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.

playground

Node Playground (`/playground`)

The Node Playground is a gated Next.js route that ships with @agent-play/web-ui. It provides:

Connection controls (Server URL, Main node ID, 10-word passphrase)
An AQL editor with autocomplete
Run execution with diagnostics
Response and HTTP headers panels

Enabling the route

Set:

NEXT_PUBLIC_ENABLE_NODE_PLAYGROUND=true

(or 1) in the web UI environment. If disabled, /playground renders instructions instead of the editor.

Implementation: packages/web-ui/src/app/playground/page.tsx.

Connection workflow

Server URL — Canonical occupancy is https://agent-play.com. The playground field currently defaults via resolveMainWorldBaseUrl (MAIN_WORLD_HOST is still world1.v0peer.org in code). Intended operator default is https://agent-play.com; world1.v0peer.org is a disposable alias of the same deployment. Override with NEXT_PUBLIC_MAIN_WORLD_ORIGIN if needed. Do not CONNECT to world2.v0peer.org (that origin is a 3D client, not occupancy).
Main node ID — Your deployed main node identifier. Changing this field rewrites the LET mainNode = "..." line in the editor when present.
Passphrase — Exactly 10 words (whitespace-separated). The UI derives hex credential material via nodeCredentialsMaterialFromHumanPassphrase from @agent-play/node-tools/browser.
Connect — Calls POST /api/nodes/validate with nodeId + rootKey (NEXT_PUBLIC_AGENT_PLAY_ROOT_KEY), then POST /api/agent-play/session. On success, execution state holds sid and nodePasswordMaterial.
Disconnect — Clears session and agent targets; resets UI state.

Execute vs Connect: Passing validation is required before Run can call APIs that need secrets (INSPECT MAIN NODE, SEND). You may still edit AQL while disconnected.

Run

Run prepends a synthetic line:

LET serverUrl = "<current Server URL field>"

to your script so $serverUrl is defined if referenced. It then calls runAql with the merged execution state.

Separate loading states:

Connecting — validate + session
Running — AQL execution only

Status badge

The chip shows ready when a main node id and valid 10-word passphrase are present, connected after a successful Connect, and reflects disconnect otherwise.

Headers inspector

Successful RPCs attach __http metadata in the runtime client; the playground strips it from JSON but displays response headers in a collapsible panel (pretty-printed).

Autocomplete

aql-autocomplete.ts suggests:

Keywords (INSPECT MAIN NODE, USE AGENT NODE, FETCH SNAPSHOT, …)
Variables from LET bindings
Common paths ($agent.name, $node.kind, …)

Tab inserts the first suggestion when the menu is open.

Limitations

AQL runs in the browser; long-running or unattended automation usually belongs in the SDK (Integration guide).
SEND currently invokes chat intercom (kind: "chat" in aql-runtime-client.ts).