Skip to content

wassermanproductions/master-canvas

Repository files navigation

Master Canvas

License: MIT Latest release Platforms Support Sam Wasserman on Ko-fi

Master Canvas

Local-first pre-production canvas for planning AI video projects. Arrange images, prompts, references, music notes, shot order, and generation handoffs on a visual board.

Developed by Sam Wasserman: WassermanProductions.com and Wasserman.AI. Released under the MIT License.

Master Canvas ships without bundled demo projects, private media, or sample user assets. New users start with a blank local canvas.

Install (macOS)

Apple Silicon (M1–M4):

curl -fsSL https://raw.githubusercontent.com/wassermanproductions/master-canvas/main/install.sh | bash

This downloads the latest release, installs Master Canvas to Applications, and launches it. Intel Macs can build from source (see Desktop App).

Preview

These screenshots show an example production board with assets, shot cards, prompts, references, handoff readiness, the movable help key, and the minimap overview. The app itself starts blank for new users.

Master Canvas board with help key open

Master Canvas board with selected shot prompt inspector

Features

  • Infinite visual canvas with zoom, pan, lasso selection, grouping, and node connections.
  • Local asset library for images, video, audio, links, style references, and music references.
  • Per-shot prompts, negative prompts, lenses, lighting, camera movement, action, sound notes, and review notes.
  • Handoff exports for Markdown, JSON, visual storyboard HTML, storyboard PDF, and a ZIP package for agents/operators.
  • Handoff ZIP includes Hermes Agent context, ComfyUI/LTX job JSON, Kling/Veo prompt sheets, source assets, shot order CSV, and scene bin plan.
  • Local-first storage in the user's browser/Electron app. No account or cloud service is required.

Privacy

All project data is stored locally in the user's browser or desktop app storage. The app does not include a backend and does not upload files anywhere. Exported ZIP/PDF/HTML/JSON files are created locally by the user.

Web App

npm install
npm run dev

Open the local URL printed by Vite.

Desktop App

npm install
npm run desktop

Package a desktop build:

npm run desktop:dir

Create distributable installers:

npm run desktop:dist

Installer output is written to release/.

For unsigned local builds on macOS, use npm run desktop:dir. For public distribution, sign and notarize the macOS app with your own Apple Developer ID.

More details:

Publishing To GitHub

This repository is set up to start from a blank canvas. Demo/private project assets are not required for the app to run.

Recommended before publishing:

  • Decide whether packaged releases should be built manually or through GitHub Actions.
  • Keep private project handoff ZIPs out of the repo.

Agent control (MCP)

Any MCP agent — Hermes, Claude Code, Codex, or any other MCP client — can read and edit a Master Canvas project (boards, cards, prompts, references, shot order, assets) and build generator-ready handoff packages headlessly, without the desktop app. It operates on the app's exported master-canvas-project.json (export from the app → let the agent work → re-import). The server lives in mcp/ here and is also published standalone as master-canvas-mcp.

# Claude Code
claude mcp add master-canvas -- node /absolute/path/to/master-canvas/mcp/master-canvas-mcp.mjs

See mcp/README.md for the full tool list and Hermes/Codex/generic setup.

Hermes Agent Plugin

A starter native Hermes plugin lives in:

plugins/mastercanvas-hermes

It lets Hermes inspect/extract Master Canvas handoff ZIPs and turn them into a ComfyUI/LTX execution plan. Copy that folder into ~/.hermes/plugins/master-canvas and restart Hermes.

Support

A few people asked if they could send tips to support my work developing open source tools. So I set up an optional way in case anyone wants to.

No pressure at all. Using the apps, sharing them, starring the repositories, and contributing code all help too. Thank you.

About

Open Source Desktop App for Local-first pre-production canvas for AI video planning, prompts, assets, and handoff packages.

Resources

License

Stars

30 stars

Watchers

1 watching

Forks

Packages

 
 
 

Contributors