00001976.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "sampling": {
        "base_frames_extracted": 83,
        "base_frames_visually_reviewed": 83,
        "dense_frames_extracted": 26,
        "dense_frames_visually_reviewed": 26,
        "unique_frames_visually_reviewed": 104,
        "source_frame_count": 1977
      },
      "playback": {
        "sha256": "ca49c6d1c776093fd5ea0fa4d41f7881ffa46385d27261f8329a8a19f301a05f",
        "bytes": 81251988,
        "playback_width": 1920,
        "playback_height": 1080,
        "source_width": 1920,
        "source_height": 1080,
        "frame_count": 1977,
        "max_timestamp_error_s": 1.4210854715202004e-14,
        "duration_s": 81.253589,
        "audio_stream_count": 0,
        "playback_time_offset_s": 0,
        "kind": "delivery_encode_served_from_r2",
        "source_sha256": "8d4cea41035c88328e014b18bbc1f8cd2f30312ea823656e7a12b6bc1d010ab2",
        "encoding": "h264-native-2pass-8mbps-relative-pts-v1",
        "url": "https://media-demo.theroboticsstack.com/media/E3.mp4?v=ca49c6d1c776"
      }
    },
    {
      "id": "E9",
      "source": "https://media-demo.theroboticsstack.com/media/E9.mp4?v=e37fa3b46f16",
      "sha256": "b84574c94d9341d40014cafdbbc465fffeb4e92a9b7b541922cc4b95c82d933c",
      "duration_s": 209.6853,
      "width": 1280,
      "height": 720,
      "fps_fraction": "20890000/698951",
      "decoded_frame_count": 6267,
      "audio_stream_count": 1,
      "source_start_pts_s": 0.0249,
      "last_frame_s": 209.65196699999998,
      "frame_interval_min_s": 0.03323299999999563,
      "frame_interval_max_s": 0.2333109999999996,
      "nonincreasing_pts_count": 0,
      "decode_diagnostics": "",
      "view_layout": "single_egocentric_full_frame",
      "sample_interval_target_s": 1.0,
      "extraction_settings": {
        "maximum_image_width": 1280,
        "crop": null,
        "frame_selection": "nearest_actual_PTS"
      },
      "title": "Mix food and cut cheese",
      "observed_activity": "Stir a coarse food mixture in a handled glass bowl, set it aside, unwrap cheese, cut it into strips, group the strips, then open and cut a second block.",
      "inferred_goal": "Prepare a mixed filling and portions of cheese for a later cooking step.",
      "assigned_instruction": null,
      "playback_time_offset_s": 0.024,
      "visible_outcome": "The bowl remains at the back of the counter and cut cheese strips accumulate on the board; cutting is still underway at the end.",
      "task_success": "unknown",
      "unknowns": [
        "The exact ingredients and recipe are not established by appearance.",
        "A visible package supports cheese identification; no nutritional or food-safety inference is made.",
        "Cutting cycles are grouped, not exhaustively counted.",
        "The clip starts mid-preparation and ends during the second cutting sequence.",
        "A separate experiment disagrees about the opening tool. Reinspection supports a white utensil from frame 0; see RECONCILIATION.md."
      ],
      "selection_rationale": "Strong reserve: many clear phases, but longer and more repetitive than the selected egg preparation example.",
      "annotation_provenance": "assistant_visual_review",
      "review_status": "pending",
      "credit": {
        "attribution": "Eidon AI \u00b7 Tracker POV \u00b7 contributor 3 \u00b7 recording 6003",
        "source_url": "https://huggingface.co/datasets/eidon-ai/tracker-pov/resolve/main/recordings/data/006/6003_video.mp4",
        "metadata_url": "https://huggingface.co/datasets/eidon-ai/tracker-pov/resolve/main/recordings/metadata.parquet",
        "license": "CC BY 4.0",
        "license_url": "https://creativecommons.org/licenses/by/4.0/",
        "changes": "Complete source published from object storage with the audio track removed; the video bitstream is copied, not re-encoded. Evidence resized to at most 1280px; assistant temporal labels added."
      },
      "protocol_role": "featured",
      "objects": [
        {
          "id": "E9.jug",
          "description": "Glass mixing bowl with gold-toned handles (legacy object ID retained)",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E9.mixture",
          "description": "Coarse mixed food, ingredients not fully identified",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E9.paddle",
          "description": "White broad mixing spoon or paddle",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E9.fork",
          "description": "Metal fork",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E9.cheese1",
          "description": "First pale cheese block and resulting pieces",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E9.cheese2",
          "description": "Second pale cheese block and resulting pieces",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E9.wrappers",
          "description": "Cheese carton and foil or paper wrappers",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E9.knife",
          "description": "Kitchen knife",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E9.board",
          "description": "Printed cutting board",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        }
      ],
      "subtasks": [
        {
          "id": "E9.s01",
          "level": "subtask",
          "parent_id": null,
          "start_s": 0,
          "end_s": 52.5,
          "label": "Mix the food in the bowl",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E9.fork",
            "E9.jug",
            "E9.mixture",
            "E9.paddle"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: mix the food in the bowl. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 0,
              "time_s": 0.0,
              "source_pts_s": 0.0249,
              "image": "assets/frames/E9/f00000000.jpg"
            },
            {
              "frame_index": 781,
              "time_s": 26.264922,
              "source_pts_s": 26.289822,
              "image": "assets/frames/E9/f00000781.jpg"
            },
            {
              "frame_index": 1580,
              "time_s": 52.996322,
              "source_pts_s": 53.021222,
              "image": "assets/frames/E9/f00001580.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.0249,
                "image": "assets/frames/E9/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 52.5,
            "range_s": [
              51.996422,
              52.996322
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1550,
                "time_s": 51.996422,
                "source_pts_s": 52.021322,
                "image": "assets/frames/E9/f00001550.jpg"
              },
              {
                "frame_index": 1580,
                "time_s": 52.996322,
                "source_pts_s": 53.021222,
                "image": "assets/frames/E9/f00001580.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.s02",
          "level": "subtask",
          "parent_id": null,
          "start_s": 52.5,
          "end_s": 73.5,
          "label": "Clear space and unwrap the first cheese block",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E9.board",
            "E9.cheese1",
            "E9.jug",
            "E9.wrappers"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: clear space and unwrap the first cheese block. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1580,
              "time_s": 52.996322,
              "source_pts_s": 53.021222,
              "image": "assets/frames/E9/f00001580.jpg"
            },
            {
              "frame_index": 1880,
              "time_s": 62.995622,
              "source_pts_s": 63.020522,
              "image": "assets/frames/E9/f00001880.jpg"
            },
            {
              "frame_index": 2210,
              "time_s": 73.994822,
              "source_pts_s": 74.019722,
              "image": "assets/frames/E9/f00002210.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 52.5,
            "range_s": [
              51.996422,
              52.996322
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1550,
                "time_s": 51.996422,
                "source_pts_s": 52.021322,
                "image": "assets/frames/E9/f00001550.jpg"
              },
              {
                "frame_index": 1580,
                "time_s": 52.996322,
                "source_pts_s": 53.021222,
                "image": "assets/frames/E9/f00001580.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 73.5,
            "range_s": [
              72.994922,
              73.994822
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2180,
                "time_s": 72.994922,
                "source_pts_s": 73.019822,
                "image": "assets/frames/E9/f00002180.jpg"
              },
              {
                "frame_index": 2210,
                "time_s": 73.994822,
                "source_pts_s": 74.019722,
                "image": "assets/frames/E9/f00002210.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.s03",
          "level": "subtask",
          "parent_id": null,
          "start_s": 73.5,
          "end_s": 130.5,
          "label": "Cut the first block into strips",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E9.board",
            "E9.cheese1",
            "E9.knife"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: cut the first block into strips. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2210,
              "time_s": 73.994822,
              "source_pts_s": 74.019722,
              "image": "assets/frames/E9/f00002210.jpg"
            },
            {
              "frame_index": 3048,
              "time_s": 101.992867,
              "source_pts_s": 102.017767,
              "image": "assets/frames/E9/f00003048.jpg"
            },
            {
              "frame_index": 3917,
              "time_s": 130.990822,
              "source_pts_s": 131.015722,
              "image": "assets/frames/E9/f00003917.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 73.5,
            "range_s": [
              72.994922,
              73.994822
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2180,
                "time_s": 72.994922,
                "source_pts_s": 73.019822,
                "image": "assets/frames/E9/f00002180.jpg"
              },
              {
                "frame_index": 2210,
                "time_s": 73.994822,
                "source_pts_s": 74.019722,
                "image": "assets/frames/E9/f00002210.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 130.5,
            "range_s": [
              129.990922,
              130.990822
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 3887,
                "time_s": 129.990922,
                "source_pts_s": 130.015822,
                "image": "assets/frames/E9/f00003887.jpg"
              },
              {
                "frame_index": 3917,
                "time_s": 130.990822,
                "source_pts_s": 131.015722,
                "image": "assets/frames/E9/f00003917.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.s04",
          "level": "subtask",
          "parent_id": null,
          "start_s": 130.5,
          "end_s": 136.5,
          "label": "Gather the cut strips",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E9.board",
            "E9.cheese1"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: gather the cut strips. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 3917,
              "time_s": 130.990822,
              "source_pts_s": 131.015722,
              "image": "assets/frames/E9/f00003917.jpg"
            },
            {
              "frame_index": 4006,
              "time_s": 133.990611,
              "source_pts_s": 134.015511,
              "image": "assets/frames/E9/f00004006.jpg"
            },
            {
              "frame_index": 4096,
              "time_s": 136.990411,
              "source_pts_s": 137.015311,
              "image": "assets/frames/E9/f00004096.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 130.5,
            "range_s": [
              129.990922,
              130.990822
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 3887,
                "time_s": 129.990922,
                "source_pts_s": 130.015822,
                "image": "assets/frames/E9/f00003887.jpg"
              },
              {
                "frame_index": 3917,
                "time_s": 130.990822,
                "source_pts_s": 131.015722,
                "image": "assets/frames/E9/f00003917.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 136.5,
            "range_s": [
              135.990411,
              136.990411
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 4066,
                "time_s": 135.990411,
                "source_pts_s": 136.015311,
                "image": "assets/frames/E9/f00004066.jpg"
              },
              {
                "frame_index": 4096,
                "time_s": 136.990411,
                "source_pts_s": 137.015311,
                "image": "assets/frames/E9/f00004096.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.s05",
          "level": "subtask",
          "parent_id": null,
          "start_s": 136.5,
          "end_s": 174.5,
          "label": "Open the second cheese package",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E9.cheese2",
            "E9.wrappers"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: open the second cheese package. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 4096,
              "time_s": 136.990411,
              "source_pts_s": 137.015311,
              "image": "assets/frames/E9/f00004096.jpg"
            },
            {
              "frame_index": 4666,
              "time_s": 155.989011,
              "source_pts_s": 156.013911,
              "image": "assets/frames/E9/f00004666.jpg"
            },
            {
              "frame_index": 5235,
              "time_s": 174.987667,
              "source_pts_s": 175.012567,
              "image": "assets/frames/E9/f00005235.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 136.5,
            "range_s": [
              135.990411,
              136.990411
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 4066,
                "time_s": 135.990411,
                "source_pts_s": 136.015311,
                "image": "assets/frames/E9/f00004066.jpg"
              },
              {
                "frame_index": 4096,
                "time_s": 136.990411,
                "source_pts_s": 137.015311,
                "image": "assets/frames/E9/f00004096.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 174.5,
            "range_s": [
              173.987811,
              174.987667
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5206,
                "time_s": 173.987811,
                "source_pts_s": 174.012711,
                "image": "assets/frames/E9/f00005206.jpg"
              },
              {
                "frame_index": 5235,
                "time_s": 174.987667,
                "source_pts_s": 175.012567,
                "image": "assets/frames/E9/f00005235.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.s06",
          "level": "subtask",
          "parent_id": null,
          "start_s": 174.5,
          "end_s": 189.5,
          "label": "Separate a portion and manage the wrapper",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E9.board",
            "E9.cheese2",
            "E9.knife",
            "E9.wrappers"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: separate a portion and manage the wrapper. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 5235,
              "time_s": 174.987667,
              "source_pts_s": 175.012567,
              "image": "assets/frames/E9/f00005235.jpg"
            },
            {
              "frame_index": 5445,
              "time_s": 181.987267,
              "source_pts_s": 182.012167,
              "image": "assets/frames/E9/f00005445.jpg"
            },
            {
              "frame_index": 5683,
              "time_s": 189.986633,
              "source_pts_s": 190.011533,
              "image": "assets/frames/E9/f00005683.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 174.5,
            "range_s": [
              173.987811,
              174.987667
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5206,
                "time_s": 173.987811,
                "source_pts_s": 174.012711,
                "image": "assets/frames/E9/f00005206.jpg"
              },
              {
                "frame_index": 5235,
                "time_s": 174.987667,
                "source_pts_s": 175.012567,
                "image": "assets/frames/E9/f00005235.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 189.5,
            "range_s": [
              188.986711,
              189.986633
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5654,
                "time_s": 188.986711,
                "source_pts_s": 189.011611,
                "image": "assets/frames/E9/f00005654.jpg"
              },
              {
                "frame_index": 5683,
                "time_s": 189.986633,
                "source_pts_s": 190.011533,
                "image": "assets/frames/E9/f00005683.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.s07",
          "level": "subtask",
          "parent_id": null,
          "start_s": 189.5,
          "end_s": 209.6853,
          "label": "Cut more cheese strips",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E9.cheese2",
            "E9.knife"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: cut more cheese strips. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 5683,
              "time_s": 189.986633,
              "source_pts_s": 190.011533,
              "image": "assets/frames/E9/f00005683.jpg"
            },
            {
              "frame_index": 5979,
              "time_s": 199.985911,
              "source_pts_s": 200.010811,
              "image": "assets/frames/E9/f00005979.jpg"
            },
            {
              "frame_index": 6266,
              "time_s": 209.651967,
              "source_pts_s": 209.676867,
              "image": "assets/frames/E9/f00006266.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 189.5,
            "range_s": [
              188.986711,
              189.986633
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5654,
                "time_s": 188.986711,
                "source_pts_s": 189.011611,
                "image": "assets/frames/E9/f00005654.jpg"
              },
              {
                "frame_index": 5683,
                "time_s": 189.986633,
                "source_pts_s": 190.011533,
                "image": "assets/frames/E9/f00005683.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 209.6853,
            "range_s": [
              209.6853,
              209.6853
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 6266,
                "time_s": 209.651967,
                "source_pts_s": 209.676867,
                "image": "assets/frames/E9/f00006266.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "actions": [
        {
          "id": "E9.a01",
          "level": "action",
          "parent_id": "E9.s01",
          "start_s": 0,
          "end_s": 23.5,
          "label": "Stir and fold the mixture with the paddle",
          "verb": "stir",
          "hands": "right",
          "hand_roles": {
            "right": "Stir and fold the mixture with the paddle"
          },
          "object_ids": [
            "E9.jug",
            "E9.mixture",
            "E9.paddle"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: stir and fold the mixture with the paddle. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 30,
              "time_s": 1.0,
              "source_pts_s": 1.0249,
              "image": "assets/frames/E9/f00000030.jpg"
            },
            {
              "frame_index": 324,
              "time_s": 10.999278,
              "source_pts_s": 11.024178,
              "image": "assets/frames/E9/f00000324.jpg"
            },
            {
              "frame_index": 654,
              "time_s": 21.998478,
              "source_pts_s": 22.023378,
              "image": "assets/frames/E9/f00000654.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.0249,
                "image": "assets/frames/E9/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 23.5,
            "range_s": [
              22.998478,
              23.998378
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 684,
                "time_s": 22.998478,
                "source_pts_s": 23.023378,
                "image": "assets/frames/E9/f00000684.jpg"
              },
              {
                "frame_index": 714,
                "time_s": 23.998378,
                "source_pts_s": 24.023278,
                "image": "assets/frames/E9/f00000714.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a02",
          "level": "action",
          "parent_id": "E9.s01",
          "start_s": 0,
          "end_s": 23.5,
          "label": "Hold the mixing bowl steady",
          "verb": "hold",
          "hands": "left",
          "hand_roles": {
            "left": "Hold the mixing bowl steady"
          },
          "object_ids": [
            "E9.jug"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: hold the mixing bowl steady. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 60,
              "time_s": 1.9999,
              "source_pts_s": 2.0248,
              "image": "assets/frames/E9/f00000060.jpg"
            },
            {
              "frame_index": 354,
              "time_s": 11.999178,
              "source_pts_s": 12.024078,
              "image": "assets/frames/E9/f00000354.jpg"
            },
            {
              "frame_index": 624,
              "time_s": 20.998578,
              "source_pts_s": 21.023478,
              "image": "assets/frames/E9/f00000624.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.0249,
                "image": "assets/frames/E9/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 23.5,
            "range_s": [
              22.998478,
              23.998378
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 684,
                "time_s": 22.998478,
                "source_pts_s": 23.023378,
                "image": "assets/frames/E9/f00000684.jpg"
              },
              {
                "frame_index": 714,
                "time_s": 23.998378,
                "source_pts_s": 24.023278,
                "image": "assets/frames/E9/f00000714.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a03",
          "level": "action",
          "parent_id": "E9.s01",
          "start_s": 24.8816,
          "end_s": 26.631539,
          "label": "Work a fork against the mixture near the paddle",
          "verb": "work",
          "hands": "right",
          "hand_roles": {
            "right": "Work a fork against the mixture near the paddle"
          },
          "object_ids": [
            "E9.fork",
            "E9.mixture",
            "E9.paddle"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: work a fork against the mixture near the paddle. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 743,
              "time_s": 24.998256,
              "source_pts_s": 25.023156,
              "image": "assets/frames/E9/f00000743.jpg"
            },
            {
              "frame_index": 773,
              "time_s": 25.998256,
              "source_pts_s": 26.023156,
              "image": "assets/frames/E9/f00000773.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 24.8816,
            "range_s": [
              24.764944,
              24.998256
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Fork moves into contact with the mixture beside the resting white utensil.",
            "evidence_frames": [
              {
                "frame_index": 737,
                "time_s": 24.764944,
                "source_pts_s": 24.789844,
                "image": "assets/frames/E9/f00000737.jpg"
              },
              {
                "frame_index": 743,
                "time_s": 24.998256,
                "source_pts_s": 25.023156,
                "image": "assets/frames/E9/f00000743.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 26.631539,
            "range_s": [
              26.498256,
              26.764822
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Fork withdraws from above the bowl; the white utensil remains in place.",
            "evidence_frames": [
              {
                "frame_index": 788,
                "time_s": 26.498256,
                "source_pts_s": 26.523156,
                "image": "assets/frames/E9/f00000788.jpg"
              },
              {
                "frame_index": 796,
                "time_s": 26.764822,
                "source_pts_s": 26.789722,
                "image": "assets/frames/E9/f00000796.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a04",
          "level": "action",
          "parent_id": "E9.s01",
          "start_s": 28.5,
          "end_s": 45.5,
          "label": "Resume stirring and folding the mixture",
          "verb": "resume",
          "hands": "right",
          "hand_roles": {
            "right": "Resume stirring and folding the mixture"
          },
          "object_ids": [
            "E9.jug",
            "E9.mixture",
            "E9.paddle"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: resume stirring and folding the mixture. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 863,
              "time_s": 28.998056,
              "source_pts_s": 29.022956,
              "image": "assets/frames/E9/f00000863.jpg"
            },
            {
              "frame_index": 1073,
              "time_s": 35.997556,
              "source_pts_s": 36.022456,
              "image": "assets/frames/E9/f00001073.jpg"
            },
            {
              "frame_index": 1312,
              "time_s": 43.996944,
