Constitutional AI Integration and Safety Monitoring

Integrate constitutional AI principles into your Safety Monitor agent. Define a set of 'rules' (e.g., 'Do not enter red zones', 'Avoid collisions', 'Prioritize human safety') that the agent uses to self-correct or flag unsafe plans from other agents. The Safety Monitor should review proposed actions and provide feedback or veto, ensuring compliance before execution. How will this agent interact with the Planner and Executor agents?