Open-source platform for 3D AI agents. Turn text or a photo into a rigged, animated GLB avatar, give it an LLM brain, memory and a wallet, and embed it anywhere with one web component. Remote MCP server, x402 payments, Three.js, browser-native.
JavaScript 3d-generationai-agentsanimationavatardigital-humangenerative-aiglbgltfimage-to-3dllmmcpmcp-servermodel-context-protocolsolanatext-to-3dthreejsweb-componentswebglwebxrx402
# syntax=docker/dockerfile:1
# agent-screen-pool: on-demand Chromium caster pool.
#
# The base image ships Chromium plus every system dependency it needs. Its tag
# MUST match the playwright version pinned in package-lock.json: the image bakes
# exactly the browser build that playwright release expects, so a drifted pair
# starts fine and then dies on the first cast with
# "Executable doesn't exist at /ms-playwright/chromium-<rev>". Bump both together.
FROM mcr.microsoft.com/playwright:v1.61.1-noble
WORKDIR /app
# Lockfile-driven install so the image gets the exact playwright the base image
# was built for, not whatever the range resolves to on build day.
COPY package.json package-lock.json ./
RUN npm ci --omit=dev
COPY index.js control.js task-runner.js ./
COPY tasks ./tasks
# Chromium is baked into the base image at /ms-playwright.
ENV PLAYWRIGHT_BROWSERS_PATH=/ms-playwright
ENV NODE_ENV=production \
BASE_URL=https://three.ws \
MAX_BROWSERS=6 \
POLL_MS=3000 \
FRAME_MS=700 \
JPEG_QUALITY=58
# SCREEN_WORKER_SECRET must be supplied at runtime (-e SCREEN_WORKER_SECRET=...).
USER pwuser
CMD ["node", "index.js"]