              "source_pts_s": 44.021844,
              "image": "assets/frames/E9/f00001312.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 28.5,
            "range_s": [
              27.998056,
              28.998056
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 833,
                "time_s": 27.998056,
                "source_pts_s": 28.022956,
                "image": "assets/frames/E9/f00000833.jpg"
              },
              {
                "frame_index": 863,
                "time_s": 28.998056,
                "source_pts_s": 29.022956,
                "image": "assets/frames/E9/f00000863.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 45.5,
            "range_s": [
              44.996844,
              45.996844
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1342,
                "time_s": 44.996844,
                "source_pts_s": 45.021744,
                "image": "assets/frames/E9/f00001342.jpg"
              },
              {
                "frame_index": 1372,
                "time_s": 45.996844,
                "source_pts_s": 46.021744,
                "image": "assets/frames/E9/f00001372.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a05",
          "level": "action",
          "parent_id": "E9.s01",
          "start_s": 28.5,
          "end_s": 45.5,
          "label": "Hold and reposition the bowl while mixing",
          "verb": "hold",
          "hands": "left",
          "hand_roles": {
            "left": "Hold and reposition the bowl while mixing"
          },
          "object_ids": [
            "E9.jug"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: hold and reposition the bowl while mixing. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 923,
              "time_s": 30.997856,
              "source_pts_s": 31.022756,
              "image": "assets/frames/E9/f00000923.jpg"
            },
            {
              "frame_index": 1163,
              "time_s": 38.997356,
              "source_pts_s": 39.022256,
              "image": "assets/frames/E9/f00001163.jpg"
            },
            {
              "frame_index": 1282,
              "time_s": 42.997044,
              "source_pts_s": 43.021944,
              "image": "assets/frames/E9/f00001282.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 28.5,
            "range_s": [
              27.998056,
              28.998056
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 833,
                "time_s": 27.998056,
                "source_pts_s": 28.022956,
                "image": "assets/frames/E9/f00000833.jpg"
              },
              {
                "frame_index": 863,
                "time_s": 28.998056,
                "source_pts_s": 29.022956,
                "image": "assets/frames/E9/f00000863.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 45.5,
            "range_s": [
              44.996844,
              45.996844
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1342,
                "time_s": 44.996844,
                "source_pts_s": 45.021744,
                "image": "assets/frames/E9/f00001342.jpg"
              },
              {
                "frame_index": 1372,
                "time_s": 45.996844,
                "source_pts_s": 46.021744,
                "image": "assets/frames/E9/f00001372.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a06",
          "level": "action",
          "parent_id": "E9.s01",
          "start_s": 45.5,
          "end_s": 52.5,
          "label": "Scrape mixture from the paddle with the fork",
          "verb": "scrape",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.paddle",
            "E9.fork",
            "E9.mixture"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: scrape mixture from the paddle with the fork. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1372,
              "time_s": 45.996844,
              "source_pts_s": 46.021744,
              "image": "assets/frames/E9/f00001372.jpg"
            },
            {
              "frame_index": 1460,
              "time_s": 48.996622,
              "source_pts_s": 49.021522,
              "image": "assets/frames/E9/f00001460.jpg"
            },
            {
              "frame_index": 1520,
              "time_s": 50.996422,
              "source_pts_s": 51.021322,
              "image": "assets/frames/E9/f00001520.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 45.5,
            "range_s": [
              44.996844,
              45.996844
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1342,
                "time_s": 44.996844,
                "source_pts_s": 45.021744,
                "image": "assets/frames/E9/f00001342.jpg"
              },
              {
                "frame_index": 1372,
                "time_s": 45.996844,
                "source_pts_s": 46.021744,
                "image": "assets/frames/E9/f00001372.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 52.5,
            "range_s": [
              51.996422,
              52.996322
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1550,
                "time_s": 51.996422,
                "source_pts_s": 52.021322,
                "image": "assets/frames/E9/f00001550.jpg"
              },
              {
                "frame_index": 1580,
                "time_s": 52.996322,
                "source_pts_s": 53.021222,
                "image": "assets/frames/E9/f00001580.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a07",
          "level": "action",
          "parent_id": "E9.s02",
          "start_s": 52.5,
          "end_s": 56.5,
          "label": "Slide the mixing bowl back to clear the cutting board",
          "verb": "slide",
          "hands": "left",
          "hand_roles": {
            "left": "Slide the mixing bowl back to clear the cutting board"
          },
          "object_ids": [
            "E9.jug",
            "E9.board"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: slide the mixing bowl back to clear the cutting board. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1550,
              "time_s": 51.996422,
              "source_pts_s": 52.021322,
              "image": "assets/frames/E9/f00001550.jpg"
            },
            {
              "frame_index": 1610,
              "time_s": 53.996222,
              "source_pts_s": 54.021122,
              "image": "assets/frames/E9/f00001610.jpg"
            },
            {
              "frame_index": 1670,
              "time_s": 55.996122,
              "source_pts_s": 56.021022,
              "image": "assets/frames/E9/f00001670.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 52.5,
            "range_s": [
              51.996422,
              52.996322
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1550,
                "time_s": 51.996422,
                "source_pts_s": 52.021322,
                "image": "assets/frames/E9/f00001550.jpg"
              },
              {
                "frame_index": 1580,
                "time_s": 52.996322,
                "source_pts_s": 53.021222,
                "image": "assets/frames/E9/f00001580.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 56.5,
            "range_s": [
              55.996122,
              56.996022
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1670,
                "time_s": 55.996122,
                "source_pts_s": 56.021022,
                "image": "assets/frames/E9/f00001670.jpg"
              },
              {
                "frame_index": 1700,
                "time_s": 56.996022,
                "source_pts_s": 57.020922,
                "image": "assets/frames/E9/f00001700.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a08",
          "level": "action",
          "parent_id": "E9.s02",
          "start_s": 56.5,
          "end_s": 69.5,
          "label": "Peel the wrapper from the first cheese block",
          "verb": "peel",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.wrappers",
            "E9.cheese1"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: peel the wrapper from the first cheese block. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1700,
              "time_s": 56.996022,
              "source_pts_s": 57.020922,
              "image": "assets/frames/E9/f00001700.jpg"
            },
            {
              "frame_index": 1880,
              "time_s": 62.995622,
              "source_pts_s": 63.020522,
              "image": "assets/frames/E9/f00001880.jpg"
            },
            {
              "frame_index": 2030,
              "time_s": 67.995222,
              "source_pts_s": 68.020122,
              "image": "assets/frames/E9/f00002030.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 56.5,
            "range_s": [
              55.996122,
              56.996022
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1670,
                "time_s": 55.996122,
                "source_pts_s": 56.021022,
                "image": "assets/frames/E9/f00001670.jpg"
              },
              {
                "frame_index": 1700,
                "time_s": 56.996022,
                "source_pts_s": 57.020922,
                "image": "assets/frames/E9/f00001700.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 69.5,
            "range_s": [
              68.995222,
              69.995122
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2060,
                "time_s": 68.995222,
                "source_pts_s": 69.020122,
                "image": "assets/frames/E9/f00002060.jpg"
              },
              {
                "frame_index": 2090,
                "time_s": 69.995122,
                "source_pts_s": 70.020022,
                "image": "assets/frames/E9/f00002090.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a09",
          "level": "action",
          "parent_id": "E9.s02",
          "start_s": 69.5,
          "end_s": 73.5,
          "label": "Set the cheese on the board and move the wrapper aside",
          "verb": "set",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.cheese1",
            "E9.wrappers",
            "E9.board"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: set the cheese on the board and move the wrapper aside. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2060,
              "time_s": 68.995222,
              "source_pts_s": 69.020122,
              "image": "assets/frames/E9/f00002060.jpg"
            },
            {
              "frame_index": 2120,
              "time_s": 70.995022,
              "source_pts_s": 71.019922,
              "image": "assets/frames/E9/f00002120.jpg"
            },
            {
              "frame_index": 2180,
              "time_s": 72.994922,
              "source_pts_s": 73.019822,
              "image": "assets/frames/E9/f00002180.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 69.5,
            "range_s": [
              68.995222,
              69.995122
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2060,
                "time_s": 68.995222,
                "source_pts_s": 69.020122,
                "image": "assets/frames/E9/f00002060.jpg"
              },
              {
                "frame_index": 2090,
                "time_s": 69.995122,
                "source_pts_s": 70.020022,
                "image": "assets/frames/E9/f00002090.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 73.5,
            "range_s": [
              72.994922,
              73.994822
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2180,
                "time_s": 72.994922,
                "source_pts_s": 73.019822,
                "image": "assets/frames/E9/f00002180.jpg"
              },
              {
                "frame_index": 2210,
                "time_s": 73.994822,
                "source_pts_s": 74.019722,
                "image": "assets/frames/E9/f00002210.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a10",
          "level": "action",
          "parent_id": "E9.s03",
          "start_s": 73.5,
          "end_s": 85.5,
          "label": "Slice the cheese block into broad pieces",
          "verb": "slice",
          "hands": "right",
          "hand_roles": {
            "right": "Slice the cheese block into broad pieces"
          },
          "object_ids": [
            "E9.cheese1",
            "E9.knife",
            "E9.board"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: slice the cheese block into broad pieces. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2210,
              "time_s": 73.994822,
              "source_pts_s": 74.019722,
              "image": "assets/frames/E9/f00002210.jpg"
            },
            {
              "frame_index": 2360,
              "time_s": 78.994422,
              "source_pts_s": 79.019322,
              "image": "assets/frames/E9/f00002360.jpg"
            },
            {
              "frame_index": 2510,
              "time_s": 83.994122,
              "source_pts_s": 84.019022,
              "image": "assets/frames/E9/f00002510.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 73.5,
            "range_s": [
              72.994922,
              73.994822
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2180,
                "time_s": 72.994922,
                "source_pts_s": 73.019822,
                "image": "assets/frames/E9/f00002180.jpg"
              },
              {
                "frame_index": 2210,
                "time_s": 73.994822,
                "source_pts_s": 74.019722,
                "image": "assets/frames/E9/f00002210.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 85.5,
            "range_s": [
              84.994022,
              85.994022
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2540,
                "time_s": 84.994022,
                "source_pts_s": 85.018922,
                "image": "assets/frames/E9/f00002540.jpg"
              },
              {
                "frame_index": 2570,
                "time_s": 85.994022,
                "source_pts_s": 86.018922,
                "image": "assets/frames/E9/f00002570.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a11",
          "level": "action",
          "parent_id": "E9.s03",
          "start_s": 85.5,
          "end_s": 130.5,
          "label": "Reposition the pieces and cut them into narrower strips",
          "verb": "reposition",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.cheese1",
            "E9.knife",
            "E9.board"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: reposition the pieces and cut them into narrower strips. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2600,
              "time_s": 86.993922,
              "source_pts_s": 87.018822,
              "image": "assets/frames/E9/f00002600.jpg"
            },
            {
              "frame_index": 3198,
              "time_s": 106.992467,
              "source_pts_s": 107.017367,
              "image": "assets/frames/E9/f00003198.jpg"
            },
            {
              "frame_index": 3827,
              "time_s": 127.9909,
              "source_pts_s": 128.0158,
              "image": "assets/frames/E9/f00003827.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 85.5,
            "range_s": [
              84.994022,
              85.994022
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2540,
                "time_s": 84.994022,
                "source_pts_s": 85.018922,
                "image": "assets/frames/E9/f00002540.jpg"
              },
              {
                "frame_index": 2570,
                "time_s": 85.994022,
                "source_pts_s": 86.018922,
                "image": "assets/frames/E9/f00002570.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 130.5,
            "range_s": [
              129.990922,
              130.990822
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 3887,
                "time_s": 129.990922,
                "source_pts_s": 130.015822,
                "image": "assets/frames/E9/f00003887.jpg"
              },
              {
                "frame_index": 3917,
                "time_s": 130.990822,
                "source_pts_s": 131.015722,
                "image": "assets/frames/E9/f00003917.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a12",
          "level": "action",
          "parent_id": "E9.s04",
          "start_s": 130.5,
          "end_s": 136.5,
          "label": "Gather and shift the cheese strips to the side",
          "verb": "gather",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.cheese1",
            "E9.board"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: gather and shift the cheese strips to the side. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 3917,
              "time_s": 130.990822,
              "source_pts_s": 131.015722,
              "image": "assets/frames/E9/f00003917.jpg"
            },
            {
              "frame_index": 3976,
              "time_s": 132.990711,
              "source_pts_s": 133.015611,
              "image": "assets/frames/E9/f00003976.jpg"
            },
            {
              "frame_index": 4036,
              "time_s": 134.990511,
              "source_pts_s": 135.015411,
              "image": "assets/frames/E9/f00004036.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 130.5,
            "range_s": [
              129.990922,
              130.990822
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 3887,
                "time_s": 129.990922,
                "source_pts_s": 130.015822,
                "image": "assets/frames/E9/f00003887.jpg"
              },
              {
                "frame_index": 3917,
                "time_s": 130.990822,
                "source_pts_s": 131.015722,
                "image": "assets/frames/E9/f00003917.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 136.5,
            "range_s": [
              135.990411,
              136.990411
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 4066,
                "time_s": 135.990411,
                "source_pts_s": 136.015311,
                "image": "assets/frames/E9/f00004066.jpg"
              },
              {
                "frame_index": 4096,
                "time_s": 136.990411,
                "source_pts_s": 137.015311,
                "image": "assets/frames/E9/f00004096.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed.",
            "Forearm obscures much of the manipulation around 134.3\u2013136.3s. Final strips remain at the board edge; transfer onto a plate is not established."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a13",
          "level": "action",
          "parent_id": "E9.s05",
          "start_s": 136.5,
          "end_s": 148.5,
          "label": "Open the carton and remove a foil-wrapped block",
          "verb": "open",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.wrappers",
            "E9.cheese2"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: open the carton and remove a foil-wrapped block. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 4126,
              "time_s": 137.990311,
              "source_pts_s": 138.015211,
              "image": "assets/frames/E9/f00004126.jpg"
            },
            {
              "frame_index": 4336,
              "time_s": 144.989811,
              "source_pts_s": 145.014711,
              "image": "assets/frames/E9/f00004336.jpg"
            },
            {
              "frame_index": 4426,
              "time_s": 147.989611,
              "source_pts_s": 148.014511,
              "image": "assets/frames/E9/f00004426.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 136.5,
            "range_s": [
              135.990411,
              136.990411
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 4066,
                "time_s": 135.990411,
                "source_pts_s": 136.015311,
                "image": "assets/frames/E9/f00004066.jpg"
              },
              {
                "frame_index": 4096,
                "time_s": 136.990411,
                "source_pts_s": 137.015311,
                "image": "assets/frames/E9/f00004096.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 148.5,
            "range_s": [
              147.989611,
              148.989511
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 4426,
                "time_s": 147.989611,
                "source_pts_s": 148.014511,
                "image": "assets/frames/E9/f00004426.jpg"
              },
              {
                "frame_index": 4456,
                "time_s": 148.989511,
                "source_pts_s": 149.014411,
                "image": "assets/frames/E9/f00004456.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a14",
          "level": "action",
          "parent_id": "E9.s05",
          "start_s": 148.5,
          "end_s": 174.5,
          "label": "Unfold the foil and paper around the second block",
          "verb": "unfold",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.wrappers",
            "E9.cheese2"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: unfold the foil and paper around the second block. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 4486,
              "time_s": 149.989411,
              "source_pts_s": 150.014311,
              "image": "assets/frames/E9/f00004486.jpg"
            },
            {
              "frame_index": 4846,
              "time_s": 161.988611,
              "source_pts_s": 162.013511,
              "image": "assets/frames/E9/f00004846.jpg"
            },
            {
              "frame_index": 5176,
              "time_s": 172.987811,
              "source_pts_s": 173.012711,
              "image": "assets/frames/E9/f00005176.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 148.5,
            "range_s": [
              147.989611,
              148.989511
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 4426,
                "time_s": 147.989611,
                "source_pts_s": 148.014511,
                "image": "assets/frames/E9/f00004426.jpg"
              },
              {
                "frame_index": 4456,
                "time_s": 148.989511,
                "source_pts_s": 149.014411,
                "image": "assets/frames/E9/f00004456.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 174.5,
            "range_s": [
              173.987811,
              174.987667
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5206,
                "time_s": 173.987811,
                "source_pts_s": 174.012711,
                "image": "assets/frames/E9/f00005206.jpg"
              },
              {
                "frame_index": 5235,
                "time_s": 174.987667,
                "source_pts_s": 175.012567,
                "image": "assets/frames/E9/f00005235.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a15",
          "level": "action",
          "parent_id": "E9.s06",
          "start_s": 174.5,
          "end_s": 177.5,
          "label": "Cut a portion from the unwrapped block",
          "verb": "cut",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.cheese2",
            "E9.knife"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: cut a portion from the unwrapped block. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 5235,
              "time_s": 174.987667,
              "source_pts_s": 175.012567,
              "image": "assets/frames/E9/f00005235.jpg"
            },
            {
              "frame_index": 5265,
              "time_s": 175.987667,
              "source_pts_s": 176.012567,
              "image": "assets/frames/E9/f00005265.jpg"
            },
            {
              "frame_index": 5295,
              "time_s": 176.987567,
              "source_pts_s": 177.012467,
              "image": "assets/frames/E9/f00005295.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 174.5,
            "range_s": [
              173.987811,
              174.987667
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5206,
                "time_s": 173.987811,
                "source_pts_s": 174.012711,
                "image": "assets/frames/E9/f00005206.jpg"
              },
              {
                "frame_index": 5235,
                "time_s": 174.987667,
                "source_pts_s": 175.012567,
                "image": "assets/frames/E9/f00005235.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 177.5,
            "range_s": [
              176.987567,
              177.987467
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5295,
                "time_s": 176.987567,
                "source_pts_s": 177.012467,
                "image": "assets/frames/E9/f00005295.jpg"
              },
              {
                "frame_index": 5325,
                "time_s": 177.987467,
                "source_pts_s": 178.012367,
                "image": "assets/frames/E9/f00005325.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed.",
            "Knife crosses the block and a portion separates; exact blade contact is partly occluded."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a16",
          "level": "action",
          "parent_id": "E9.s06",
          "start_s": 177.5,
          "end_s": 184.5,
          "label": "Fold wrapper around the separated portion and move it aside",
          "verb": "fold",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.cheese2",
            "E9.wrappers"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: fold wrapper around the separated portion and move it aside. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 5325,
              "time_s": 177.987467,
              "source_pts_s": 178.012367,
              "image": "assets/frames/E9/f00005325.jpg"
            },
            {
              "frame_index": 5415,
              "time_s": 180.987267,
              "source_pts_s": 181.012167,
              "image": "assets/frames/E9/f00005415.jpg"
            },
            {
              "frame_index": 5505,
