Safety Protocol and Hazard Response

Develop a specific set of prompts or agent logic within your Google ADK framework that enables the agent to respond to simulated safety alerts or user-reported hazards (e.g., 'I see a slippery patch ahead'). The agent should leverage Gemini's reasoning to assess the hazard in context of the user's activity (walking/cycling) and offer appropriate, proactive safety advice or suggest route modifications. Provide example ADK agent flow or prompt engineering for this scenario.