teal-picture

As a user I want to be able to see an interactive domino galaxy showcase on the home page with a gallery entry point so that I can explore the app and navigate to the camera or history sections. Implement based on the existing Home (v2) JSX design.

As a user I want to be able to take a photo of multiple dominoes using the camera page so that I can submit the image for pip count analysis and AI rendering. Implement based on the existing Camera (v2) JSX design.

As a user I want to be able to view the pip count for each detected domino alongside its AI-rendered image on the results page so that I can review the analysis output clearly. Implement based on the existing Results (v2) JSX design.

As a user I want to be able to browse, save, and review my history of saved pip counts and AI-rendered domino images on the history page so that I can revisit past analyses at any time. Implement based on the existing History (v2) JSX design.

As a backend developer I want to be able to define and implement the MySQL database schema for storing domino analysis sessions, individual domino pip counts, AI-rendered image URLs, and history records so that all data is persistently and efficiently stored.

As a user I want to be able to upload a captured domino image via the FastAPI backend so that the image can be processed for pip detection and AI rendering.

As a developer I want to be able to set up Langchain workflow orchestration with Litellm routing to coordinate GPT 5.4 pip validation and Google Nano Banana image generation in a unified AI pipeline so that both AI models operate reliably and efficiently.

As a user I want to be able to have my captured domino photo automatically sent to the backend image upload endpoint from the camera page so that the analysis pipeline is triggered seamlessly.

As a developer I want to be able to configure Litellm to route requests between GPT 5.4 and Google Nano Banana models with proper fallback and load handling so that the AI pipeline is resilient and model-agnostic.

As a user I want to be able to have my domino pip counts validated by GPT 5.4 via Litellm routing and Langchain orchestration so that the detected pip values are accurate and reliable.

As a user I want to be able to have multiple individual dominoes automatically detected and segmented from my uploaded photo so that each domino is independently analyzed for pip count and AI rendering.

As a user I want to be able to receive an AI-rendered image of each detected domino generated by Google Nano Banana via Langchain orchestration so that I can view a clean visual representation alongside my pip count.

As a user I want to be able to retrieve the full analysis results including pip counts per domino and AI-rendered image URLs from the FastAPI backend so that the results page can display accurate data.

As a user I want to be able to save my pip count results and AI-rendered images to the history via the FastAPI backend so that my analyses are persisted and retrievable later.

As a user I want to be able to see live pip counts and AI-rendered images on the results page fetched from the backend analysis results API so that my results are dynamically populated after capture.

As a user I want to be able to fetch my full history of saved pip counts and AI-rendered domino images from the FastAPI backend so that the history page displays all past analyses.

As a user I want to be able to see my saved history of pip counts and AI-rendered images on the history page fetched from the backend so that I can browse all past domino analyses.

As a user I want to be able to view a curated gallery of recent AI-rendered domino images on the home page pulled from my history so that the home page feels dynamic and personalized.