              "time_s": 183.987067,
              "source_pts_s": 184.011967,
              "image": "assets/frames/E9/f00005505.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 177.5,
            "range_s": [
              176.987567,
              177.987467
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5295,
                "time_s": 176.987567,
                "source_pts_s": 177.012467,
                "image": "assets/frames/E9/f00005295.jpg"
              },
              {
                "frame_index": 5325,
                "time_s": 177.987467,
                "source_pts_s": 178.012367,
                "image": "assets/frames/E9/f00005325.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 184.5,
            "range_s": [
              183.987067,
              184.986967
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5505,
                "time_s": 183.987067,
                "source_pts_s": 184.011967,
                "image": "assets/frames/E9/f00005505.jpg"
              },
              {
                "frame_index": 5535,
                "time_s": 184.986967,
                "source_pts_s": 185.011867,
                "image": "assets/frames/E9/f00005535.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a17",
          "level": "action",
          "parent_id": "E9.s06",
          "start_s": 184.5,
          "end_s": 189.5,
          "label": "Remove remaining paper and set cheese on the board",
          "verb": "remove",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.cheese2",
            "E9.wrappers",
            "E9.board"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: remove remaining paper and set cheese on the board. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 5535,
              "time_s": 184.986967,
              "source_pts_s": 185.011867,
              "image": "assets/frames/E9/f00005535.jpg"
            },
            {
              "frame_index": 5595,
              "time_s": 186.986867,
              "source_pts_s": 187.011767,
              "image": "assets/frames/E9/f00005595.jpg"
            },
            {
              "frame_index": 5654,
              "time_s": 188.986711,
              "source_pts_s": 189.011611,
              "image": "assets/frames/E9/f00005654.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 184.5,
            "range_s": [
              183.987067,
              184.986967
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5505,
                "time_s": 183.987067,
                "source_pts_s": 184.011967,
                "image": "assets/frames/E9/f00005505.jpg"
              },
              {
                "frame_index": 5535,
                "time_s": 184.986967,
                "source_pts_s": 185.011867,
                "image": "assets/frames/E9/f00005535.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 189.5,
            "range_s": [
              188.986711,
              189.986633
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5654,
                "time_s": 188.986711,
                "source_pts_s": 189.011611,
                "image": "assets/frames/E9/f00005654.jpg"
              },
              {
                "frame_index": 5683,
                "time_s": 189.986633,
                "source_pts_s": 190.011533,
                "image": "assets/frames/E9/f00005683.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a18",
          "level": "action",
          "parent_id": "E9.s07",
          "start_s": 189.5,
          "end_s": 199.5,
          "label": "Slice the second cheese portion into broad pieces",
          "verb": "slice",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.cheese2",
            "E9.knife"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: slice the second cheese portion into broad pieces. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 5712,
              "time_s": 190.986544,
              "source_pts_s": 191.011444,
              "image": "assets/frames/E9/f00005712.jpg"
            },
            {
              "frame_index": 5831,
              "time_s": 194.986267,
              "source_pts_s": 195.011167,
              "image": "assets/frames/E9/f00005831.jpg"
            },
            {
              "frame_index": 5921,
              "time_s": 197.986067,
              "source_pts_s": 198.010967,
              "image": "assets/frames/E9/f00005921.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 189.5,
            "range_s": [
              188.986711,
              189.986633
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5654,
                "time_s": 188.986711,
                "source_pts_s": 189.011611,
                "image": "assets/frames/E9/f00005654.jpg"
              },
              {
                "frame_index": 5683,
                "time_s": 189.986633,
                "source_pts_s": 190.011533,
                "image": "assets/frames/E9/f00005683.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 199.5,
            "range_s": [
              198.986011,
              199.985911
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5949,
                "time_s": 198.986011,
                "source_pts_s": 199.010911,
                "image": "assets/frames/E9/f00005949.jpg"
              },
              {
                "frame_index": 5979,
                "time_s": 199.985911,
                "source_pts_s": 200.010811,
                "image": "assets/frames/E9/f00005979.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E9.a19",
          "level": "action",
          "parent_id": "E9.s07",
          "start_s": 199.5,
          "end_s": 209.6853,
          "label": "Cut the broad pieces into narrower strips",
          "verb": "cut",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E9.cheese2",
            "E9.knife"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: cut the broad pieces into narrower strips. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 5979,
              "time_s": 199.985911,
              "source_pts_s": 200.010811,
              "image": "assets/frames/E9/f00005979.jpg"
            },
            {
              "frame_index": 6099,
              "time_s": 203.985711,
              "source_pts_s": 204.010611,
              "image": "assets/frames/E9/f00006099.jpg"
            },
            {
              "frame_index": 6246,
              "time_s": 208.9853,
              "source_pts_s": 209.0102,
              "image": "assets/frames/E9/f00006246.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 199.5,
            "range_s": [
              198.986011,
              199.985911
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 5949,
                "time_s": 198.986011,
                "source_pts_s": 199.010911,
                "image": "assets/frames/E9/f00005949.jpg"
              },
              {
                "frame_index": 5979,
                "time_s": 199.985911,
                "source_pts_s": 200.010811,
                "image": "assets/frames/E9/f00005979.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 209.6853,
            "range_s": [
              209.6853,
              209.6853
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 6266,
                "time_s": 209.651967,
                "source_pts_s": 209.676867,
                "image": "assets/frames/E9/f00006266.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "sampling": {
        "base_frames_extracted": 211,
        "base_frames_visually_reviewed": 211,
        "dense_frames_extracted": 38,
        "dense_frames_visually_reviewed": 38,
        "unique_frames_visually_reviewed": 241,
        "source_frame_count": 6267
      },
      "playback": {
        "sha256": "e37fa3b46f160c090720cbb3d39b227fcac2bea156f53c25c9916b059e718a97",
        "bytes": 187292829,
        "playback_width": 1280,
        "playback_height": 720,
        "source_width": 1280,
        "source_height": 720,
        "frame_count": 6267,
        "max_timestamp_error_s": 2.842170943040401e-14,
        "duration_s": 209.6853,
        "audio_stream_count": 0,
        "playback_time_offset_s": 0.024,
        "kind": "original_video_audio_removed_served_from_r2",
        "source_sha256": "b84574c94d9341d40014cafdbbc465fffeb4e92a9b7b541922cc4b95c82d933c",
        "url": "https://media-demo.theroboticsstack.com/media/E9.mp4?v=e37fa3b46f16"
      }
    },
    {
      "id": "E6",
      "source": "https://media-demo.theroboticsstack.com/media/E6.mp4?v=f293e148ebdd",
      "sha256": "e8ae83f772fda809a962e40344096335d46d3c96f7ff70aef7d46330eac7fde7",
      "duration_s": 115.890378,
      "width": 1280,
      "height": 720,
      "fps_fraction": "14220000/474097",
      "decoded_frame_count": 3476,
      "audio_stream_count": 1,
      "source_start_pts_s": 0.0,
      "last_frame_s": 115.857089,
      "frame_interval_min_s": 0.023377999999993904,
      "frame_interval_max_s": 0.071456,
      "nonincreasing_pts_count": 0,
      "decode_diagnostics": "",
      "view_layout": "single_egocentric_full_frame",
      "sample_interval_target_s": 1.0,
      "extraction_settings": {
        "maximum_image_width": 1280,
        "crop": null,
        "frame_selection": "nearest_actual_PTS"
      },
      "title": "Reposition a bedsheet",
      "observed_activity": "Gather a pink blanket and two pillows toward one end of a bed, then pull, fit and smooth the white sheet around the mattress.",
      "inferred_goal": "Straighten the sheet around the mattress after moving the loose bedding aside.",
      "assigned_instruction": null,
      "playback_time_offset_s": 0,
      "visible_outcome": "The white sheet covers more of the mattress edges; the blanket and pillows remain piled at one end.",
      "task_success": "unknown",
      "unknowns": [
        "This is a partial bed-making sequence; the blanket is not spread again.",
        "Tucks beneath the mattress are partly occluded.",
        "Repeated pulls are grouped; the exact number of regrips is not resolved.",
        "A cat moves around the bed; its movements are not hand-object task actions."
      ],
      "selection_rationale": "Strong: large state changes, distinct object handling and a clear preparation-to-sheet transition.",
      "annotation_provenance": "assistant_visual_review",
      "review_status": "pending",
      "credit": {
        "attribution": "Eidon AI \u00b7 Tracker POV \u00b7 contributor 8 \u00b7 recording 11817",
        "source_url": "https://huggingface.co/datasets/eidon-ai/tracker-pov/resolve/main/recordings/data/011/11817_video.mp4",
        "metadata_url": "https://huggingface.co/datasets/eidon-ai/tracker-pov/resolve/main/recordings/metadata.parquet",
        "license": "CC BY 4.0",
        "license_url": "https://creativecommons.org/licenses/by/4.0/",
        "changes": "Complete video re-encoded as H.264 at 8 Mbps for delivery, native resolution, audio omitted, first video frame normalised to zero; every source frame time verified. Evidence resized to at most 1280px; assistant temporal labels added."
      },
      "protocol_role": "featured",
      "objects": [
        {
          "id": "E6.blanket",
          "description": "Pink blanket",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E6.plain_pillow",
          "description": "Plain cream pillow",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E6.pattern_pillow",
          "description": "Cream pillow with blue and yellow pattern",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E6.sheet",
          "description": "White sheet",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "E6.mattress",
          "description": "Mattress",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        }
      ],
      "subtasks": [
        {
          "id": "E6.s01",
          "level": "subtask",
          "parent_id": null,
          "start_s": 0,
          "end_s": 15.5,
          "label": "Gather the blanket to one end",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E6.blanket"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: gather the blanket to one end. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 0,
              "time_s": 0.0,
              "source_pts_s": 0.0,
              "image": "assets/frames/E6/f00000000.jpg"
            },
            {
              "frame_index": 239,
              "time_s": 7.994656,
              "source_pts_s": 7.994656,
              "image": "assets/frames/E6/f00000239.jpg"
            },
            {
              "frame_index": 478,
              "time_s": 15.987789,
              "source_pts_s": 15.987789,
              "image": "assets/frames/E6/f00000478.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.0,
                "image": "assets/frames/E6/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 15.5,
            "range_s": [
              14.989056,
              15.987789
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 448,
                "time_s": 14.989056,
                "source_pts_s": 14.989056,
                "image": "assets/frames/E6/f00000448.jpg"
              },
              {
                "frame_index": 478,
                "time_s": 15.987789,
                "source_pts_s": 15.987789,
                "image": "assets/frames/E6/f00000478.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.s02",
          "level": "subtask",
          "parent_id": null,
          "start_s": 15.5,
          "end_s": 20.382167,
          "label": "Move the pillows onto the blanket",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E6.blanket",
            "E6.pattern_pillow",
            "E6.plain_pillow"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: move the pillows onto the blanket. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 478,
              "time_s": 15.987789,
              "source_pts_s": 15.987789,
              "image": "assets/frames/E6/f00000478.jpg"
            },
            {
              "frame_index": 538,
              "time_s": 17.985233,
              "source_pts_s": 17.985233,
              "image": "assets/frames/E6/f00000538.jpg"
            },
            {
              "frame_index": 614,
              "time_s": 20.515322,
              "source_pts_s": 20.515322,
              "image": "assets/frames/E6/f00000614.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 15.5,
            "range_s": [
              14.989056,
              15.987789
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 448,
                "time_s": 14.989056,
                "source_pts_s": 14.989056,
                "image": "assets/frames/E6/f00000448.jpg"
              },
              {
                "frame_index": 478,
                "time_s": 15.987789,
                "source_pts_s": 15.987789,
                "image": "assets/frames/E6/f00000478.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 20.382167,
            "range_s": [
              20.249011,
              20.515322
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Patterned pillow rests on the pile and the hands withdraw from it.",
            "evidence_frames": [
              {
                "frame_index": 606,
                "time_s": 20.249011,
                "source_pts_s": 20.249011,
                "image": "assets/frames/E6/f00000606.jpg"
              },
              {
                "frame_index": 614,
                "time_s": 20.515322,
                "source_pts_s": 20.515322,
                "image": "assets/frames/E6/f00000614.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.s03",
          "level": "subtask",
          "parent_id": null,
          "start_s": 20.382167,
          "end_s": 52.5,
          "label": "Pull the sheet across the mattress and fit a corner",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E6.mattress",
            "E6.sheet"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: pull the sheet across the mattress and fit a corner. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 614,
              "time_s": 20.515322,
              "source_pts_s": 20.515322,
              "image": "assets/frames/E6/f00000614.jpg"
            },
            {
              "frame_index": 1079,
              "time_s": 35.995556,
              "source_pts_s": 35.995556,
              "image": "assets/frames/E6/f00001079.jpg"
            },
            {
              "frame_index": 1559,
              "time_s": 52.008522,
              "source_pts_s": 52.008522,
              "image": "assets/frames/E6/f00001559.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 20.382167,
            "range_s": [
              20.249011,
              20.515322
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Patterned pillow rests on the pile and the hands withdraw from it.",
            "evidence_frames": [
              {
                "frame_index": 606,
                "time_s": 20.249011,
                "source_pts_s": 20.249011,
                "image": "assets/frames/E6/f00000606.jpg"
              },
              {
                "frame_index": 614,
                "time_s": 20.515322,
                "source_pts_s": 20.515322,
                "image": "assets/frames/E6/f00000614.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 52.5,
            "range_s": [
              52.008522,
              53.0073
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1559,
                "time_s": 52.008522,
                "source_pts_s": 52.008522,
                "image": "assets/frames/E6/f00001559.jpg"
              },
              {
                "frame_index": 1589,
                "time_s": 53.0073,
                "source_pts_s": 53.0073,
                "image": "assets/frames/E6/f00001589.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.s04",
          "level": "subtask",
          "parent_id": null,
          "start_s": 52.5,
          "end_s": 80.5,
          "label": "Stretch and fit the adjacent edges",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E6.mattress",
            "E6.sheet"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: stretch and fit the adjacent edges. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1559,
              "time_s": 52.008522,
              "source_pts_s": 52.008522,
              "image": "assets/frames/E6/f00001559.jpg"
            },
            {
              "frame_index": 2009,
              "time_s": 66.986211,
              "source_pts_s": 66.986211,
              "image": "assets/frames/E6/f00002009.jpg"
            },
            {
              "frame_index": 2400,
              "time_s": 80.002933,
              "source_pts_s": 80.002933,
              "image": "assets/frames/E6/f00002400.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 52.5,
            "range_s": [
              52.008522,
              53.0073
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1559,
                "time_s": 52.008522,
                "source_pts_s": 52.008522,
                "image": "assets/frames/E6/f00001559.jpg"
              },
              {
                "frame_index": 1589,
                "time_s": 53.0073,
                "source_pts_s": 53.0073,
                "image": "assets/frames/E6/f00001589.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 80.5,
            "range_s": [
              80.002933,
              81.0016
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2400,
                "time_s": 80.002933,
                "source_pts_s": 80.002933,
                "image": "assets/frames/E6/f00002400.jpg"
              },
              {
                "frame_index": 2430,
                "time_s": 81.0016,
                "source_pts_s": 81.0016,
                "image": "assets/frames/E6/f00002430.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.s05",
          "level": "subtask",
          "parent_id": null,
          "start_s": 80.5,
          "end_s": 103.5,
          "label": "Continue pulling around the foot and sides",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E6.mattress",
            "E6.sheet"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: continue pulling around the foot and sides. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2400,
              "time_s": 80.002933,
              "source_pts_s": 80.002933,
              "image": "assets/frames/E6/f00002400.jpg"
            },
            {
              "frame_index": 2760,
              "time_s": 91.987556,
              "source_pts_s": 91.987556,
              "image": "assets/frames/E6/f00002760.jpg"
            },
            {
              "frame_index": 3090,
              "time_s": 103.006878,
              "source_pts_s": 103.006878,
              "image": "assets/frames/E6/f00003090.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 80.5,
            "range_s": [
              80.002933,
              81.0016
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2400,
                "time_s": 80.002933,
                "source_pts_s": 80.002933,
                "image": "assets/frames/E6/f00002400.jpg"
              },
              {
                "frame_index": 2430,
                "time_s": 81.0016,
                "source_pts_s": 81.0016,
                "image": "assets/frames/E6/f00002430.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 103.5,
            "range_s": [
              103.006878,
              104.005544
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 3090,
                "time_s": 103.006878,
                "source_pts_s": 103.006878,
                "image": "assets/frames/E6/f00003090.jpg"
              },
              {
                "frame_index": 3120,
                "time_s": 104.005544,
                "source_pts_s": 104.005544,
                "image": "assets/frames/E6/f00003120.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.s06",
          "level": "subtask",
          "parent_id": null,
          "start_s": 103.5,
          "end_s": 115.890378,
          "label": "Retuck and smooth the wall-side corner",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "E6.mattress",
            "E6.sheet"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: retuck and smooth the wall-side corner. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 3090,
              "time_s": 103.006878,
              "source_pts_s": 103.006878,
              "image": "assets/frames/E6/f00003090.jpg"
            },
            {
              "frame_index": 3300,
              "time_s": 109.992844,
              "source_pts_s": 109.992844,
              "image": "assets/frames/E6/f00003300.jpg"
            },
            {
              "frame_index": 3475,
              "time_s": 115.857089,
              "source_pts_s": 115.857089,
              "image": "assets/frames/E6/f00003475.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 103.5,
            "range_s": [
              103.006878,
              104.005544
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 3090,
                "time_s": 103.006878,
                "source_pts_s": 103.006878,
                "image": "assets/frames/E6/f00003090.jpg"
              },
              {
                "frame_index": 3120,
                "time_s": 104.005544,
                "source_pts_s": 104.005544,
                "image": "assets/frames/E6/f00003120.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 115.890378,
            "range_s": [
              115.890378,
              115.890378
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 3475,
                "time_s": 115.857089,
                "source_pts_s": 115.857089,
                "image": "assets/frames/E6/f00003475.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "actions": [
        {
          "id": "E6.a01",
          "level": "action",
          "parent_id": "E6.s01",
          "start_s": 0,
          "end_s": 7.5,
          "label": "Gather and lift the near part of the pink blanket",
          "verb": "gather",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.blanket"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: gather and lift the near part of the pink blanket. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 29,
              "time_s": 1.003544,
              "source_pts_s": 1.003544,
              "image": "assets/frames/E6/f00000029.jpg"
            },
            {
              "frame_index": 149,
              "time_s": 4.998433,
              "source_pts_s": 4.998433,
              "image": "assets/frames/E6/f00000149.jpg"
            },
            {
              "frame_index": 209,
              "time_s": 6.995878,
              "source_pts_s": 6.995878,
              "image": "assets/frames/E6/f00000209.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.0,
                "image": "assets/frames/E6/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 7.5,
            "range_s": [
              6.995878,
              7.994656
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 209,
                "time_s": 6.995878,
                "source_pts_s": 6.995878,
                "image": "assets/frames/E6/f00000209.jpg"
              },
              {
                "frame_index": 239,
                "time_s": 7.994656,
                "source_pts_s": 7.994656,
                "image": "assets/frames/E6/f00000239.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a02",
          "level": "action",
          "parent_id": "E6.s01",
          "start_s": 9.5,
          "end_s": 13.5,
          "label": "Lift another section of blanket onto the pile",
          "verb": "lift",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.blanket"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: lift another section of blanket onto the pile. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 299,
              "time_s": 9.995478,
              "source_pts_s": 9.995478,
              "image": "assets/frames/E6/f00000299.jpg"
            },
            {
