hero8152/Infinite-Canvas

The gap in what exists

While many AI image generators exist, few offer a truly freeform, collaborative, and composable infinite canvas experience directly integrated with advanced backends like ComfyUI for prosumers.

Wedge to win

Target creative prosumers and small design agencies who value a flexible, composable workspace for AI image generation over rigid standalone tools, emphasizing the ComfyUI/API integration and collaborative features.

Reputation value (worth doing free for proof) — 70/100

Launching a well-executed, open-source infinite canvas for AI art, especially with robust backend integrations, would be a strong 'Show HN' and proof-of-work for AI-tool and creative-tool development.

Likely monetization

SaaS subscription (per user / project slots)

Incumbents to beat

ExcalidrawMiroLeonardo.aiMidjourney

You are an expert full-stack developer. Your task is to build a web-based 'Infinite Canvas' application, similar to the concept of hero8152/Infinite-Canvas, focusing on generative AI image creation. The application should allow users to pan, zoom, and place 'nodes' on an infinite canvas. Each node can be an image, a text prompt, or a workflow definition. The core functionality must involve integrating with external AI image generation services. 

Use the following stack: Next.js 16 App Router, React 19, Tailwind v4, AI SDK v6 with Gemini (for potential future prompt generation or embedding analysis), Neon Postgres on Vercel for data persistence. For the UI components, use Shadcn UI. For the infinite canvas itself, research and integrate a suitable React library like 'react-flow-renderer' or similar alternatives.

**MVP Slice:**
1.  **Basic Canvas:** Implement an infinite canvas with panning and zooming capabilities.
2.  **Image Node Creation:** Allow users to drag-and-drop or upload images to create image nodes on the canvas. These images should be stored in cloud storage (e.g., Vercel Blob or S3).
3.  **Prompt Node Creation:** Allow users to create text nodes for prompts.
4.  **Basic AI Integration:** Implement a server-side API endpoint that accepts a text prompt and uses a public generative AI API (e.g., stability.ai, replicate.com) to generate an image. Allow this generated image to be placed back on the canvas as an image node.
5.  **User Accounts:** Implement basic user registration and login using NextAuth.js or Clerk.
6.  **Persistence:** Save and load canvases for authenticated users to/from Neon Postgres. 

**Build and Verify Gate:**
-   A user can sign up, create a new canvas, add a text prompt, trigger an AI image generation, and see the result appear as an image node on their persistent canvas. The canvas should be savable and loadable.
-   The UI should be responsive and visually appealing using Tailwind CSS and Shadcn UI components.