ganbo-gab/open-storyboard-canvas

Open Storyboard Canvas – What It Is

Open Storyboard Canvas is a locally‑run desktop application that lets creators build, organise and run AI‑powered image‑ and video‑generation pipelines on a visual “node canvas”. It combines a Tauri‑based GUI (React + TypeScript front‑end, Rust back‑end) with plug‑in support for multiple text‑to‑image / text‑to‑video providers (OpenAI‑compatible APIs, Claude, Dreamina CLI, etc.). The canvas can hold reference media, prompts, generated assets, a 3‑D “director studio” for scene layout, and a conversational Canvas Agent that can manipulate nodes, submit generation jobs and report status.

---

Core Capabilities (as described in the README)

Area	What the app does
Canvas Agent (beta)	Chat‑based assistant on the right‑hand side that can read the current project, create/move/edit nodes, submit generation tasks, track their status, and show a concise inference summary. Supports both manual and auto‑confirm modes.
AI image/video workflow	Create images from scratch, edit with reference images, generate video, and derive storyboard frames. All provider‑level settings (model, resolution, aspect ratio, number of outputs, extra params) are managed centrally. Supports custom OpenAI‑compatible providers, Agnes, and local Dreamina CLI.
Node canvas & tag groups	Upload media, AI‑generated assets, storyboards, panoramas, and director‑studio scenes as nodes. Tag groups let you group references without inflating prompt text. Unlimited canvas size with geometric indexing and caching for performance.
Director studio & panoramas	3‑D scene editor where you place characters, props, lights and cameras. Export screenshots back to the canvas, generate or import panoramas, and record camera moves.
Productivity & diagnostics	Bulk import prompts from spreadsheets, prompt libraries with presets, undo/redo, project persistence, searchable logs with anonymised details, and explicit handling of API keys and cost‑related actions.
Cross‑platform desktop	Packaged for macOS (universal dmg) and Windows (exe). Built with Tauri 2, so the UI runs in a lightweight WebView and the heavy lifting (network calls, file I/O) is done in Rust.

---

How to Get Started

1. Download the latest installer from the GitHub Releases page (separate binaries for macOS and Windows). macOS users may need to bypass Gatekeeper because the app isn’t notarised yet.
2. Configure providers in Settings → My Config – add your OpenAI‑compatible API keys, Claude credentials, or point to a local Dreamina CLI installation.
3. Create a project (or open an existing one) and start adding nodes from the left‑hand panel – upload media, add AI‑image/video nodes, or drop a director‑studio scene.
4. Generate either by clicking the node’s generate button or by invoking the Canvas Agent (chat window) and asking it to “create a sunset background” or similar. The Agent will create the node, submit the request, and update the canvas when the result arrives.
5. Inspect logs on the left side if something fails; the logs are automatically anonymised but can be expanded for debugging.

---

Project Structure (high‑level)

src/                     # TypeScript front‑end
  commands/              # Bridge to Tauri (Rust) commands
  components/            # Reusable UI pieces
  features/canvas/       # Canvas UI, Agent, director studio, node logic
  features/promptLibrary # Prompt library UI & storage
  stores/                # Zustand state management
  i18n/                  # Chinese / English translations
src‑tauri/               # Rust back‑end (Tauri)
  src/commands/          # System‑level commands, network proxy
  src/ai/                # Provider adapters (OpenAI, Claude, Dreamina…)
  tauri.conf.json        # Tauri configuration
docs/                    # User & developer documentation

---

Development Quick‑Start

# Clone & install dependencies
git clone https://github.com/ganbo-gab/open-storyboard-canvas.git
cd open-storyboard-canvas
npm install

# Run the UI in development mode
npm run dev            # hot‑reload front‑end
npm run tauri dev      # launch the desktop app with the dev server

The repo also provides scripts for linting, type‑checking, building, and Rust‑side tests (cargo test).

---

License & Attribution

The code is released under the MIT License. It is a fork of Storyboard‑Copilot (original author henjicc) with explicit permission to continue development and open‑source the derivative work. The original author and upstream repository are credited in the NOTICE file.

---

Who Might Use This?

• Storyboard artists & animators who want a single place to manage prompts, generated assets, and scene layout.
• Content creators building AI‑generated visuals for games, marketing, or video production.
• Developers looking to extend a Tauri‑based AI desktop app (the repo includes clear Rust/TS boundaries and provider adapters).

---

Bottom Line

Open Storyboard Canvas is a genuine, feature‑rich desktop tool that integrates AI image/video generation, a visual node‑based workflow, and a conversational assistant. It is fully open‑source, cross‑platform, and designed for creators who need more control than a simple “prompt‑to‑image” web UI.