              "frame_index": 359,
              "time_s": 11.992922,
              "source_pts_s": 11.992922,
              "image": "assets/frames/E6/f00000359.jpg"
            },
            {
              "frame_index": 389,
              "time_s": 12.991589,
              "source_pts_s": 12.991589,
              "image": "assets/frames/E6/f00000389.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 9.5,
            "range_s": [
              8.993433,
              9.995478
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 269,
                "time_s": 8.993433,
                "source_pts_s": 8.993433,
                "image": "assets/frames/E6/f00000269.jpg"
              },
              {
                "frame_index": 299,
                "time_s": 9.995478,
                "source_pts_s": 9.995478,
                "image": "assets/frames/E6/f00000299.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 13.5,
            "range_s": [
              12.991589,
              13.990389
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 389,
                "time_s": 12.991589,
                "source_pts_s": 12.991589,
                "image": "assets/frames/E6/f00000389.jpg"
              },
              {
                "frame_index": 418,
                "time_s": 13.990389,
                "source_pts_s": 13.990389,
                "image": "assets/frames/E6/f00000418.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a03",
          "level": "action",
          "parent_id": "E6.s02",
          "start_s": 15.5,
          "end_s": 18.5,
          "label": "Lift the plain pillow onto the gathered blanket",
          "verb": "lift",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.plain_pillow",
            "E6.blanket"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: lift the plain pillow onto the gathered blanket. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 478,
              "time_s": 15.987789,
              "source_pts_s": 15.987789,
              "image": "assets/frames/E6/f00000478.jpg"
            },
            {
              "frame_index": 508,
              "time_s": 16.986456,
              "source_pts_s": 16.986456,
              "image": "assets/frames/E6/f00000508.jpg"
            },
            {
              "frame_index": 538,
              "time_s": 17.985233,
              "source_pts_s": 17.985233,
              "image": "assets/frames/E6/f00000538.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 15.5,
            "range_s": [
              14.989056,
              15.987789
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 448,
                "time_s": 14.989056,
                "source_pts_s": 14.989056,
                "image": "assets/frames/E6/f00000448.jpg"
              },
              {
                "frame_index": 478,
                "time_s": 15.987789,
                "source_pts_s": 15.987789,
                "image": "assets/frames/E6/f00000478.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 18.5,
            "range_s": [
              17.985233,
              18.9839
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 538,
                "time_s": 17.985233,
                "source_pts_s": 17.985233,
                "image": "assets/frames/E6/f00000538.jpg"
              },
              {
                "frame_index": 568,
                "time_s": 18.9839,
                "source_pts_s": 18.9839,
                "image": "assets/frames/E6/f00000568.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a04",
          "level": "action",
          "parent_id": "E6.s02",
          "start_s": 18.5,
          "end_s": 20.382167,
          "label": "Move the patterned pillow onto the same pile",
          "verb": "move",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.pattern_pillow",
            "E6.blanket"
          ],
          "state_before": "Patterned pillow at the head of the bed.",
          "state_after": "Patterned pillow rests with the plain pillow on the blanket.",
          "evidence_description": "Ordered sampled images support: move the patterned pillow onto the same pile. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 568,
              "time_s": 18.9839,
              "source_pts_s": 18.9839,
              "image": "assets/frames/E6/f00000568.jpg"
            },
            {
              "frame_index": 599,
              "time_s": 20.015989,
              "source_pts_s": 20.015989,
              "image": "assets/frames/E6/f00000599.jpg"
            },
            {
              "frame_index": 629,
              "time_s": 21.014656,
              "source_pts_s": 21.014656,
              "image": "assets/frames/E6/f00000629.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 18.5,
            "range_s": [
              17.985233,
              18.9839
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 538,
                "time_s": 17.985233,
                "source_pts_s": 17.985233,
                "image": "assets/frames/E6/f00000538.jpg"
              },
              {
                "frame_index": 568,
                "time_s": 18.9839,
                "source_pts_s": 18.9839,
                "image": "assets/frames/E6/f00000568.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 20.382167,
            "range_s": [
              20.249011,
              20.515322
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Patterned pillow rests on the pile and the hands withdraw from it.",
            "evidence_frames": [
              {
                "frame_index": 606,
                "time_s": 20.249011,
                "source_pts_s": 20.249011,
                "image": "assets/frames/E6/f00000606.jpg"
              },
              {
                "frame_index": 614,
                "time_s": 20.515322,
                "source_pts_s": 20.515322,
                "image": "assets/frames/E6/f00000614.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a05",
          "level": "action",
          "parent_id": "E6.s03",
          "start_s": 21.5,
          "end_s": 32.5,
          "label": "Pull the white sheet outward along the exposed mattress",
          "verb": "pull",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: pull the white sheet outward along the exposed mattress. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 659,
              "time_s": 22.013433,
              "source_pts_s": 22.013433,
              "image": "assets/frames/E6/f00000659.jpg"
            },
            {
              "frame_index": 839,
              "time_s": 28.002456,
              "source_pts_s": 28.002456,
              "image": "assets/frames/E6/f00000839.jpg"
            },
            {
              "frame_index": 929,
              "time_s": 30.998567,
              "source_pts_s": 30.998567,
              "image": "assets/frames/E6/f00000929.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 21.5,
            "range_s": [
              21.014656,
              22.013433
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 629,
                "time_s": 21.014656,
                "source_pts_s": 21.014656,
                "image": "assets/frames/E6/f00000629.jpg"
              },
              {
                "frame_index": 659,
                "time_s": 22.013433,
                "source_pts_s": 22.013433,
                "image": "assets/frames/E6/f00000659.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 32.5,
            "range_s": [
              31.997344,
              32.999444
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 959,
                "time_s": 31.997344,
                "source_pts_s": 31.997344,
                "image": "assets/frames/E6/f00000959.jpg"
              },
              {
                "frame_index": 989,
                "time_s": 32.999444,
                "source_pts_s": 32.999444,
                "image": "assets/frames/E6/f00000989.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a06",
          "level": "action",
          "parent_id": "E6.s03",
          "start_s": 32.5,
          "end_s": 39.5,
          "label": "Draw the sheet across the near side",
          "verb": "draw",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: draw the sheet across the near side. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1019,
              "time_s": 33.998111,
              "source_pts_s": 33.998111,
              "image": "assets/frames/E6/f00001019.jpg"
            },
            {
              "frame_index": 1079,
              "time_s": 35.995556,
              "source_pts_s": 35.995556,
              "image": "assets/frames/E6/f00001079.jpg"
            },
            {
              "frame_index": 1139,
              "time_s": 37.993111,
              "source_pts_s": 37.993111,
              "image": "assets/frames/E6/f00001139.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 32.5,
            "range_s": [
              31.997344,
              32.999444
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 959,
                "time_s": 31.997344,
                "source_pts_s": 31.997344,
                "image": "assets/frames/E6/f00000959.jpg"
              },
              {
                "frame_index": 989,
                "time_s": 32.999444,
                "source_pts_s": 32.999444,
                "image": "assets/frames/E6/f00000989.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 39.5,
            "range_s": [
              38.991778,
              39.990556
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1169,
                "time_s": 38.991778,
                "source_pts_s": 38.991778,
                "image": "assets/frames/E6/f00001169.jpg"
              },
              {
                "frame_index": 1199,
                "time_s": 39.990556,
                "source_pts_s": 39.990556,
                "image": "assets/frames/E6/f00001199.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a07",
          "level": "action",
          "parent_id": "E6.s03",
          "start_s": 39.5,
          "end_s": 48.5,
          "label": "Gather and fit the sheet around the corner beside the pillows",
          "verb": "gather",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: gather and fit the sheet around the corner beside the pillows. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1199,
              "time_s": 39.990556,
              "source_pts_s": 39.990556,
              "image": "assets/frames/E6/f00001199.jpg"
            },
            {
              "frame_index": 1320,
              "time_s": 44.015478,
              "source_pts_s": 44.015478,
              "image": "assets/frames/E6/f00001320.jpg"
            },
            {
              "frame_index": 1409,
              "time_s": 47.011667,
              "source_pts_s": 47.011667,
              "image": "assets/frames/E6/f00001409.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 39.5,
            "range_s": [
              38.991778,
              39.990556
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1169,
                "time_s": 38.991778,
                "source_pts_s": 38.991778,
                "image": "assets/frames/E6/f00001169.jpg"
              },
              {
                "frame_index": 1199,
                "time_s": 39.990556,
                "source_pts_s": 39.990556,
                "image": "assets/frames/E6/f00001199.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 48.5,
            "range_s": [
              48.010333,
              49.012411
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1439,
                "time_s": 48.010333,
                "source_pts_s": 48.010333,
                "image": "assets/frames/E6/f00001439.jpg"
              },
              {
                "frame_index": 1469,
                "time_s": 49.012411,
                "source_pts_s": 49.012411,
                "image": "assets/frames/E6/f00001469.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a08",
          "level": "action",
          "parent_id": "E6.s03",
          "start_s": 48.5,
          "end_s": 52.5,
          "label": "Pull the adjacent hanging edge downward",
          "verb": "pull",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: pull the adjacent hanging edge downward. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1469,
              "time_s": 49.012411,
              "source_pts_s": 49.012411,
              "image": "assets/frames/E6/f00001469.jpg"
            },
            {
              "frame_index": 1499,
              "time_s": 50.011189,
              "source_pts_s": 50.011189,
              "image": "assets/frames/E6/f00001499.jpg"
            },
            {
              "frame_index": 1529,
              "time_s": 51.009856,
              "source_pts_s": 51.009856,
              "image": "assets/frames/E6/f00001529.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 48.5,
            "range_s": [
              48.010333,
              49.012411
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1439,
                "time_s": 48.010333,
                "source_pts_s": 48.010333,
                "image": "assets/frames/E6/f00001439.jpg"
              },
              {
                "frame_index": 1469,
                "time_s": 49.012411,
                "source_pts_s": 49.012411,
                "image": "assets/frames/E6/f00001469.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 52.5,
            "range_s": [
              52.008522,
              53.0073
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1559,
                "time_s": 52.008522,
                "source_pts_s": 52.008522,
                "image": "assets/frames/E6/f00001559.jpg"
              },
              {
                "frame_index": 1589,
                "time_s": 53.0073,
                "source_pts_s": 53.0073,
                "image": "assets/frames/E6/f00001589.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a09",
          "level": "action",
          "parent_id": "E6.s04",
          "start_s": 52.5,
          "end_s": 65.5,
          "label": "Move around the bed and stretch the sheet along the next edge",
          "verb": "move",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: move around the bed and stretch the sheet along the next edge. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1619,
              "time_s": 54.002722,
              "source_pts_s": 54.002722,
              "image": "assets/frames/E6/f00001619.jpg"
            },
            {
              "frame_index": 1769,
              "time_s": 58.996389,
              "source_pts_s": 58.996389,
              "image": "assets/frames/E6/f00001769.jpg"
            },
            {
              "frame_index": 1919,
              "time_s": 63.9933,
              "source_pts_s": 63.9933,
              "image": "assets/frames/E6/f00001919.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 52.5,
            "range_s": [
              52.008522,
              53.0073
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1559,
                "time_s": 52.008522,
                "source_pts_s": 52.008522,
                "image": "assets/frames/E6/f00001559.jpg"
              },
              {
                "frame_index": 1589,
                "time_s": 53.0073,
                "source_pts_s": 53.0073,
                "image": "assets/frames/E6/f00001589.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 65.5,
            "range_s": [
              64.992078,
              65.987433
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1949,
                "time_s": 64.992078,
                "source_pts_s": 64.992078,
                "image": "assets/frames/E6/f00001949.jpg"
              },
              {
                "frame_index": 1979,
                "time_s": 65.987433,
                "source_pts_s": 65.987433,
                "image": "assets/frames/E6/f00001979.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a10",
          "level": "action",
          "parent_id": "E6.s04",
          "start_s": 65.5,
          "end_s": 76.5,
          "label": "Pull and tuck the sheet around another mattress corner",
          "verb": "pull",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: pull and tuck the sheet around another mattress corner. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2009,
              "time_s": 66.986211,
              "source_pts_s": 66.986211,
              "image": "assets/frames/E6/f00002009.jpg"
            },
            {
              "frame_index": 2160,
              "time_s": 72.013056,
              "source_pts_s": 72.013056,
              "image": "assets/frames/E6/f00002160.jpg"
            },
            {
              "frame_index": 2250,
              "time_s": 74.999289,
              "source_pts_s": 74.999289,
              "image": "assets/frames/E6/f00002250.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 65.5,
            "range_s": [
              64.992078,
              65.987433
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1949,
                "time_s": 64.992078,
                "source_pts_s": 64.992078,
                "image": "assets/frames/E6/f00001949.jpg"
              },
              {
                "frame_index": 1979,
                "time_s": 65.987433,
                "source_pts_s": 65.987433,
                "image": "assets/frames/E6/f00001979.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 76.5,
            "range_s": [
              75.997956,
              77.006711
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2280,
                "time_s": 75.997956,
                "source_pts_s": 75.997956,
                "image": "assets/frames/E6/f00002280.jpg"
              },
              {
                "frame_index": 2310,
                "time_s": 77.006711,
                "source_pts_s": 77.006711,
                "image": "assets/frames/E6/f00002310.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a11",
          "level": "action",
          "parent_id": "E6.s04",
          "start_s": 76.5,
          "end_s": 80.5,
          "label": "Press and pull the corner seam into place",
          "verb": "press",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: press and pull the corner seam into place. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2310,
              "time_s": 77.006711,
              "source_pts_s": 77.006711,
              "image": "assets/frames/E6/f00002310.jpg"
            },
            {
              "frame_index": 2340,
              "time_s": 78.005489,
              "source_pts_s": 78.005489,
              "image": "assets/frames/E6/f00002340.jpg"
            },
            {
              "frame_index": 2370,
              "time_s": 79.004156,
              "source_pts_s": 79.004156,
              "image": "assets/frames/E6/f00002370.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 76.5,
            "range_s": [
              75.997956,
              77.006711
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2280,
                "time_s": 75.997956,
                "source_pts_s": 75.997956,
                "image": "assets/frames/E6/f00002280.jpg"
              },
              {
                "frame_index": 2310,
                "time_s": 77.006711,
                "source_pts_s": 77.006711,
                "image": "assets/frames/E6/f00002310.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 80.5,
            "range_s": [
              80.002933,
              81.0016
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2400,
                "time_s": 80.002933,
                "source_pts_s": 80.002933,
                "image": "assets/frames/E6/f00002400.jpg"
              },
              {
                "frame_index": 2430,
                "time_s": 81.0016,
                "source_pts_s": 81.0016,
                "image": "assets/frames/E6/f00002430.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a12",
          "level": "action",
          "parent_id": "E6.s05",
          "start_s": 80.5,
          "end_s": 90.5,
          "label": "Move along the foot of the bed while stretching the sheet",
          "verb": "move",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: move along the foot of the bed while stretching the sheet. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2490,
              "time_s": 82.999044,
              "source_pts_s": 82.999044,
              "image": "assets/frames/E6/f00002490.jpg"
            },
            {
              "frame_index": 2580,
              "time_s": 85.998622,
              "source_pts_s": 85.998622,
              "image": "assets/frames/E6/f00002580.jpg"
            },
            {
              "frame_index": 2670,
              "time_s": 88.991444,
              "source_pts_s": 88.991444,
              "image": "assets/frames/E6/f00002670.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 80.5,
            "range_s": [
              80.002933,
              81.0016
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2400,
                "time_s": 80.002933,
                "source_pts_s": 80.002933,
                "image": "assets/frames/E6/f00002400.jpg"
              },
              {
                "frame_index": 2430,
                "time_s": 81.0016,
                "source_pts_s": 81.0016,
                "image": "assets/frames/E6/f00002430.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 90.5,
            "range_s": [
              89.990111,
              90.988889
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2700,
                "time_s": 89.990111,
                "source_pts_s": 89.990111,
                "image": "assets/frames/E6/f00002700.jpg"
              },
              {
                "frame_index": 2730,
                "time_s": 90.988889,
                "source_pts_s": 90.988889,
                "image": "assets/frames/E6/f00002730.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a13",
          "level": "action",
          "parent_id": "E6.s05",
          "start_s": 90.5,
          "end_s": 99.5,
          "label": "Tuck and adjust the side edge beside the bedding pile",
          "verb": "tuck",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: tuck and adjust the side edge beside the bedding pile. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2790,
              "time_s": 92.986333,
              "source_pts_s": 92.986333,
              "image": "assets/frames/E6/f00002790.jpg"
            },
            {
              "frame_index": 2881,
              "time_s": 96.015733,
              "source_pts_s": 96.015733,
              "image": "assets/frames/E6/f00002881.jpg"
            },
            {
              "frame_index": 2941,
              "time_s": 98.013289,
              "source_pts_s": 98.013289,
              "image": "assets/frames/E6/f00002941.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 90.5,
            "range_s": [
              89.990111,
              90.988889
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2700,
                "time_s": 89.990111,
                "source_pts_s": 89.990111,
                "image": "assets/frames/E6/f00002700.jpg"
              },
              {
                "frame_index": 2730,
                "time_s": 90.988889,
                "source_pts_s": 90.988889,
                "image": "assets/frames/E6/f00002730.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 99.5,
            "range_s": [
              99.011956,
              100.010733
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2971,
                "time_s": 99.011956,
                "source_pts_s": 99.011956,
                "image": "assets/frames/E6/f00002971.jpg"
              },
              {
                "frame_index": 3001,
                "time_s": 100.010733,
                "source_pts_s": 100.010733,
                "image": "assets/frames/E6/f00003001.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a14",
          "level": "action",
          "parent_id": "E6.s05",
          "start_s": 99.5,
          "end_s": 103.5,
          "label": "Pull the lower corner and move along the bed",
          "verb": "pull",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: pull the lower corner and move along the bed. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 3001,
              "time_s": 100.010733,
              "source_pts_s": 100.010733,
              "image": "assets/frames/E6/f00003001.jpg"
            },
            {
              "frame_index": 3031,
              "time_s": 101.0094,
              "source_pts_s": 101.0094,
              "image": "assets/frames/E6/f00003031.jpg"
            },
            {
              "frame_index": 3061,
              "time_s": 102.008178,
              "source_pts_s": 102.008178,
              "image": "assets/frames/E6/f00003061.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 99.5,
            "range_s": [
              99.011956,
              100.010733
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2971,
                "time_s": 99.011956,
                "source_pts_s": 99.011956,
                "image": "assets/frames/E6/f00002971.jpg"
              },
              {
                "frame_index": 3001,
                "time_s": 100.010733,
                "source_pts_s": 100.010733,
                "image": "assets/frames/E6/f00003001.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 103.5,
            "range_s": [
              103.006878,
              104.005544
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 3090,
                "time_s": 103.006878,
                "source_pts_s": 103.006878,
                "image": "assets/frames/E6/f00003090.jpg"
              },
              {
                "frame_index": 3120,
                "time_s": 104.005544,
                "source_pts_s": 104.005544,
