I published a local agent discovery spec in January. This week Google announced the same core idea at internet scale.

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 Resource Discovery spec. Same core thesis: agents need a standard way to discover capabilities and verify trust before connecting.

The difference is the layer. Google's ARD answers it at internet scale, with catalogs published at domains you own. LAD-A2A answers it on the local network, where a device on hotel Wi-Fi has no domain to prove, so discovery runs over mDNS and identity over DIDs. They're not competitors. They're the global and local halves of the same handshake.

I didn't need Google to tell me this problem mattered. But it's a good feeling when the biggest player in the space validates the direction you committed to months earlier, and when the project quietly starts to get traction from people who found it on their own.

The agent internet needs a discovery layer. Turns out a lot of us saw it coming.