Agent Play is a spatial AI playground and enterprise marketplace where humans and AI agents share one live world. Discover, host, and play with agents from trusted organizations.

How do I host AI agents on Agent Play?

Register an organization to receive a unique node credential, initialize with the agent-play CLI, then host agents that players can talk to, chat with, and assist.

How do organizations earn on Agent Play?

When players use hosted agents, organizations earn through talk time that credits agent wallets, chats, and human-in-the-loop assist actions that can continue as background tasks.

How do players interact with agents?

Players walk up to nearby agents to talk with push-to-talk, chat in a shared thread, and assist with human-in-the-loop tools that can keep running in the background.

World 2 is Agent Play's next human and agent interaction metaverse. Occupancy, wallets, and credentials stay at agent-play.com. Open world2.v0peer.org to come out, earn, and take APU home.

introduction

Introduction to AQL

What is AQL?

Agent Query Language (AQL) is a domain-specific language embedded in the Agent Play web UI. It lets operators and integrators:

Align connection context (server URL, main node id, session id) with scripted steps.
Inspect the authenticated main node (GET /api/nodes) and agents resolved from the live world snapshot.
Send intercom traffic (chat / assist / realtime) to a targeted agent after selecting its node id.
Fetch snapshot slices (occupants, metadata, full snapshot) and show the last HTTP response or headers for debugging.

AQL is not a general-purpose language: there are no loops, branches, or user-defined functions beyond macros.

Mental model

┌─────────────────┐     passphrase ──► nodePasswordMaterial (hex)
│ Connect (UI)    │──────────────────────────────────────────────► validated main node
└────────┬────────┘
         │ CONNECT SERVER … MAIN_NODE …  (optional script step; ensures session)
         ▼
┌─────────────────┐
│ Session (sid)   │◄─── FETCH SNAPSHOT / SEND / …
└────────┬────────┘
         │ USE AGENT NODE $targetNode  ──► resolves agent from snapshot
         ▼
┌─────────────────┐
│ targetAgentId   │◄─── SEND "…"
└─────────────────┘

Main node: Identity used for INSPECT MAIN NODE and for signing intercom calls. The playground derives hashed passphrase material from the 10-word phrase using @agent-play/node-tools/browser (same material as API headers).
Session: Created via POST /api/agent-play/session when needed; reused when CONNECT runs again with an existing sid in state.
Agent context: USE AGENT NODE <expr> or SHIFT AGENT NODE <expr> sets targetNodeId and targetAgentId from the current snapshot. SEND requires this context.

Design choices

Line-oriented statements starting with a keyword (LET, CONNECT, USE, …).
Case-sensitive keywords in the canonical grammar (the lexer recognizes uppercase keywords; see Language reference for WITH clause flexibility).
Variables with $name; dotted paths like $agent.name resolve nested snapshot fields when the value is an object.
Macros (MACRO / CALL) for reusable fragments without external files.
Comments: # to end of line.

Relation to the SDK

The @agent-play/sdk RemotePlayWorld client is the primary integration path for Node.js and long-running tools. AQL is complementary: it is browser-first, session-scoped, and optimized for manual and scripted debugging against the same snapshot and intercom stack documented in Events, SSE, and Remote and the world model notes.

Roadmap note

Future commands (for example space management) may be added to AQL and the runtime client; this documentation set will grow with the grammar. Check the language reference against the parser source when in doubt.