[system]
You are proficient in multi-agent collaborative planning and task design. Your goal is to generate a practical collaborative task consisting of multiple navigation and interaction subtasks, based on the provided scene graph (which contains spatial layout, room connections, and object distributions) and the single agent task.
[user]
You will receive two types of input to design a multi-robot collaborative task:
"""
scene graph:
- "region": A list of region identifiers (e.g., "Kitchen_1", "Bedroom_2").
- "link": A list of bidirectional region connections indicating navigability (e.g., "Kitchen_1 <-> Bedroom_2").
- "item": A list of region-specific items, including:
  - "region": Region identifier.
  - "asset": Non-movable, fixed furniture such as tables, shelves, or counters. These serve as reference locations or delivery targets.
  - "object": Movable, portable items that robots can manipulate. Only objects may be picked up or released.
single agent task:
- "Single robot start region": Starting region of robot_1.
- "Single robot target object region": Region where the target object is located.
- "Target object": target object.
- "Single robot end region": Region where the target object is delivered.
- "Single robot travel path": Ordered region-to-region path followed by robot_1.
- "Task instruction": Natural language description of the task for robot_1.
"""