                "image": "assets/frames/E6/f00003120.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a15",
          "level": "action",
          "parent_id": "E6.s06",
          "start_s": 105.5,
          "end_s": 111.5,
          "label": "Push the sheet around the wall-side corner and smooth the top",
          "verb": "push",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: push the sheet around the wall-side corner and smooth the top. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 3180,
              "time_s": 106.002989,
              "source_pts_s": 106.002989,
              "image": "assets/frames/E6/f00003180.jpg"
            },
            {
              "frame_index": 3270,
              "time_s": 108.994178,
              "source_pts_s": 108.994178,
              "image": "assets/frames/E6/f00003270.jpg"
            },
            {
              "frame_index": 3300,
              "time_s": 109.992844,
              "source_pts_s": 109.992844,
              "image": "assets/frames/E6/f00003300.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 105.5,
            "range_s": [
              105.004322,
              106.002989
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 3150,
                "time_s": 105.004322,
                "source_pts_s": 105.004322,
                "image": "assets/frames/E6/f00003150.jpg"
              },
              {
                "frame_index": 3180,
                "time_s": 106.002989,
                "source_pts_s": 106.002989,
                "image": "assets/frames/E6/f00003180.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 111.5,
            "range_s": [
              110.996644,
              111.995311
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 3330,
                "time_s": 110.996644,
                "source_pts_s": 110.996644,
                "image": "assets/frames/E6/f00003330.jpg"
              },
              {
                "frame_index": 3360,
                "time_s": 111.995311,
                "source_pts_s": 111.995311,
                "image": "assets/frames/E6/f00003360.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "E6.a16",
          "level": "action",
          "parent_id": "E6.s06",
          "start_s": 113.5,
          "end_s": 115.890378,
          "label": "Make another short pull at the near edge",
          "verb": "make",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "E6.sheet",
            "E6.mattress"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: make another short pull at the near edge. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 3419,
              "time_s": 113.9928,
              "source_pts_s": 113.9928,
              "image": "assets/frames/E6/f00003419.jpg"
            },
            {
              "frame_index": 3449,
              "time_s": 114.991467,
              "source_pts_s": 114.991467,
              "image": "assets/frames/E6/f00003449.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 113.5,
            "range_s": [
              112.994089,
              113.9928
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 3390,
                "time_s": 112.994089,
                "source_pts_s": 112.994089,
                "image": "assets/frames/E6/f00003390.jpg"
              },
              {
                "frame_index": 3419,
                "time_s": 113.9928,
                "source_pts_s": 113.9928,
                "image": "assets/frames/E6/f00003419.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 115.890378,
            "range_s": [
              115.890378,
              115.890378
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 3475,
                "time_s": 115.857089,
                "source_pts_s": 115.857089,
                "image": "assets/frames/E6/f00003475.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "sampling": {
        "base_frames_extracted": 117,
        "base_frames_visually_reviewed": 117,
        "dense_frames_extracted": 26,
        "dense_frames_visually_reviewed": 26,
        "unique_frames_visually_reviewed": 138,
        "source_frame_count": 3476
      },
      "playback": {
        "sha256": "f293e148ebddcf808640c3c410ba10d49f39acca53483161ad92321a5eff2012",
        "bytes": 115996209,
        "playback_width": 1280,
        "playback_height": 720,
        "source_width": 1280,
        "source_height": 720,
        "frame_count": 3476,
        "max_timestamp_error_s": 0.0,
        "duration_s": 115.890378,
        "audio_stream_count": 0,
        "playback_time_offset_s": 0,
        "kind": "delivery_encode_served_from_r2",
        "source_sha256": "e8ae83f772fda809a962e40344096335d46d3c96f7ff70aef7d46330eac7fde7",
        "encoding": "h264-native-2pass-8mbps-relative-pts-v1",
        "url": "https://media-demo.theroboticsstack.com/media/E6.mp4?v=f293e148ebdd"
      }
    },
    {
      "id": "M3",
      "source": "https://media-demo.theroboticsstack.com/media/M3.mp4?v=a283cddd55ae",
      "sha256": "a283cddd55ae93c659922c18943c94bd307d45aa2c47126995e0ff2d243916d0",
      "duration_s": 87.933333,
      "width": 640,
      "height": 360,
      "fps_fraction": "30/1",
      "decoded_frame_count": 2638,
      "audio_stream_count": 0,
      "source_start_pts_s": 0.0,
      "last_frame_s": 87.9,
      "frame_interval_min_s": 0.03333299999999895,
      "frame_interval_max_s": 0.03333400000001063,
      "nonincreasing_pts_count": 0,
      "decode_diagnostics": "",
      "view_layout": "single_egocentric_full_frame",
      "sample_interval_target_s": 1.0,
      "extraction_settings": {
        "maximum_image_width": 1280,
        "crop": null,
        "frame_selection": "nearest_actual_PTS"
      },
      "title": "Clean and pack two rackets",
      "observed_activity": "Open a purple racket bag, wipe a red-framed racket and then a dark-framed racket with a blue cloth, align both rackets and return them to the bag.",
      "inferred_goal": "Wipe the rackets and pack them together.",
      "assigned_instruction": null,
      "playback_time_offset_s": 0.0,
      "visible_outcome": "Both rackets enter the purple bag, which is closed and lowered toward the chair.",
      "task_success": "unknown",
      "unknowns": [
        "Two rackets are visible despite the singular wording in the publisher description.",
        "No particular racket sport or cleanliness standard is inferred.",
        "Small zipper and grip details are limited by the preview resolution."
      ],
      "selection_rationale": "Featured: clear unpack, clean, exchange, clean and repack structure, including an object-count correction.",
      "annotation_provenance": "assistant_visual_review",
      "review_status": "pending",
      "credit": {
        "attribution": "Mecka AI \u00b7 Egoverse \u00b7 recording 69bb0ec4a86b4425674ca0e3",
        "source_url": "https://partners.mecka.ai/api/egoverse/uploads/69bb0ec4a86b4425674ca0e3/video?redirect=1",
        "metadata_url": "https://partners.mecka.ai/egoverse?embodiment=human_bimanual&page=2",
        "license": "CC BY-SA 4.0",
        "license_url": "https://creativecommons.org/licenses/by-sa/4.0/",
        "changes": "Complete source published unchanged from object storage; no re-encoding. Evidence resized to at most 1280px; assistant temporal labels added."
      },
      "protocol_role": "featured",
      "objects": [
        {
          "id": "M3.red",
          "description": "Red-framed racket",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M3.dark",
          "description": "Dark-framed racket with orange-toned strings",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M3.cloth",
          "description": "Blue cleaning cloth",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M3.bag",
          "description": "Purple racket bag",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M3.chair",
          "description": "Chair supporting the bag",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        }
      ],
      "subtasks": [
        {
          "id": "M3.s01",
          "level": "subtask",
          "parent_id": null,
          "start_s": 0,
          "end_s": 18.5,
          "label": "Open the bag and retrieve the first racket",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M3.bag",
            "M3.chair",
            "M3.cloth",
            "M3.red"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: open the bag and retrieve the first racket. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 0,
              "time_s": 0.0,
              "source_pts_s": 0.0,
              "image": "assets/frames/M3/f00000000.jpg"
            },
            {
              "frame_index": 270,
              "time_s": 9.0,
              "source_pts_s": 9.0,
              "image": "assets/frames/M3/f00000270.jpg"
            },
            {
              "frame_index": 540,
              "time_s": 18.0,
              "source_pts_s": 18.0,
              "image": "assets/frames/M3/f00000540.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.0,
                "image": "assets/frames/M3/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 18.5,
            "range_s": [
              18.0,
              19.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 540,
                "time_s": 18.0,
                "source_pts_s": 18.0,
                "image": "assets/frames/M3/f00000540.jpg"
              },
              {
                "frame_index": 570,
                "time_s": 19.0,
                "source_pts_s": 19.0,
                "image": "assets/frames/M3/f00000570.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.s02",
          "level": "subtask",
          "parent_id": null,
          "start_s": 18.5,
          "end_s": 39.5,
          "label": "Wipe the red-framed racket",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M3.cloth",
            "M3.red"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: wipe the red-framed racket. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 540,
              "time_s": 18.0,
              "source_pts_s": 18.0,
              "image": "assets/frames/M3/f00000540.jpg"
            },
            {
              "frame_index": 870,
              "time_s": 29.0,
              "source_pts_s": 29.0,
              "image": "assets/frames/M3/f00000870.jpg"
            },
            {
              "frame_index": 1170,
              "time_s": 39.0,
              "source_pts_s": 39.0,
              "image": "assets/frames/M3/f00001170.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 18.5,
            "range_s": [
              18.0,
              19.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 540,
                "time_s": 18.0,
                "source_pts_s": 18.0,
                "image": "assets/frames/M3/f00000540.jpg"
              },
              {
                "frame_index": 570,
                "time_s": 19.0,
                "source_pts_s": 19.0,
                "image": "assets/frames/M3/f00000570.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 39.5,
            "range_s": [
              39.0,
              40.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1170,
                "time_s": 39.0,
                "source_pts_s": 39.0,
                "image": "assets/frames/M3/f00001170.jpg"
              },
              {
                "frame_index": 1200,
                "time_s": 40.0,
                "source_pts_s": 40.0,
                "image": "assets/frames/M3/f00001200.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.s03",
          "level": "subtask",
          "parent_id": null,
          "start_s": 39.5,
          "end_s": 44.5,
          "label": "Exchange rackets",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M3.bag",
            "M3.chair",
            "M3.dark",
            "M3.red"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: exchange rackets. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1170,
              "time_s": 39.0,
              "source_pts_s": 39.0,
              "image": "assets/frames/M3/f00001170.jpg"
            },
            {
              "frame_index": 1260,
              "time_s": 42.0,
              "source_pts_s": 42.0,
              "image": "assets/frames/M3/f00001260.jpg"
            },
            {
              "frame_index": 1320,
              "time_s": 44.0,
              "source_pts_s": 44.0,
              "image": "assets/frames/M3/f00001320.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 39.5,
            "range_s": [
              39.0,
              40.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1170,
                "time_s": 39.0,
                "source_pts_s": 39.0,
                "image": "assets/frames/M3/f00001170.jpg"
              },
              {
                "frame_index": 1200,
                "time_s": 40.0,
                "source_pts_s": 40.0,
                "image": "assets/frames/M3/f00001200.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 44.5,
            "range_s": [
              44.0,
              45.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1320,
                "time_s": 44.0,
                "source_pts_s": 44.0,
                "image": "assets/frames/M3/f00001320.jpg"
              },
              {
                "frame_index": 1350,
                "time_s": 45.0,
                "source_pts_s": 45.0,
                "image": "assets/frames/M3/f00001350.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.s04",
          "level": "subtask",
          "parent_id": null,
          "start_s": 44.5,
          "end_s": 62.5,
          "label": "Wipe the dark-framed racket",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M3.cloth",
            "M3.dark"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: wipe the dark-framed racket. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1320,
              "time_s": 44.0,
              "source_pts_s": 44.0,
              "image": "assets/frames/M3/f00001320.jpg"
            },
            {
              "frame_index": 1590,
              "time_s": 53.0,
              "source_pts_s": 53.0,
              "image": "assets/frames/M3/f00001590.jpg"
            },
            {
              "frame_index": 1860,
              "time_s": 62.0,
              "source_pts_s": 62.0,
              "image": "assets/frames/M3/f00001860.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 44.5,
            "range_s": [
              44.0,
              45.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1320,
                "time_s": 44.0,
                "source_pts_s": 44.0,
                "image": "assets/frames/M3/f00001320.jpg"
              },
              {
                "frame_index": 1350,
                "time_s": 45.0,
                "source_pts_s": 45.0,
                "image": "assets/frames/M3/f00001350.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 62.5,
            "range_s": [
              62.0,
              63.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1860,
                "time_s": 62.0,
                "source_pts_s": 62.0,
                "image": "assets/frames/M3/f00001860.jpg"
              },
              {
                "frame_index": 1890,
                "time_s": 63.0,
                "source_pts_s": 63.0,
                "image": "assets/frames/M3/f00001890.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.s05",
          "level": "subtask",
          "parent_id": null,
          "start_s": 62.5,
          "end_s": 78.5,
          "label": "Gather and align both rackets",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M3.cloth",
            "M3.dark",
            "M3.red"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: gather and align both rackets. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1860,
              "time_s": 62.0,
              "source_pts_s": 62.0,
              "image": "assets/frames/M3/f00001860.jpg"
            },
            {
              "frame_index": 2100,
              "time_s": 70.0,
              "source_pts_s": 70.0,
              "image": "assets/frames/M3/f00002100.jpg"
            },
            {
              "frame_index": 2340,
              "time_s": 78.0,
              "source_pts_s": 78.0,
              "image": "assets/frames/M3/f00002340.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 62.5,
            "range_s": [
              62.0,
              63.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1860,
                "time_s": 62.0,
                "source_pts_s": 62.0,
                "image": "assets/frames/M3/f00001860.jpg"
              },
              {
                "frame_index": 1890,
                "time_s": 63.0,
                "source_pts_s": 63.0,
                "image": "assets/frames/M3/f00001890.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 78.5,
            "range_s": [
              78.0,
              79.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2340,
                "time_s": 78.0,
                "source_pts_s": 78.0,
                "image": "assets/frames/M3/f00002340.jpg"
              },
              {
                "frame_index": 2370,
                "time_s": 79.0,
                "source_pts_s": 79.0,
                "image": "assets/frames/M3/f00002370.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.s06",
          "level": "subtask",
          "parent_id": null,
          "start_s": 78.5,
          "end_s": 87.933333,
          "label": "Return both rackets to the bag",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M3.bag",
            "M3.chair",
            "M3.dark",
            "M3.red"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: return both rackets to the bag. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2340,
              "time_s": 78.0,
              "source_pts_s": 78.0,
              "image": "assets/frames/M3/f00002340.jpg"
            },
            {
              "frame_index": 2490,
              "time_s": 83.0,
              "source_pts_s": 83.0,
              "image": "assets/frames/M3/f00002490.jpg"
            },
            {
              "frame_index": 2637,
              "time_s": 87.9,
              "source_pts_s": 87.9,
              "image": "assets/frames/M3/f00002637.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 78.5,
            "range_s": [
              78.0,
              79.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2340,
                "time_s": 78.0,
                "source_pts_s": 78.0,
                "image": "assets/frames/M3/f00002340.jpg"
              },
              {
                "frame_index": 2370,
                "time_s": 79.0,
                "source_pts_s": 79.0,
                "image": "assets/frames/M3/f00002370.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 87.933333,
            "range_s": [
              87.933333,
              87.933333
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 2637,
                "time_s": 87.9,
                "source_pts_s": 87.9,
                "image": "assets/frames/M3/f00002637.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "actions": [
        {
          "id": "M3.a01",
          "level": "action",
          "parent_id": "M3.s01",
          "start_s": 0,
          "end_s": 7.5,
          "label": "Approach the chair and move the blue cloth aside",
          "verb": "approach",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M3.chair",
            "M3.cloth"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: approach the chair and move the blue cloth aside. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 30,
              "time_s": 1.0,
              "source_pts_s": 1.0,
              "image": "assets/frames/M3/f00000030.jpg"
            },
            {
              "frame_index": 120,
              "time_s": 4.0,
              "source_pts_s": 4.0,
              "image": "assets/frames/M3/f00000120.jpg"
            },
            {
              "frame_index": 210,
              "time_s": 7.0,
              "source_pts_s": 7.0,
              "image": "assets/frames/M3/f00000210.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.0,
                "image": "assets/frames/M3/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 7.5,
            "range_s": [
              7.0,
              8.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 210,
                "time_s": 7.0,
                "source_pts_s": 7.0,
                "image": "assets/frames/M3/f00000210.jpg"
              },
              {
                "frame_index": 240,
                "time_s": 8.0,
                "source_pts_s": 8.0,
                "image": "assets/frames/M3/f00000240.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a02",
          "level": "action",
          "parent_id": "M3.s01",
          "start_s": 7.5,
          "end_s": 14.5,
          "label": "Lift and open the purple racket bag",
          "verb": "lift",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M3.bag"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: lift and open the purple racket bag. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 240,
              "time_s": 8.0,
              "source_pts_s": 8.0,
              "image": "assets/frames/M3/f00000240.jpg"
            },
            {
              "frame_index": 330,
              "time_s": 11.0,
              "source_pts_s": 11.0,
              "image": "assets/frames/M3/f00000330.jpg"
            },
            {
              "frame_index": 420,
              "time_s": 14.0,
              "source_pts_s": 14.0,
              "image": "assets/frames/M3/f00000420.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 7.5,
            "range_s": [
              7.0,
              8.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 210,
                "time_s": 7.0,
                "source_pts_s": 7.0,
                "image": "assets/frames/M3/f00000210.jpg"
              },
              {
                "frame_index": 240,
                "time_s": 8.0,
                "source_pts_s": 8.0,
                "image": "assets/frames/M3/f00000240.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 14.5,
            "range_s": [
              14.0,
              15.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 420,
                "time_s": 14.0,
                "source_pts_s": 14.0,
                "image": "assets/frames/M3/f00000420.jpg"
              },
              {
                "frame_index": 450,
                "time_s": 15.0,
                "source_pts_s": 15.0,
                "image": "assets/frames/M3/f00000450.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a03",
          "level": "action",
          "parent_id": "M3.s01",
          "start_s": 14.5,
          "end_s": 18.5,
          "label": "Pull out the red-framed racket and retrieve the cloth",
          "verb": "pull",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M3.red",
            "M3.bag",
            "M3.cloth"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: pull out the red-framed racket and retrieve the cloth. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 450,
              "time_s": 15.0,
              "source_pts_s": 15.0,
              "image": "assets/frames/M3/f00000450.jpg"
            },
            {
              "frame_index": 510,
              "time_s": 17.0,
              "source_pts_s": 17.0,
              "image": "assets/frames/M3/f00000510.jpg"
            },
            {
              "frame_index": 540,
              "time_s": 18.0,
              "source_pts_s": 18.0,
              "image": "assets/frames/M3/f00000540.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 14.5,
            "range_s": [
              14.0,
              15.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 420,
                "time_s": 14.0,
                "source_pts_s": 14.0,
                "image": "assets/frames/M3/f00000420.jpg"
              },
              {
                "frame_index": 450,
                "time_s": 15.0,
                "source_pts_s": 15.0,
                "image": "assets/frames/M3/f00000450.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 18.5,
            "range_s": [
              18.0,
              19.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 540,
                "time_s": 18.0,
                "source_pts_s": 18.0,
                "image": "assets/frames/M3/f00000540.jpg"
              },
              {
                "frame_index": 570,
                "time_s": 19.0,
                "source_pts_s": 19.0,
                "image": "assets/frames/M3/f00000570.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a04",
          "level": "action",
          "parent_id": "M3.s02",
          "start_s": 18.5,
          "end_s": 25.5,
          "label": "Wipe the first racket shaft, throat and frame",
          "verb": "wipe",
          "hands": "right",
          "hand_roles": {
            "right": "Wipe the first racket shaft, throat and frame"
          },
          "object_ids": [
            "M3.red",
            "M3.cloth"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: wipe the first racket shaft, throat and frame. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 600,
              "time_s": 20.0,
              "source_pts_s": 20.0,
              "image": "assets/frames/M3/f00000600.jpg"
            },
            {
              "frame_index": 690,
              "time_s": 23.0,
              "source_pts_s": 23.0,
              "image": "assets/frames/M3/f00000690.jpg"
            },
            {
              "frame_index": 750,
              "time_s": 25.0,
              "source_pts_s": 25.0,
              "image": "assets/frames/M3/f00000750.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 18.5,
            "range_s": [
              18.0,
              19.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 540,
                "time_s": 18.0,
                "source_pts_s": 18.0,
