version: "1.0.0" name: cut-object description: Cut a target object using a knife or axe already held by the robot. Use when the next task step requires cutting, chopping, or slicing an object.

Cut object

<!-- codex-coordinate-contract:start --> The latest image in the conversation represents the current observed state. Older images are historical context; their points, bounding boxes, and object-part locations do not describe the current view. After any motion, re-locate the target in the latest image; do not shift, rotate, or scale an older location estimate to choose a new click.

Before EVERY point selection, inspect the latest attached clickable head image and re-locate the physical target. Use the full attached image as a relative coordinate canvas: integer u (column) and v (row) in 0..1000, with top-left (0,0) and bottom-right (1000,1000). Do not use encoded image pixels or crop coordinates. Bind every point, including every row of a multi-point call, to that exact current image_id. Never copy locations from older images, earlier reasoning, reference photos, tracking text, or overlay labels. Wrist images are observation-only and cannot be clicked. If the point is rejected, obtain a fresh head image and ground again; do not guess an image_id. <!-- codex-coordinate-contract:end -->

<!-- codex-task-lifecycle:start --> Follow this procedure only while cut-object is the active task Skill loaded by activate_skill. Its presence in history, including after compaction, does not make it active. After success and all required post-actions, return to the parent task; activate another matching Skill only when its procedure is needed. Instructions below to stay, retry, or exit describe pursuit of success and do not prohibit a change of approach or handoff to the parent task. <!-- codex-task-lifecycle:end -->

1. Use one current head capture where the held knife or axe and the target

object are both visible. Identify which arm holds the cutting tool and keep that gripper closed.

1. If the cutting edge is not visible, or is not yet at a convenient cutting

orientation, call control_wrist_roll on the holding arm until the blade faces the most convenient cut pose. Recapture the head image after each roll. Do not call set_arm_to_grasp_position before or during the cut: that call zeros the previous J8 wrist-roll amount and undoes this orientation.

1. Select exactly two points on the same current image:

• cutting_tool_point: any visible point on the knife or axe.
• target_object_point: the point to touch on the target object.

1. Call track_object_distance once with that capture's image_id and both

named points.

1. Read both current xyz_in_robot_base_coord_m values from

persistent_tracking. Compute target_object_point - cutting_tool_point.

1. Move the holding arm by that robot-base XYZ difference with

adjust_left_eef_pose_in_head_frame or adjust_right_eef_pose_in_head_frame. Pass the difference as x, y, and z; keep roll, pitch, and yaw unchanged.

1. After each move, read the updated tracked points and repeat steps 5-6 until

the cutting-tool point touches the target point. If either point is lost, capture a new head image and bind both points again.

1. After the whole cut is complete, call set_arm_to_grasp_position once for

the holding arm with gripper="keep". This is the required last step. It also zeros the wrist roll from step 2. Then return control to the parent task; this is Skill exit, not session end.