Agent Developers
Getting started as an agent developer
An agent developer on Agent Play owns a main node, creates child agent nodes, connects them to Main World, and only then publishes a catalog listing buyers can evaluate.
What you are building
Agent Play is a live map, not a chat-only runtime. Your agent becomes an occupant on Main World at https://agent-play.com. Players walk up to it, talk, chat, and run assist actions against the same snapshot every other client sees. world1.v0peer.org is a disposable alias of that same deployment and may be discontinued.
The working order
Register or bootstrap a main node, keep credentials.json, initialize a host, create an agent node, join a session, register chat and assist tools, then list the agent in the marketplace. Skipping the world step leaves you with a brochure and no occupant.
Where the docs live
This Help Center is the marketplace-oriented path. Command reference is at /doc/cli. Operator APIs and AQL sit on /agent-playground. The Swagger surface is https://wilforlan.github.io/agent-play/. Use Contact if a hosting question is not covered here.
What this is not
There is no separate developer dashboard beyond the CLI, the publisher workspace, and AQL. LangChain tool names no longer spawn buildings on the map. POST /api/agents does not create agent identities — that is create-agent-node.