                "image": "assets/frames/M3/f00000540.jpg"
              },
              {
                "frame_index": 570,
                "time_s": 19.0,
                "source_pts_s": 19.0,
                "image": "assets/frames/M3/f00000570.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 25.5,
            "range_s": [
              25.0,
              26.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 750,
                "time_s": 25.0,
                "source_pts_s": 25.0,
                "image": "assets/frames/M3/f00000750.jpg"
              },
              {
                "frame_index": 780,
                "time_s": 26.0,
                "source_pts_s": 26.0,
                "image": "assets/frames/M3/f00000780.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a05",
          "level": "action",
          "parent_id": "M3.s02",
          "start_s": 25.5,
          "end_s": 32.5,
          "label": "Wipe across the first racket strings while turning it",
          "verb": "wipe",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M3.red",
            "M3.cloth"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: wipe across the first racket strings while turning it. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 780,
              "time_s": 26.0,
              "source_pts_s": 26.0,
              "image": "assets/frames/M3/f00000780.jpg"
            },
            {
              "frame_index": 870,
              "time_s": 29.0,
              "source_pts_s": 29.0,
              "image": "assets/frames/M3/f00000870.jpg"
            },
            {
              "frame_index": 960,
              "time_s": 32.0,
              "source_pts_s": 32.0,
              "image": "assets/frames/M3/f00000960.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 25.5,
            "range_s": [
              25.0,
              26.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 750,
                "time_s": 25.0,
                "source_pts_s": 25.0,
                "image": "assets/frames/M3/f00000750.jpg"
              },
              {
                "frame_index": 780,
                "time_s": 26.0,
                "source_pts_s": 26.0,
                "image": "assets/frames/M3/f00000780.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 32.5,
            "range_s": [
              32.0,
              33.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 960,
                "time_s": 32.0,
                "source_pts_s": 32.0,
                "image": "assets/frames/M3/f00000960.jpg"
              },
              {
                "frame_index": 990,
                "time_s": 33.0,
                "source_pts_s": 33.0,
                "image": "assets/frames/M3/f00000990.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a06",
          "level": "action",
          "parent_id": "M3.s02",
          "start_s": 32.5,
          "end_s": 39.5,
          "label": "Wipe the first racket shaft and grip",
          "verb": "wipe",
          "hands": "right",
          "hand_roles": {
            "right": "Wipe the first racket shaft and grip"
          },
          "object_ids": [
            "M3.red",
            "M3.cloth"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: wipe the first racket shaft and grip. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 990,
              "time_s": 33.0,
              "source_pts_s": 33.0,
              "image": "assets/frames/M3/f00000990.jpg"
            },
            {
              "frame_index": 1080,
              "time_s": 36.0,
              "source_pts_s": 36.0,
              "image": "assets/frames/M3/f00001080.jpg"
            },
            {
              "frame_index": 1140,
              "time_s": 38.0,
              "source_pts_s": 38.0,
              "image": "assets/frames/M3/f00001140.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 32.5,
            "range_s": [
              32.0,
              33.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 960,
                "time_s": 32.0,
                "source_pts_s": 32.0,
                "image": "assets/frames/M3/f00000960.jpg"
              },
              {
                "frame_index": 990,
                "time_s": 33.0,
                "source_pts_s": 33.0,
                "image": "assets/frames/M3/f00000990.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 39.5,
            "range_s": [
              39.0,
              40.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1170,
                "time_s": 39.0,
                "source_pts_s": 39.0,
                "image": "assets/frames/M3/f00001170.jpg"
              },
              {
                "frame_index": 1200,
                "time_s": 40.0,
                "source_pts_s": 40.0,
                "image": "assets/frames/M3/f00001200.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a07",
          "level": "action",
          "parent_id": "M3.s03",
          "start_s": 39.5,
          "end_s": 40.866667,
          "label": "Rest the red-framed racket against the bag and chair",
          "verb": "rest",
          "hands": "left",
          "hand_roles": {
            "left": "Rest the red-framed racket against the bag and chair"
          },
          "object_ids": [
            "M3.red",
            "M3.bag",
            "M3.chair"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: rest the red-framed racket against the bag and chair. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1200,
              "time_s": 40.0,
              "source_pts_s": 40.0,
              "image": "assets/frames/M3/f00001200.jpg"
            },
            {
              "frame_index": 1230,
              "time_s": 41.0,
              "source_pts_s": 41.0,
              "image": "assets/frames/M3/f00001230.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 39.5,
            "range_s": [
              39.0,
              40.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1170,
                "time_s": 39.0,
                "source_pts_s": 39.0,
                "image": "assets/frames/M3/f00001170.jpg"
              },
              {
                "frame_index": 1200,
                "time_s": 40.0,
                "source_pts_s": 40.0,
                "image": "assets/frames/M3/f00001200.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 40.866667,
            "range_s": [
              40.733333,
              41.0
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Bag flap opens to expose the second dark-framed racket; start of retrieval phase.",
            "evidence_frames": [
              {
                "frame_index": 1222,
                "time_s": 40.733333,
                "source_pts_s": 40.733333,
                "image": "assets/frames/M3/f00001222.jpg"
              },
              {
                "frame_index": 1230,
                "time_s": 41.0,
                "source_pts_s": 41.0,
                "image": "assets/frames/M3/f00001230.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a08",
          "level": "action",
          "parent_id": "M3.s03",
          "start_s": 40.866667,
          "end_s": 44.5,
          "label": "Reach into the bag and extract the dark-framed racket",
          "verb": "reach",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M3.dark",
            "M3.bag"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: reach into the bag and extract the dark-framed racket. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1260,
              "time_s": 42.0,
              "source_pts_s": 42.0,
              "image": "assets/frames/M3/f00001260.jpg"
            },
            {
              "frame_index": 1290,
              "time_s": 43.0,
              "source_pts_s": 43.0,
              "image": "assets/frames/M3/f00001290.jpg"
            },
            {
              "frame_index": 1320,
              "time_s": 44.0,
              "source_pts_s": 44.0,
              "image": "assets/frames/M3/f00001320.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 40.866667,
            "range_s": [
              40.733333,
              41.0
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Bag flap opens to expose the second dark-framed racket; start of retrieval phase.",
            "evidence_frames": [
              {
                "frame_index": 1222,
                "time_s": 40.733333,
                "source_pts_s": 40.733333,
                "image": "assets/frames/M3/f00001222.jpg"
              },
              {
                "frame_index": 1230,
                "time_s": 41.0,
                "source_pts_s": 41.0,
                "image": "assets/frames/M3/f00001230.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 44.5,
            "range_s": [
              44.0,
              45.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1320,
                "time_s": 44.0,
                "source_pts_s": 44.0,
                "image": "assets/frames/M3/f00001320.jpg"
              },
              {
                "frame_index": 1350,
                "time_s": 45.0,
                "source_pts_s": 45.0,
                "image": "assets/frames/M3/f00001350.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a09",
          "level": "action",
          "parent_id": "M3.s04",
          "start_s": 44.5,
          "end_s": 52.5,
          "label": "Wipe the second racket shaft and frame",
          "verb": "wipe",
          "hands": "right",
          "hand_roles": {
            "right": "Wipe the second racket shaft and frame"
          },
          "object_ids": [
            "M3.dark",
            "M3.cloth"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: wipe the second racket shaft and frame. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1350,
              "time_s": 45.0,
              "source_pts_s": 45.0,
              "image": "assets/frames/M3/f00001350.jpg"
            },
            {
              "frame_index": 1470,
              "time_s": 49.0,
              "source_pts_s": 49.0,
              "image": "assets/frames/M3/f00001470.jpg"
            },
            {
              "frame_index": 1560,
              "time_s": 52.0,
              "source_pts_s": 52.0,
              "image": "assets/frames/M3/f00001560.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 44.5,
            "range_s": [
              44.0,
              45.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1320,
                "time_s": 44.0,
                "source_pts_s": 44.0,
                "image": "assets/frames/M3/f00001320.jpg"
              },
              {
                "frame_index": 1350,
                "time_s": 45.0,
                "source_pts_s": 45.0,
                "image": "assets/frames/M3/f00001350.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 52.5,
            "range_s": [
              52.0,
              53.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1560,
                "time_s": 52.0,
                "source_pts_s": 52.0,
                "image": "assets/frames/M3/f00001560.jpg"
              },
              {
                "frame_index": 1590,
                "time_s": 53.0,
                "source_pts_s": 53.0,
                "image": "assets/frames/M3/f00001590.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a10",
          "level": "action",
          "parent_id": "M3.s04",
          "start_s": 52.5,
          "end_s": 59.5,
          "label": "Wipe the second racket strings",
          "verb": "wipe",
          "hands": "right",
          "hand_roles": {
            "right": "Wipe the second racket strings"
          },
          "object_ids": [
            "M3.dark",
            "M3.cloth"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: wipe the second racket strings. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1590,
              "time_s": 53.0,
              "source_pts_s": 53.0,
              "image": "assets/frames/M3/f00001590.jpg"
            },
            {
              "frame_index": 1680,
              "time_s": 56.0,
              "source_pts_s": 56.0,
              "image": "assets/frames/M3/f00001680.jpg"
            },
            {
              "frame_index": 1770,
              "time_s": 59.0,
              "source_pts_s": 59.0,
              "image": "assets/frames/M3/f00001770.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 52.5,
            "range_s": [
              52.0,
              53.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1560,
                "time_s": 52.0,
                "source_pts_s": 52.0,
                "image": "assets/frames/M3/f00001560.jpg"
              },
              {
                "frame_index": 1590,
                "time_s": 53.0,
                "source_pts_s": 53.0,
                "image": "assets/frames/M3/f00001590.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 59.5,
            "range_s": [
              59.0,
              60.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1770,
                "time_s": 59.0,
                "source_pts_s": 59.0,
                "image": "assets/frames/M3/f00001770.jpg"
              },
              {
                "frame_index": 1800,
                "time_s": 60.0,
                "source_pts_s": 60.0,
                "image": "assets/frames/M3/f00001800.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a11",
          "level": "action",
          "parent_id": "M3.s04",
          "start_s": 59.5,
          "end_s": 62.5,
          "label": "Wipe the second racket grip",
          "verb": "wipe",
          "hands": "right",
          "hand_roles": {
            "right": "Wipe the second racket grip"
          },
          "object_ids": [
            "M3.dark",
            "M3.cloth"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: wipe the second racket grip. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1800,
              "time_s": 60.0,
              "source_pts_s": 60.0,
              "image": "assets/frames/M3/f00001800.jpg"
            },
            {
              "frame_index": 1830,
              "time_s": 61.0,
              "source_pts_s": 61.0,
              "image": "assets/frames/M3/f00001830.jpg"
            },
            {
              "frame_index": 1860,
              "time_s": 62.0,
              "source_pts_s": 62.0,
              "image": "assets/frames/M3/f00001860.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 59.5,
            "range_s": [
              59.0,
              60.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1770,
                "time_s": 59.0,
                "source_pts_s": 59.0,
                "image": "assets/frames/M3/f00001770.jpg"
              },
              {
                "frame_index": 1800,
                "time_s": 60.0,
                "source_pts_s": 60.0,
                "image": "assets/frames/M3/f00001800.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 62.5,
            "range_s": [
              62.0,
              63.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1860,
                "time_s": 62.0,
                "source_pts_s": 62.0,
                "image": "assets/frames/M3/f00001860.jpg"
              },
              {
                "frame_index": 1890,
                "time_s": 63.0,
                "source_pts_s": 63.0,
                "image": "assets/frames/M3/f00001890.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a12",
          "level": "action",
          "parent_id": "M3.s05",
          "start_s": 62.5,
          "end_s": 65.5,
          "label": "Put the cloth aside and pick up the first racket",
          "verb": "put",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M3.cloth",
            "M3.red",
            "M3.dark"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: put the cloth aside and pick up the first racket. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1890,
              "time_s": 63.0,
              "source_pts_s": 63.0,
              "image": "assets/frames/M3/f00001890.jpg"
            },
            {
              "frame_index": 1920,
              "time_s": 64.0,
              "source_pts_s": 64.0,
              "image": "assets/frames/M3/f00001920.jpg"
            },
            {
              "frame_index": 1950,
              "time_s": 65.0,
              "source_pts_s": 65.0,
              "image": "assets/frames/M3/f00001950.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 62.5,
            "range_s": [
              62.0,
              63.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1860,
                "time_s": 62.0,
                "source_pts_s": 62.0,
                "image": "assets/frames/M3/f00001860.jpg"
              },
              {
                "frame_index": 1890,
                "time_s": 63.0,
                "source_pts_s": 63.0,
                "image": "assets/frames/M3/f00001890.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 65.5,
            "range_s": [
              65.0,
              66.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1950,
                "time_s": 65.0,
                "source_pts_s": 65.0,
                "image": "assets/frames/M3/f00001950.jpg"
              },
              {
                "frame_index": 1980,
                "time_s": 66.0,
                "source_pts_s": 66.0,
                "image": "assets/frames/M3/f00001980.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a13",
          "level": "action",
          "parent_id": "M3.s05",
          "start_s": 65.5,
          "end_s": 78.5,
          "label": "Bring the two rackets together and align their heads and grips",
          "verb": "bring",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M3.red",
            "M3.dark"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: bring the two rackets together and align their heads and grips. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2010,
              "time_s": 67.0,
              "source_pts_s": 67.0,
              "image": "assets/frames/M3/f00002010.jpg"
            },
            {
              "frame_index": 2190,
              "time_s": 73.0,
              "source_pts_s": 73.0,
              "image": "assets/frames/M3/f00002190.jpg"
            },
            {
              "frame_index": 2340,
              "time_s": 78.0,
              "source_pts_s": 78.0,
              "image": "assets/frames/M3/f00002340.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 65.5,
            "range_s": [
              65.0,
              66.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1950,
                "time_s": 65.0,
                "source_pts_s": 65.0,
                "image": "assets/frames/M3/f00001950.jpg"
              },
              {
                "frame_index": 1980,
                "time_s": 66.0,
                "source_pts_s": 66.0,
                "image": "assets/frames/M3/f00001980.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 78.5,
            "range_s": [
              78.0,
              79.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2340,
                "time_s": 78.0,
                "source_pts_s": 78.0,
                "image": "assets/frames/M3/f00002340.jpg"
              },
              {
                "frame_index": 2370,
                "time_s": 79.0,
                "source_pts_s": 79.0,
                "image": "assets/frames/M3/f00002370.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a14",
          "level": "action",
          "parent_id": "M3.s06",
          "start_s": 78.5,
          "end_s": 83.5,
          "label": "Slide both rackets into the open bag",
          "verb": "slide",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M3.red",
            "M3.dark",
            "M3.bag"
          ],
          "state_before": "Two rackets held together outside the bag.",
          "state_after": "Both racket heads are inside the bag.",
          "evidence_description": "Ordered sampled images support: slide both rackets into the open bag. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2370,
              "time_s": 79.0,
              "source_pts_s": 79.0,
              "image": "assets/frames/M3/f00002370.jpg"
            },
            {
              "frame_index": 2430,
              "time_s": 81.0,
              "source_pts_s": 81.0,
              "image": "assets/frames/M3/f00002430.jpg"
            },
            {
              "frame_index": 2490,
              "time_s": 83.0,
              "source_pts_s": 83.0,
              "image": "assets/frames/M3/f00002490.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 78.5,
            "range_s": [
              78.0,
              79.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2340,
                "time_s": 78.0,
                "source_pts_s": 78.0,
                "image": "assets/frames/M3/f00002340.jpg"
              },
              {
                "frame_index": 2370,
                "time_s": 79.0,
                "source_pts_s": 79.0,
                "image": "assets/frames/M3/f00002370.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 83.5,
            "range_s": [
              83.0,
              84.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2490,
                "time_s": 83.0,
                "source_pts_s": 83.0,
                "image": "assets/frames/M3/f00002490.jpg"
              },
              {
                "frame_index": 2520,
                "time_s": 84.0,
                "source_pts_s": 84.0,
                "image": "assets/frames/M3/f00002520.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M3.a15",
          "level": "action",
          "parent_id": "M3.s06",
          "start_s": 83.5,
          "end_s": 87.933333,
          "label": "Close the bag and lower it toward the chair",
          "verb": "close",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M3.bag",
            "M3.chair"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: close the bag and lower it toward the chair. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2520,
              "time_s": 84.0,
              "source_pts_s": 84.0,
              "image": "assets/frames/M3/f00002520.jpg"
            },
            {
              "frame_index": 2580,
              "time_s": 86.0,
              "source_pts_s": 86.0,
              "image": "assets/frames/M3/f00002580.jpg"
            },
            {
              "frame_index": 2637,
              "time_s": 87.9,
              "source_pts_s": 87.9,
              "image": "assets/frames/M3/f00002637.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 83.5,
            "range_s": [
              83.0,
              84.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2490,
                "time_s": 83.0,
                "source_pts_s": 83.0,
                "image": "assets/frames/M3/f00002490.jpg"
              },
              {
                "frame_index": 2520,
                "time_s": 84.0,
                "source_pts_s": 84.0,
                "image": "assets/frames/M3/f00002520.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 87.933333,
            "range_s": [
              87.933333,
              87.933333
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 2637,
                "time_s": 87.9,
                "source_pts_s": 87.9,
                "image": "assets/frames/M3/f00002637.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "sampling": {
        "base_frames_extracted": 89,
        "base_frames_visually_reviewed": 89,
        "dense_frames_extracted": 26,
        "dense_frames_visually_reviewed": 17,
        "unique_frames_visually_reviewed": 101,
        "source_frame_count": 2638
      },
      "playback": {
        "sha256": "a283cddd55ae93c659922c18943c94bd307d45aa2c47126995e0ff2d243916d0",
        "bytes": 5933528,
        "playback_width": 640,
        "playback_height": 360,
        "source_width": 640,
        "source_height": 360,
        "frame_count": 2638,
        "max_timestamp_error_s": 0.0,
        "duration_s": 87.933333,
        "audio_stream_count": 0,
        "playback_time_offset_s": 0.0,
        "kind": "original_served_from_r2",
        "source_sha256": "a283cddd55ae93c659922c18943c94bd307d45aa2c47126995e0ff2d243916d0",
        "url": "https://media-demo.theroboticsstack.com/media/M3.mp4?v=a283cddd55ae"
      }
    },
    {
      "id": "M1",
      "source": "https://media-demo.theroboticsstack.com/media/M1.mp4?v=3e212c2bf54e",
      "sha256": "3e212c2bf54e1e153604a02e843ea380f4003a12c979c9df4b95608d0187115e",
      "duration_s": 69.966667,
      "width": 640,
      "height": 360,
      "fps_fraction": "30/1",
      "decoded_frame_count": 2099,
      "audio_stream_count": 0,
      "source_start_pts_s": 0.0,
      "last_frame_s": 69.933333,
      "frame_interval_min_s": 0.03333299999999895,
      "frame_interval_max_s": 0.03333400000001063,
      "nonincreasing_pts_count": 0,
      "decode_diagnostics": "",
      "view_layout": "single_egocentric_full_frame",
      "sample_interval_target_s": 1.0,
      "extraction_settings": {
        "maximum_image_width": 1280,
        "crop": null,
        "frame_selection": "nearest_actual_PTS"
      },
      "title": "Arrange plush toys on a shelf",
      "observed_activity": "Adjust a purple plush already on a shelf, then carry five distinguishable plush toys from a bed to the shelf and position each one.",
      "inferred_goal": "Move the loose plush toys from the bed into a shelf arrangement.",
      "assigned_instruction": null,
      "playback_time_offset_s": 0.0,
      "visible_outcome": "The transferred toys remain on the shelf and the hands withdraw at the end.",
      "task_success": "unknown",
      "unknowns": [
        "The opening placement starts before the clip.",
        "The 320 by 180 preview limits small contact details.",
        "The shelf contains other toys; these are not counted as newly transferred."
