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.

monorepo

Monorepo layout

The workspace root lists packages/* as npm workspaces. The package @agent-play/sdk contains the published Node SDK under packages/sdk/src (built to dist/ with tsup for npm), integration examples under packages/sdk/examples, and Vitest suites.

The package @agent-play/play-ui is the Vite application under packages/play-ui. It resolves shared math helpers from the SDK source through a Vite alias so the UI and server agree on world bounds and clamping without duplicating files.

Install dependencies once from the repository root with npm install. That links workspaces so local development does not require publishing to npm.

For downstream projects outside this monorepo (for example ~/Documents/agent-service consuming @agent-play/sdk), use the bundled helper to build the publishable packages and symlink them into the consumer in one step:

npm run link:local -- --consumer ~/Documents/agent-service

See Local linking for downstream projects for the full workflow, including the npm run unlink:local reverse and the npm install caveat.

npm publish order (each package depends on the previous where applicable):

@agent-play/node-tools — identity and credentials helpers.
@agent-play/intercom — intercom wire types and Zod parsers (depends only on zod; conceptually second after node-tools for release planning).
@agent-play/geography-mesh — AOI / Yjs geography helpers (depends on yjs + zod; required by play-ui).
@agent-play/sdk — depends on node-tools and intercom.
@agent-play/cli, @agent-play/play-ui, and other consumers.

The GitHub Action .github/workflows/publish-npm.yml builds and publishes in that order when the corresponding path filters or manual toggles apply.

Build order for a full stack release is typically: run tests, build the play UI to packages/play-ui/dist, then build or start web-ui (@agent-play/web-ui). The Next build runs copy-sources so play-ui sources are available under src/canvas/vendor, and the app serves /agent-play/watch and API routes for snapshot, RPC, and SSE.

Root package.json scripts forward common tasks: npm run build builds the SDK, CLI, web UI, and play UI (npm run build:play-ui); npm run test runs tests in every workspace that defines a test script; npm run example runs the primary SDK example from @agent-play/sdk (expects npm run dev for @agent-play/web-ui in another terminal). npm run dev and npm start both start the web UI dev server (@agent-play/web-ui).