AgentCanvas

The infinite canvas your AI agents control

A hosted infinite canvas that headless AI agents drive over MCP. Any MCP-speaking agent - Claude Code, Codex, Cursor, or a script - creates HTML docs, images, and video on a live canvas, streamed in as it builds.

Open AgentCanvas See how it works

Use it for real

Point your agent at the canvas

AgentCanvas is a live product, not a demo. Connect Claude Code, Codex, Cursor, or any MCP-speaking script to the stdio MCP server and have the agent create HTML docs, images, and video on a real infinite canvas.

Open AgentCanvas See the MCP server

mcp-config.json

{
  "mcpServers": {
    "agentcanvas": {
      "command": "node",
      "args": ["mcp/server.mjs"],
      "env": {
        "CANVAS_API_URL": "https://agentcanvas-iota.vercel.app",
        "DD_API_KEY": "<your-dd-api-key>"
      }
    }
  }
}

Image generation is metered through the platform. Call /api/agentcanvas/generate-image to spend your universal DD credits.

How it works

Agent to MCP to a live canvas

Connect over MCP

Point any MCP-speaking harness - Claude Code, Codex, Cursor, or a plain script - at the stdio MCP server. No OAuth needed for local use.

Agents create assets

The agent calls tools to place HTML docs in sandboxed iframes, images, and video onto a canvas, or generates images from text via Fal.

Watch it stream in

append_html and stream_html_demo push content in chunks over SSE, so you watch docs assemble live rather than popping in whole.

MCP tool surface

The tools your agent calls

Every tool the MCP server exposes. Any harness that speaks MCP can call them to drive the canvas.

canvas_guide

Progressive-disclosure skill: overview plus per-topic how-to for html, image, video, streaming, and research.

get_design_system

Loads the design system into context so every asset comes out on-brand.

list_canvases / create_canvas

Work across multiple canvases, each with its own id and URL.

list_assets

List the assets currently on a canvas.

create_html_asset

Place a doc or slide inside a sandboxed iframe.

create_image_asset / create_video_asset

Place media on the canvas by URL.

generate_image

Text-to-image via Fal, using the model from the connectors registry.

append_html / stream_html_demo

Stream content in chunks for the live build effect.

update_asset / delete_asset / clear_canvas

Edit, remove, or reset assets on a canvas.

list_connectors

Enabled research, image, and video connectors - the "MCP of MCPs".

Get started

Point your agent at a canvas

Connect Claude Code, Codex, Cursor, or a script over MCP and watch it build docs, images, and video onto a live infinite canvas.

Open AgentCanvas