      ],
      "selection_rationale": "Featured: repeated transfers with distinguishable objects, clear destinations and visible release states.",
      "annotation_provenance": "assistant_visual_review",
      "review_status": "pending",
      "credit": {
        "attribution": "Mecka AI \u00b7 Egoverse \u00b7 recording 69bb1239efeadec2abedad96",
        "source_url": "https://partners.mecka.ai/api/egoverse/uploads/69bb1239efeadec2abedad96/video?redirect=1",
        "metadata_url": "https://partners.mecka.ai/egoverse?embodiment=human_bimanual",
        "license": "CC BY-SA 4.0",
        "license_url": "https://creativecommons.org/licenses/by-sa/4.0/",
        "changes": "Complete source published unchanged from object storage; no re-encoding. Evidence resized to at most 1280px; assistant temporal labels added."
      },
      "protocol_role": "featured",
      "objects": [
        {
          "id": "M1.purple",
          "description": "Purple rabbit-like plush already at the shelf",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M1.green",
          "description": "Green-and-white plush moved from the bed",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M1.pink",
          "description": "Pink bear-like plush",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M1.red",
          "description": "Red bird-like plush",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M1.spotted",
          "description": "Spotted cat-like plush",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M1.black_white",
          "description": "Black-and-white plush",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M1.shelf",
          "description": "Toy shelf",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        },
        {
          "id": "M1.bed",
          "description": "Bed holding the loose toys",
          "identity_scope": "within_clip",
          "identity_uncertainty": "Groups remain grouped; no unsupported count."
        }
      ],
      "subtasks": [
        {
          "id": "M1.s01",
          "level": "subtask",
          "parent_id": null,
          "start_s": 0,
          "end_s": 6.5,
          "label": "Adjust the purple plush",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M1.purple",
            "M1.shelf"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: adjust the purple plush. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 0,
              "time_s": 0.0,
              "source_pts_s": 0.0,
              "image": "assets/frames/M1/f00000000.jpg"
            },
            {
              "frame_index": 90,
              "time_s": 3.0,
              "source_pts_s": 3.0,
              "image": "assets/frames/M1/f00000090.jpg"
            },
            {
              "frame_index": 180,
              "time_s": 6.0,
              "source_pts_s": 6.0,
              "image": "assets/frames/M1/f00000180.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.0,
                "image": "assets/frames/M1/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 6.5,
            "range_s": [
              6.0,
              7.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 180,
                "time_s": 6.0,
                "source_pts_s": 6.0,
                "image": "assets/frames/M1/f00000180.jpg"
              },
              {
                "frame_index": 210,
                "time_s": 7.0,
                "source_pts_s": 7.0,
                "image": "assets/frames/M1/f00000210.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.s02",
          "level": "subtask",
          "parent_id": null,
          "start_s": 6.5,
          "end_s": 23.5,
          "label": "Transfer the green-and-white plush",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M1.bed",
            "M1.green",
            "M1.shelf"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: transfer the green-and-white plush. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 180,
              "time_s": 6.0,
              "source_pts_s": 6.0,
              "image": "assets/frames/M1/f00000180.jpg"
            },
            {
              "frame_index": 450,
              "time_s": 15.0,
              "source_pts_s": 15.0,
              "image": "assets/frames/M1/f00000450.jpg"
            },
            {
              "frame_index": 690,
              "time_s": 23.0,
              "source_pts_s": 23.0,
              "image": "assets/frames/M1/f00000690.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 6.5,
            "range_s": [
              6.0,
              7.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 180,
                "time_s": 6.0,
                "source_pts_s": 6.0,
                "image": "assets/frames/M1/f00000180.jpg"
              },
              {
                "frame_index": 210,
                "time_s": 7.0,
                "source_pts_s": 7.0,
                "image": "assets/frames/M1/f00000210.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 23.5,
            "range_s": [
              23.0,
              24.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 690,
                "time_s": 23.0,
                "source_pts_s": 23.0,
                "image": "assets/frames/M1/f00000690.jpg"
              },
              {
                "frame_index": 720,
                "time_s": 24.0,
                "source_pts_s": 24.0,
                "image": "assets/frames/M1/f00000720.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.s03",
          "level": "subtask",
          "parent_id": null,
          "start_s": 23.5,
          "end_s": 34.483333,
          "label": "Transfer the pink plush",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M1.bed",
            "M1.pink",
            "M1.shelf"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: transfer the pink plush. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 690,
              "time_s": 23.0,
              "source_pts_s": 23.0,
              "image": "assets/frames/M1/f00000690.jpg"
            },
            {
              "frame_index": 870,
              "time_s": 29.0,
              "source_pts_s": 29.0,
              "image": "assets/frames/M1/f00000870.jpg"
            },
            {
              "frame_index": 1035,
              "time_s": 34.5,
              "source_pts_s": 34.5,
              "image": "assets/frames/M1/f00001035.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 23.5,
            "range_s": [
              23.0,
              24.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 690,
                "time_s": 23.0,
                "source_pts_s": 23.0,
                "image": "assets/frames/M1/f00000690.jpg"
              },
              {
                "frame_index": 720,
                "time_s": 24.0,
                "source_pts_s": 24.0,
                "image": "assets/frames/M1/f00000720.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 34.483333,
            "range_s": [
              34.233333,
              34.733333
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Hands leave the pink plush resting on the shelf.",
            "evidence_frames": [
              {
                "frame_index": 1027,
                "time_s": 34.233333,
                "source_pts_s": 34.233333,
                "image": "assets/frames/M1/f00001027.jpg"
              },
              {
                "frame_index": 1042,
                "time_s": 34.733333,
                "source_pts_s": 34.733333,
                "image": "assets/frames/M1/f00001042.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.s04",
          "level": "subtask",
          "parent_id": null,
          "start_s": 34.483333,
          "end_s": 45.5,
          "label": "Transfer the red plush",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M1.bed",
            "M1.red",
            "M1.shelf"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: transfer the red plush. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1035,
              "time_s": 34.5,
              "source_pts_s": 34.5,
              "image": "assets/frames/M1/f00001035.jpg"
            },
            {
              "frame_index": 1200,
              "time_s": 40.0,
              "source_pts_s": 40.0,
              "image": "assets/frames/M1/f00001200.jpg"
            },
            {
              "frame_index": 1350,
              "time_s": 45.0,
              "source_pts_s": 45.0,
              "image": "assets/frames/M1/f00001350.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 34.483333,
            "range_s": [
              34.233333,
              34.733333
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Hands leave the pink plush resting on the shelf.",
            "evidence_frames": [
              {
                "frame_index": 1027,
                "time_s": 34.233333,
                "source_pts_s": 34.233333,
                "image": "assets/frames/M1/f00001027.jpg"
              },
              {
                "frame_index": 1042,
                "time_s": 34.733333,
                "source_pts_s": 34.733333,
                "image": "assets/frames/M1/f00001042.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 45.5,
            "range_s": [
              45.0,
              46.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1350,
                "time_s": 45.0,
                "source_pts_s": 45.0,
                "image": "assets/frames/M1/f00001350.jpg"
              },
              {
                "frame_index": 1380,
                "time_s": 46.0,
                "source_pts_s": 46.0,
                "image": "assets/frames/M1/f00001380.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.s05",
          "level": "subtask",
          "parent_id": null,
          "start_s": 45.5,
          "end_s": 55.5,
          "label": "Transfer the spotted plush",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M1.bed",
            "M1.shelf",
            "M1.spotted"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: transfer the spotted plush. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1350,
              "time_s": 45.0,
              "source_pts_s": 45.0,
              "image": "assets/frames/M1/f00001350.jpg"
            },
            {
              "frame_index": 1500,
              "time_s": 50.0,
              "source_pts_s": 50.0,
              "image": "assets/frames/M1/f00001500.jpg"
            },
            {
              "frame_index": 1650,
              "time_s": 55.0,
              "source_pts_s": 55.0,
              "image": "assets/frames/M1/f00001650.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 45.5,
            "range_s": [
              45.0,
              46.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1350,
                "time_s": 45.0,
                "source_pts_s": 45.0,
                "image": "assets/frames/M1/f00001350.jpg"
              },
              {
                "frame_index": 1380,
                "time_s": 46.0,
                "source_pts_s": 46.0,
                "image": "assets/frames/M1/f00001380.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 55.5,
            "range_s": [
              55.0,
              56.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1650,
                "time_s": 55.0,
                "source_pts_s": 55.0,
                "image": "assets/frames/M1/f00001650.jpg"
              },
              {
                "frame_index": 1680,
                "time_s": 56.0,
                "source_pts_s": 56.0,
                "image": "assets/frames/M1/f00001680.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.s06",
          "level": "subtask",
          "parent_id": null,
          "start_s": 55.5,
          "end_s": 67.5,
          "label": "Transfer the black-and-white plush",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [
            "M1.bed",
            "M1.black_white",
            "M1.shelf"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: transfer the black-and-white plush. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1650,
              "time_s": 55.0,
              "source_pts_s": 55.0,
              "image": "assets/frames/M1/f00001650.jpg"
            },
            {
              "frame_index": 1830,
              "time_s": 61.0,
              "source_pts_s": 61.0,
              "image": "assets/frames/M1/f00001830.jpg"
            },
            {
              "frame_index": 2010,
              "time_s": 67.0,
              "source_pts_s": 67.0,
              "image": "assets/frames/M1/f00002010.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 55.5,
            "range_s": [
              55.0,
              56.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1650,
                "time_s": 55.0,
                "source_pts_s": 55.0,
                "image": "assets/frames/M1/f00001650.jpg"
              },
              {
                "frame_index": 1680,
                "time_s": 56.0,
                "source_pts_s": 56.0,
                "image": "assets/frames/M1/f00001680.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 67.5,
            "range_s": [
              67.0,
              68.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2010,
                "time_s": 67.0,
                "source_pts_s": 67.0,
                "image": "assets/frames/M1/f00002010.jpg"
              },
              {
                "frame_index": 2040,
                "time_s": 68.0,
                "source_pts_s": 68.0,
                "image": "assets/frames/M1/f00002040.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.s07",
          "level": "subtask",
          "parent_id": null,
          "start_s": 67.5,
          "end_s": 69.966667,
          "label": "Look at the shelf arrangement",
          "verb": null,
          "hands": "unknown",
          "hand_roles": {},
          "object_ids": [],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: look at the shelf arrangement. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 2010,
              "time_s": 67.0,
              "source_pts_s": 67.0,
              "image": "assets/frames/M1/f00002010.jpg"
            },
            {
              "frame_index": 2070,
              "time_s": 69.0,
              "source_pts_s": 69.0,
              "image": "assets/frames/M1/f00002070.jpg"
            },
            {
              "frame_index": 2098,
              "time_s": 69.933333,
              "source_pts_s": 69.933333,
              "image": "assets/frames/M1/f00002098.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 67.5,
            "range_s": [
              67.0,
              68.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2010,
                "time_s": 67.0,
                "source_pts_s": 67.0,
                "image": "assets/frames/M1/f00002010.jpg"
              },
              {
                "frame_index": 2040,
                "time_s": 68.0,
                "source_pts_s": 68.0,
                "image": "assets/frames/M1/f00002040.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 69.966667,
            "range_s": [
              69.966667,
              69.966667
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 2098,
                "time_s": 69.933333,
                "source_pts_s": 69.933333,
                "image": "assets/frames/M1/f00002098.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "actions": [
        {
          "id": "M1.a01",
          "level": "action",
          "parent_id": "M1.s01",
          "start_s": 0,
          "end_s": 6.5,
          "label": "Adjust the purple plush on the shelf",
          "verb": "adjust",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.purple",
            "M1.shelf"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: adjust the purple plush on the shelf. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 0,
              "time_s": 0.0,
              "source_pts_s": 0.0,
              "image": "assets/frames/M1/f00000000.jpg"
            },
            {
              "frame_index": 90,
              "time_s": 3.0,
              "source_pts_s": 3.0,
              "image": "assets/frames/M1/f00000090.jpg"
            },
            {
              "frame_index": 180,
              "time_s": 6.0,
              "source_pts_s": 6.0,
              "image": "assets/frames/M1/f00000180.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 0,
            "range_s": [
              0,
              0
            ],
            "basis": "clip_edge",
            "clip_edge": true,
            "evidence_frames": [
              {
                "frame_index": 0,
                "time_s": 0.0,
                "source_pts_s": 0.0,
                "image": "assets/frames/M1/f00000000.jpg"
              }
            ],
            "uncertainty": "Recording edge; the physical event may extend beyond the clip."
