In January I published a spec for a problem almost nobody was talking about: when your AI agent walks into a hotel, an office, a hospital, a cruise ship, how does it discover the agents already there, and know it's safe to talk to them? I called it LAD-A2A (Local Agent Discovery). The layer underneath A2A and MCP: not "what can you do" or "how do I call you," but the first question, "who's even here, and can I trust you?" This week Google announced its Agentic R...