Trigger this skill when the user wants embodied data where the task can be completed in an unsafe way unless the agent reasons about hazards, cleanup, shutdown, or risky object placement. Plain-language triggers include: 'do the task but safely,' 'avoid dangerous plans,' 'make the robot notice fire or slip risk,' 'questions about safe completion not just success,' and 'test if it leaves the environment in a risky state.'
[Case 1]
[Case 2]
[Case 3]
To synthesize data for this capability, you must strictly follow a 3-phase pipeline. Do not hallucinate steps. Read the corresponding reference file for each phase sequentially:
Phase 1: Environment Exploration
Read the exploration guidelines to discover raw knowledge seeds:
references/EXPLORATION.md
Phase 2: Trajectory Selection
Once Phase 1 is complete, read the selection criteria to evaluate the trajectory:
references/SELECTION.md
Phase 3: Data Synthesis
Once a trajectory passes Phase 2, read the synthesis instructions to generate the final data:
references/SYNTHESIS.md