          },
          "end_boundary": {
            "estimate_s": 6.5,
            "range_s": [
              6.0,
              7.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 180,
                "time_s": 6.0,
                "source_pts_s": 6.0,
                "image": "assets/frames/M1/f00000180.jpg"
              },
              {
                "frame_index": 210,
                "time_s": 7.0,
                "source_pts_s": 7.0,
                "image": "assets/frames/M1/f00000210.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a02",
          "level": "action",
          "parent_id": "M1.s02",
          "start_s": 6.5,
          "end_s": 14.5,
          "label": "Return to the bed and grasp the green-and-white plush",
          "verb": "return",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.green",
            "M1.bed"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: return to the bed and grasp the green-and-white plush. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 240,
              "time_s": 8.0,
              "source_pts_s": 8.0,
              "image": "assets/frames/M1/f00000240.jpg"
            },
            {
              "frame_index": 330,
              "time_s": 11.0,
              "source_pts_s": 11.0,
              "image": "assets/frames/M1/f00000330.jpg"
            },
            {
              "frame_index": 420,
              "time_s": 14.0,
              "source_pts_s": 14.0,
              "image": "assets/frames/M1/f00000420.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 6.5,
            "range_s": [
              6.0,
              7.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 180,
                "time_s": 6.0,
                "source_pts_s": 6.0,
                "image": "assets/frames/M1/f00000180.jpg"
              },
              {
                "frame_index": 210,
                "time_s": 7.0,
                "source_pts_s": 7.0,
                "image": "assets/frames/M1/f00000210.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 14.5,
            "range_s": [
              14.0,
              15.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 420,
                "time_s": 14.0,
                "source_pts_s": 14.0,
                "image": "assets/frames/M1/f00000420.jpg"
              },
              {
                "frame_index": 450,
                "time_s": 15.0,
                "source_pts_s": 15.0,
                "image": "assets/frames/M1/f00000450.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a03",
          "level": "action",
          "parent_id": "M1.s02",
          "start_s": 14.5,
          "end_s": 17.5,
          "label": "Carry the green-and-white plush toward the shelf",
          "verb": "carry",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.green"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: carry the green-and-white plush toward the shelf. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 450,
              "time_s": 15.0,
              "source_pts_s": 15.0,
              "image": "assets/frames/M1/f00000450.jpg"
            },
            {
              "frame_index": 480,
              "time_s": 16.0,
              "source_pts_s": 16.0,
              "image": "assets/frames/M1/f00000480.jpg"
            },
            {
              "frame_index": 510,
              "time_s": 17.0,
              "source_pts_s": 17.0,
              "image": "assets/frames/M1/f00000510.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 14.5,
            "range_s": [
              14.0,
              15.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 420,
                "time_s": 14.0,
                "source_pts_s": 14.0,
                "image": "assets/frames/M1/f00000420.jpg"
              },
              {
                "frame_index": 450,
                "time_s": 15.0,
                "source_pts_s": 15.0,
                "image": "assets/frames/M1/f00000450.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 17.5,
            "range_s": [
              17.0,
              18.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 510,
                "time_s": 17.0,
                "source_pts_s": 17.0,
                "image": "assets/frames/M1/f00000510.jpg"
              },
              {
                "frame_index": 540,
                "time_s": 18.0,
                "source_pts_s": 18.0,
                "image": "assets/frames/M1/f00000540.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a04",
          "level": "action",
          "parent_id": "M1.s02",
          "start_s": 17.5,
          "end_s": 23.5,
          "label": "Place and adjust the green-and-white plush on the shelf",
          "verb": "place",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.green",
            "M1.shelf"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: place and adjust the green-and-white plush on the shelf. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 540,
              "time_s": 18.0,
              "source_pts_s": 18.0,
              "image": "assets/frames/M1/f00000540.jpg"
            },
            {
              "frame_index": 630,
              "time_s": 21.0,
              "source_pts_s": 21.0,
              "image": "assets/frames/M1/f00000630.jpg"
            },
            {
              "frame_index": 690,
              "time_s": 23.0,
              "source_pts_s": 23.0,
              "image": "assets/frames/M1/f00000690.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 17.5,
            "range_s": [
              17.0,
              18.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 510,
                "time_s": 17.0,
                "source_pts_s": 17.0,
                "image": "assets/frames/M1/f00000510.jpg"
              },
              {
                "frame_index": 540,
                "time_s": 18.0,
                "source_pts_s": 18.0,
                "image": "assets/frames/M1/f00000540.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 23.5,
            "range_s": [
              23.0,
              24.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 690,
                "time_s": 23.0,
                "source_pts_s": 23.0,
                "image": "assets/frames/M1/f00000690.jpg"
              },
              {
                "frame_index": 720,
                "time_s": 24.0,
                "source_pts_s": 24.0,
                "image": "assets/frames/M1/f00000720.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a05",
          "level": "action",
          "parent_id": "M1.s03",
          "start_s": 23.5,
          "end_s": 27.5,
          "label": "Return to the bed and pick up the pink plush",
          "verb": "return",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.pink",
            "M1.bed"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: return to the bed and pick up the pink plush. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 720,
              "time_s": 24.0,
              "source_pts_s": 24.0,
              "image": "assets/frames/M1/f00000720.jpg"
            },
            {
              "frame_index": 780,
              "time_s": 26.0,
              "source_pts_s": 26.0,
              "image": "assets/frames/M1/f00000780.jpg"
            },
            {
              "frame_index": 810,
              "time_s": 27.0,
              "source_pts_s": 27.0,
              "image": "assets/frames/M1/f00000810.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 23.5,
            "range_s": [
              23.0,
              24.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 690,
                "time_s": 23.0,
                "source_pts_s": 23.0,
                "image": "assets/frames/M1/f00000690.jpg"
              },
              {
                "frame_index": 720,
                "time_s": 24.0,
                "source_pts_s": 24.0,
                "image": "assets/frames/M1/f00000720.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 27.5,
            "range_s": [
              27.0,
              28.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 810,
                "time_s": 27.0,
                "source_pts_s": 27.0,
                "image": "assets/frames/M1/f00000810.jpg"
              },
              {
                "frame_index": 840,
                "time_s": 28.0,
                "source_pts_s": 28.0,
                "image": "assets/frames/M1/f00000840.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a06",
          "level": "action",
          "parent_id": "M1.s03",
          "start_s": 27.5,
          "end_s": 30.5,
          "label": "Carry the pink plush to the shelf",
          "verb": "carry",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.pink"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: carry the pink plush to the shelf. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 840,
              "time_s": 28.0,
              "source_pts_s": 28.0,
              "image": "assets/frames/M1/f00000840.jpg"
            },
            {
              "frame_index": 870,
              "time_s": 29.0,
              "source_pts_s": 29.0,
              "image": "assets/frames/M1/f00000870.jpg"
            },
            {
              "frame_index": 900,
              "time_s": 30.0,
              "source_pts_s": 30.0,
              "image": "assets/frames/M1/f00000900.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 27.5,
            "range_s": [
              27.0,
              28.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 810,
                "time_s": 27.0,
                "source_pts_s": 27.0,
                "image": "assets/frames/M1/f00000810.jpg"
              },
              {
                "frame_index": 840,
                "time_s": 28.0,
                "source_pts_s": 28.0,
                "image": "assets/frames/M1/f00000840.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 30.5,
            "range_s": [
              30.0,
              31.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 900,
                "time_s": 30.0,
                "source_pts_s": 30.0,
                "image": "assets/frames/M1/f00000900.jpg"
              },
              {
                "frame_index": 930,
                "time_s": 31.0,
                "source_pts_s": 31.0,
                "image": "assets/frames/M1/f00000930.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a07",
          "level": "action",
          "parent_id": "M1.s03",
          "start_s": 30.5,
          "end_s": 34.483333,
          "label": "Place and adjust the pink plush on the shelf",
          "verb": "place",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.pink",
            "M1.shelf"
          ],
          "state_before": "Pink plush held in front of the shelf.",
          "state_after": "Pink plush rests on the shelf after the hands withdraw.",
          "evidence_description": "Ordered sampled images support: place and adjust the pink plush on the shelf. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 930,
              "time_s": 31.0,
              "source_pts_s": 31.0,
              "image": "assets/frames/M1/f00000930.jpg"
            },
            {
              "frame_index": 990,
              "time_s": 33.0,
              "source_pts_s": 33.0,
              "image": "assets/frames/M1/f00000990.jpg"
            },
            {
              "frame_index": 1050,
              "time_s": 35.0,
              "source_pts_s": 35.0,
              "image": "assets/frames/M1/f00001050.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 30.5,
            "range_s": [
              30.0,
              31.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 900,
                "time_s": 30.0,
                "source_pts_s": 30.0,
                "image": "assets/frames/M1/f00000900.jpg"
              },
              {
                "frame_index": 930,
                "time_s": 31.0,
                "source_pts_s": 31.0,
                "image": "assets/frames/M1/f00000930.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 34.483333,
            "range_s": [
              34.233333,
              34.733333
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Hands leave the pink plush resting on the shelf.",
            "evidence_frames": [
              {
                "frame_index": 1027,
                "time_s": 34.233333,
                "source_pts_s": 34.233333,
                "image": "assets/frames/M1/f00001027.jpg"
              },
              {
                "frame_index": 1042,
                "time_s": 34.733333,
                "source_pts_s": 34.733333,
                "image": "assets/frames/M1/f00001042.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a08",
          "level": "action",
          "parent_id": "M1.s04",
          "start_s": 34.483333,
          "end_s": 39.5,
          "label": "Return to the bed and pick up the red plush",
          "verb": "return",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.red",
            "M1.bed"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: return to the bed and pick up the red plush. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1080,
              "time_s": 36.0,
              "source_pts_s": 36.0,
              "image": "assets/frames/M1/f00001080.jpg"
            },
            {
              "frame_index": 1140,
              "time_s": 38.0,
              "source_pts_s": 38.0,
              "image": "assets/frames/M1/f00001140.jpg"
            },
            {
              "frame_index": 1170,
              "time_s": 39.0,
              "source_pts_s": 39.0,
              "image": "assets/frames/M1/f00001170.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 34.483333,
            "range_s": [
              34.233333,
              34.733333
            ],
            "basis": "dense_visual_transition_bracket",
            "clip_edge": false,
            "criterion": "Hands leave the pink plush resting on the shelf.",
            "evidence_frames": [
              {
                "frame_index": 1027,
                "time_s": 34.233333,
                "source_pts_s": 34.233333,
                "image": "assets/frames/M1/f00001027.jpg"
              },
              {
                "frame_index": 1042,
                "time_s": 34.733333,
                "source_pts_s": 34.733333,
                "image": "assets/frames/M1/f00001042.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 39.5,
            "range_s": [
              39.0,
              40.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1170,
                "time_s": 39.0,
                "source_pts_s": 39.0,
                "image": "assets/frames/M1/f00001170.jpg"
              },
              {
                "frame_index": 1200,
                "time_s": 40.0,
                "source_pts_s": 40.0,
                "image": "assets/frames/M1/f00001200.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a09",
          "level": "action",
          "parent_id": "M1.s04",
          "start_s": 39.5,
          "end_s": 41.5,
          "label": "Carry the red plush to the shelf",
          "verb": "carry",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.red"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: carry the red plush to the shelf. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1200,
              "time_s": 40.0,
              "source_pts_s": 40.0,
              "image": "assets/frames/M1/f00001200.jpg"
            },
            {
              "frame_index": 1230,
              "time_s": 41.0,
              "source_pts_s": 41.0,
              "image": "assets/frames/M1/f00001230.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 39.5,
            "range_s": [
              39.0,
              40.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1170,
                "time_s": 39.0,
                "source_pts_s": 39.0,
                "image": "assets/frames/M1/f00001170.jpg"
              },
              {
                "frame_index": 1200,
                "time_s": 40.0,
                "source_pts_s": 40.0,
                "image": "assets/frames/M1/f00001200.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 41.5,
            "range_s": [
              41.0,
              42.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1230,
                "time_s": 41.0,
                "source_pts_s": 41.0,
                "image": "assets/frames/M1/f00001230.jpg"
              },
              {
                "frame_index": 1260,
                "time_s": 42.0,
                "source_pts_s": 42.0,
                "image": "assets/frames/M1/f00001260.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a10",
          "level": "action",
          "parent_id": "M1.s04",
          "start_s": 41.5,
          "end_s": 45.5,
          "label": "Place and adjust the red plush on the shelf",
          "verb": "place",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.red",
            "M1.shelf"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: place and adjust the red plush on the shelf. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1260,
              "time_s": 42.0,
              "source_pts_s": 42.0,
              "image": "assets/frames/M1/f00001260.jpg"
            },
            {
              "frame_index": 1320,
              "time_s": 44.0,
              "source_pts_s": 44.0,
              "image": "assets/frames/M1/f00001320.jpg"
            },
            {
              "frame_index": 1350,
              "time_s": 45.0,
              "source_pts_s": 45.0,
              "image": "assets/frames/M1/f00001350.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 41.5,
            "range_s": [
              41.0,
              42.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1230,
                "time_s": 41.0,
                "source_pts_s": 41.0,
                "image": "assets/frames/M1/f00001230.jpg"
              },
              {
                "frame_index": 1260,
                "time_s": 42.0,
                "source_pts_s": 42.0,
                "image": "assets/frames/M1/f00001260.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 45.5,
            "range_s": [
              45.0,
              46.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1350,
                "time_s": 45.0,
                "source_pts_s": 45.0,
                "image": "assets/frames/M1/f00001350.jpg"
              },
              {
                "frame_index": 1380,
                "time_s": 46.0,
                "source_pts_s": 46.0,
                "image": "assets/frames/M1/f00001380.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a11",
          "level": "action",
          "parent_id": "M1.s05",
          "start_s": 45.5,
          "end_s": 48.5,
          "label": "Return to the bed and pick up the spotted plush",
          "verb": "return",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.spotted",
            "M1.bed"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: return to the bed and pick up the spotted plush. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1380,
              "time_s": 46.0,
              "source_pts_s": 46.0,
              "image": "assets/frames/M1/f00001380.jpg"
            },
            {
              "frame_index": 1410,
              "time_s": 47.0,
              "source_pts_s": 47.0,
              "image": "assets/frames/M1/f00001410.jpg"
            },
            {
              "frame_index": 1440,
              "time_s": 48.0,
              "source_pts_s": 48.0,
              "image": "assets/frames/M1/f00001440.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 45.5,
            "range_s": [
              45.0,
              46.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1350,
                "time_s": 45.0,
                "source_pts_s": 45.0,
                "image": "assets/frames/M1/f00001350.jpg"
              },
              {
                "frame_index": 1380,
                "time_s": 46.0,
                "source_pts_s": 46.0,
                "image": "assets/frames/M1/f00001380.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 48.5,
            "range_s": [
              48.0,
              49.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1440,
                "time_s": 48.0,
                "source_pts_s": 48.0,
                "image": "assets/frames/M1/f00001440.jpg"
              },
              {
                "frame_index": 1470,
                "time_s": 49.0,
                "source_pts_s": 49.0,
                "image": "assets/frames/M1/f00001470.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a12",
          "level": "action",
          "parent_id": "M1.s05",
          "start_s": 48.5,
          "end_s": 50.5,
          "label": "Carry the spotted plush to the shelf",
          "verb": "carry",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.spotted"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: carry the spotted plush to the shelf. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1470,
              "time_s": 49.0,
              "source_pts_s": 49.0,
              "image": "assets/frames/M1/f00001470.jpg"
            },
            {
              "frame_index": 1500,
              "time_s": 50.0,
              "source_pts_s": 50.0,
              "image": "assets/frames/M1/f00001500.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 48.5,
            "range_s": [
              48.0,
              49.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1440,
                "time_s": 48.0,
                "source_pts_s": 48.0,
                "image": "assets/frames/M1/f00001440.jpg"
              },
              {
                "frame_index": 1470,
                "time_s": 49.0,
                "source_pts_s": 49.0,
                "image": "assets/frames/M1/f00001470.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 50.5,
            "range_s": [
              50.0,
              51.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1500,
                "time_s": 50.0,
                "source_pts_s": 50.0,
                "image": "assets/frames/M1/f00001500.jpg"
              },
              {
                "frame_index": 1530,
                "time_s": 51.0,
                "source_pts_s": 51.0,
                "image": "assets/frames/M1/f00001530.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a13",
          "level": "action",
          "parent_id": "M1.s05",
          "start_s": 50.5,
          "end_s": 55.5,
          "label": "Place and adjust the spotted plush on the shelf",
          "verb": "place",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.spotted",
            "M1.shelf"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: place and adjust the spotted plush on the shelf. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1530,
              "time_s": 51.0,
              "source_pts_s": 51.0,
              "image": "assets/frames/M1/f00001530.jpg"
            },
            {
              "frame_index": 1590,
              "time_s": 53.0,
              "source_pts_s": 53.0,
              "image": "assets/frames/M1/f00001590.jpg"
            },
            {
              "frame_index": 1650,
              "time_s": 55.0,
              "source_pts_s": 55.0,
              "image": "assets/frames/M1/f00001650.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 50.5,
            "range_s": [
              50.0,
              51.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1500,
                "time_s": 50.0,
                "source_pts_s": 50.0,
                "image": "assets/frames/M1/f00001500.jpg"
              },
              {
                "frame_index": 1530,
                "time_s": 51.0,
                "source_pts_s": 51.0,
                "image": "assets/frames/M1/f00001530.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 55.5,
            "range_s": [
              55.0,
              56.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1650,
                "time_s": 55.0,
                "source_pts_s": 55.0,
                "image": "assets/frames/M1/f00001650.jpg"
              },
              {
                "frame_index": 1680,
                "time_s": 56.0,
                "source_pts_s": 56.0,
                "image": "assets/frames/M1/f00001680.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a14",
          "level": "action",
          "parent_id": "M1.s06",
          "start_s": 55.5,
          "end_s": 58.5,
          "label": "Return to the bed and pick up the black-and-white plush",
          "verb": "return",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.black_white",
            "M1.bed"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: return to the bed and pick up the black-and-white plush. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1680,
              "time_s": 56.0,
              "source_pts_s": 56.0,
              "image": "assets/frames/M1/f00001680.jpg"
            },
            {
              "frame_index": 1710,
              "time_s": 57.0,
              "source_pts_s": 57.0,
              "image": "assets/frames/M1/f00001710.jpg"
            },
            {
              "frame_index": 1740,
              "time_s": 58.0,
              "source_pts_s": 58.0,
              "image": "assets/frames/M1/f00001740.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 55.5,
            "range_s": [
              55.0,
              56.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1650,
                "time_s": 55.0,
                "source_pts_s": 55.0,
                "image": "assets/frames/M1/f00001650.jpg"
              },
              {
                "frame_index": 1680,
                "time_s": 56.0,
                "source_pts_s": 56.0,
                "image": "assets/frames/M1/f00001680.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 58.5,
            "range_s": [
              58.0,
              59.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1740,
                "time_s": 58.0,
                "source_pts_s": 58.0,
                "image": "assets/frames/M1/f00001740.jpg"
              },
              {
                "frame_index": 1770,
                "time_s": 59.0,
                "source_pts_s": 59.0,
                "image": "assets/frames/M1/f00001770.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a15",
          "level": "action",
          "parent_id": "M1.s06",
          "start_s": 58.5,
          "end_s": 61.5,
          "label": "Carry the black-and-white plush to the shelf",
          "verb": "carry",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.black_white"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: carry the black-and-white plush to the shelf. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1770,
              "time_s": 59.0,
              "source_pts_s": 59.0,
              "image": "assets/frames/M1/f00001770.jpg"
            },
            {
              "frame_index": 1800,
              "time_s": 60.0,
              "source_pts_s": 60.0,
              "image": "assets/frames/M1/f00001800.jpg"
            },
            {
              "frame_index": 1830,
              "time_s": 61.0,
              "source_pts_s": 61.0,
              "image": "assets/frames/M1/f00001830.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 58.5,
            "range_s": [
              58.0,
              59.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1740,
                "time_s": 58.0,
                "source_pts_s": 58.0,
                "image": "assets/frames/M1/f00001740.jpg"
              },
              {
                "frame_index": 1770,
                "time_s": 59.0,
                "source_pts_s": 59.0,
                "image": "assets/frames/M1/f00001770.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 61.5,
            "range_s": [
              61.0,
              62.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1830,
                "time_s": 61.0,
                "source_pts_s": 61.0,
                "image": "assets/frames/M1/f00001830.jpg"
              },
              {
                "frame_index": 1860,
                "time_s": 62.0,
                "source_pts_s": 62.0,
                "image": "assets/frames/M1/f00001860.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        },
        {
          "id": "M1.a16",
          "level": "action",
          "parent_id": "M1.s06",
          "start_s": 61.5,
          "end_s": 67.5,
          "label": "Place and adjust the black-and-white plush on the shelf",
          "verb": "place",
          "hands": "both",
          "hand_roles": {},
          "object_ids": [
            "M1.black_white",
            "M1.shelf"
          ],
          "state_before": null,
          "state_after": null,
          "evidence_description": "Ordered sampled images support: place and adjust the black-and-white plush on the shelf. Context images may bracket the span; intermediate motions are not continuously observed.",
          "evidence_frames": [
            {
              "frame_index": 1860,
              "time_s": 62.0,
              "source_pts_s": 62.0,
              "image": "assets/frames/M1/f00001860.jpg"
            },
            {
              "frame_index": 1950,
              "time_s": 65.0,
              "source_pts_s": 65.0,
              "image": "assets/frames/M1/f00001950.jpg"
            },
            {
              "frame_index": 2010,
              "time_s": 67.0,
              "source_pts_s": 67.0,
              "image": "assets/frames/M1/f00002010.jpg"
            }
          ],
          "start_boundary": {
            "estimate_s": 61.5,
            "range_s": [
              61.0,
              62.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 1830,
                "time_s": 61.0,
                "source_pts_s": 61.0,
                "image": "assets/frames/M1/f00001830.jpg"
              },
              {
                "frame_index": 1860,
                "time_s": 62.0,
                "source_pts_s": 62.0,
                "image": "assets/frames/M1/f00001860.jpg"
              }
            ]
          },
          "end_boundary": {
            "estimate_s": 67.5,
            "range_s": [
              67.0,
              68.0
            ],
            "basis": "sampled_visual_estimate",
            "clip_edge": false,
            "evidence_frames": [
              {
                "frame_index": 2010,
                "time_s": 67.0,
                "source_pts_s": 67.0,
                "image": "assets/frames/M1/f00002010.jpg"
              },
              {
                "frame_index": 2040,
                "time_s": 68.0,
                "source_pts_s": 68.0,
                "image": "assets/frames/M1/f00002040.jpg"
              }
            ]
          },
          "uncertainty": [
            "Semantic span estimate from sampled imagery; brief intervening actions may be missed."
          ],
          "provenance": "assistant_visual_review",
          "review_status": "pending",
          "review_note": ""
        }
      ],
      "sampling": {
        "base_frames_extracted": 71,
        "base_frames_visually_reviewed": 71,
        "dense_frames_extracted": 22,
        "dense_frames_visually_reviewed": 13,
        "unique_frames_visually_reviewed": 80,
        "source_frame_count": 2099
      },
      "playback": {
        "sha256": "3e212c2bf54e1e153604a02e843ea380f4003a12c979c9df4b95608d0187115e",
        "bytes": 3502694,
        "playback_width": 640,
        "playback_height": 360,
        "source_width": 640,
        "source_height": 360,
        "frame_count": 2099,
        "max_timestamp_error_s": 0.0,
        "duration_s": 69.966667,
        "audio_stream_count": 0,
        "playback_time_offset_s": 0.0,
        "kind": "original_served_from_r2",
        "source_sha256": "3e212c2bf54e1e153604a02e843ea380f4003a12c979c9df4b95608d0187115e",
        "url": "https://media-demo.theroboticsstack.com/media/M1.mp4?v=3e212c2bf54e"
      }
    }
  ]
}