infinite-canvas
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
Infinite Canvas — A live spatial board for AI and people
An AI-assisted freeform canvas for mind maps, architecture diagrams, research boards, project plans, customer journeys, and agent designs. Models and browser users edit the same positioned cards and connections incrementally—without generating Mermaid source.
Features
Shared live board — MCP tools and the browser edit the same canvas
Freeform cards — create, select, edit, drag, and move cards with the keyboard
Directed connections — link cards with optional labels
Canvas navigation — pan, zoom, and reset the viewport
Live synchronization — browser polling picks up model changes once per second
Portable exports — download editable JSON or standalone SVG
Dependency-free UI — the browser editor is plain HTML, CSS, and JavaScript
Related MCP server: sketchbord-mcp
Tools
Tool | Description |
| Return a canvas's revision, positioned cards, and connections |
| Add a card, creating a canvas when |
| Patch selected card fields or its position |
| Add an optionally labeled, directed connection |
| Export editable JSON or standalone SVG |
How it works
Start a board by calling add_node without a canvas_id. The result includes the generated canvas ID. Open the corresponding browser board and pass that ID to later tool calls:
https://YOUR_HOST/?canvas=CANVAS_IDThe MCP endpoint is:
https://YOUR_HOST/mcpInfinite Canvas is a standalone collaborative browser experience, not an embedded MCP App widget.
Local development
Requires Python 3.11 or newer.
git clone https://github.com/skundu42/infinite-canvas.git
cd infinite-canvas
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
python server.pyOpen the canvas at http://localhost:8000. Visit http://localhost:8000/mcp in a browser for the connection guide, or use that same URL in an MCP client.
Run the test suite:
python -m unittest -vHOST and PORT override the default 0.0.0.0:8000 listener.
Deploy to Manufact
Install the Manufact GitHub App for this repository, then deploy from the project directory:
npx mcp-use login
npx mcp-use deploy --name infinite-canvas --start-command "python server.py" --openOnce the deployment is running, copy its generated MCP URL from the Manufact dashboard.
Setup on ChatGPT
Open Settings > Apps and Connectors > Advanced Settings and enable Developer Mode
Go to Connectors > Create, name it "Infinite Canvas," and paste the hosted
/mcpURLIn a new chat, click + > More and select the Infinite Canvas connector
Setup on Claude
Open Settings > Connectors > Add custom connector
Paste the hosted
/mcpURL and saveThe Infinite Canvas tools will be available in new conversations
Built with
mcp-use — MCP server framework
Python standard library — canvas state, validation, JSON, and SVG generation
Plain HTML, CSS, and JavaScript — browser canvas
Current limits
State is intentionally process-local and resets when the server restarts. Canvas IDs isolate boards within one process, but this template does not provide authentication, durable storage, deletion, or multi-worker synchronization.
Related MCP Connectors
Create, read and live-edit visual boards, Kanban plans, Gantt timelines and diagrams with AI agents.
A live board you share with Claude, ChatGPT and Cursor: notes, frames, arrows and comments.
Read, create and edit noddle diagram boards, browse versions and comment as an AI collaborator.
Whiteboard where any shape opens into another canvas. Claude Code and Codex plugins render plans on it.
Related MCP Servers
- AlicenseNot gradedqualityBmaintenanceEnables AI agents to collaboratively read, write, draw, and diagram on a shared real-time whiteboard canvas with live human edits.1MIT
- AlicenseNot gradedqualityCmaintenanceEnables AI agents to visualize tasks on a collaborative Sketchbord whiteboard by composing content into diagrams, editing them incrementally, and reading back the board including user-drawn additions.MIT
- AlicenseAqualityBmaintenanceEnables AI assistants to read and edit a spatial canvas board of notes, tasks and ideas through the running desktop app, creating, updating, linking, grouping and arranging nodes, importing files, focusing the view and undoing changes so edits appear instantly and stay reversible.2510 npm1MIT
- AlicenseNot gradedqualityAmaintenanceEnables agents to draw on a self-hosted canvas by creating and editing frames, text, shapes, images, and SVGs with auto-layout, then inspecting, exporting, and sharing the results. Also supports live viewing of agent edits, pinned comments, screenshots, and revision history so people and agents can collaborate on boards.1,034 npm1MIT