ProductFlow

Lazy cat OIDC login has been integrated. ##Functional characteristics - Product workbench: supports product list, paginated browsing, new product and details workbench; the product deletion entry is closed in the current measured deployment. - Node Canvas: Organize product data, reference drawings, copy nodes and production drawings nodes in the same canvas, supporting drag and drop, zoom, connection, multiple selection and mobile touch operations. - AI copywriting and template poster: Generate structured copywriting based on the product context, and continue to feed copywriting and reference drawings to subsequent posters or graphics nodes. - Continuous picture generation sessions: Support reference picture uploading, historical picture reuse, multi-candidate comparison, cancellation/retry, download and casting to gallery. - Gallery archiving: Centrally save the generated results, retaining the source session, associated products, prompt words, size, model and download portal. - Run-time configuration: `/settings` supports switching between provider, model, image size, prompt template, upload restrictions and global concurrency, and is protected twice by independently unlocking tokens. ###Usage Tutorial This is the first time to enter ProductFlow, and it is recommended to complete it in the following order. This can not only complete the basic configuration, but also quickly produce the first available picture. 1. ** Log in to enter the workbench ** - Open the application home page. - Click "Log in using lazy cat account" to enter the background. - After successfully logging in, you will enter `/products`, which is `Products/Workstation` in the top navigation. 2. ** Unlock the configuration page first ** - Click `Configure` at the top. - Enter the settings_access_token set during installation. - After the unlocking is successful, you can see the configuration groups on the left: `Supplier Files`,`Copy Generation`,`Picture Generation`,`Prompt Words`,`Picture Generation Parameters`,`Generation Queue`,`Security`,`Import and Export`. 3. ** Check whether the copy and picture model are ready ** - First look at `Configuration-> Copy Generation` and confirm: - API interface - supplier file - model version - Look at Configuration-> Image Generation again and confirm: - API interface - supplier file - model version - If you are using the OpenAI Responses solution, you must also confirm whether the background response mode is turned on. - If you use OpenAI to be compatible with upstream, the Base URL in the supplier file generally contains `/v1`; for example,`https://aiinterface.site/v1`. 4. ** Create the first product ** - Go back to `Product/workbench` and click `New Product`. - Upload a main product map, fill in the product name, and select a template. It is recommended to directly use the 'E-commerce Main Map'. - After the creation is completed, you will enter the product details workbench. - Note: If you need to delete, you need to go to the Settings-Security section first, open the permissions and save them before you can delete them. 5. ** Supplement product information ** - Add in the details panel on the right: - category - price - part description - The page will be automatically saved; as long as the fields can be read back normally, it means that the product can already participate in the copywriting and drawing process. 6. ** Run a complete workflow once ** - On the product details page, click `Run entire workflow`. - Under normal circumstances, the nodes on the canvas will advance in turn: - `Commodity information` - `Main picture selling points` - `Generate main picture` - `Main picture output` - `Refining the main picture` - `Refining output` - If you fail halfway, first look at the actual error reported in the 'Log' panel on the right, and then decide whether to adjust the prompt or go back to the configuration page to check the model. 7. ** Check the results and decide whether to continue to make up the map ** - If this round of workflow has produced satisfactory pictures, you can continue to keep them and use them in the product details. - If you still want to change your style, composition, or continue to do secondary creations, go to the 'Wen/Tu Generation Picture' at the top. 8. ** Make additional generation in `Text/Picture Generation`* - It is recommended to create a new session directly for the first time. - Select the target product and add the prompt words; if necessary, you can also upload the session reference map. - Click `Start Generation` and wait for the candidate graph to be returned from the current result area. - If the old session appears in a state such as the button is grayed out or a historical base image is required, priority will be given to creating a new session and retry. 9. ** Submit satisfactory results to the gallery ** - In the current result area of the `Text/Picture Generation Picture ', select the candidate picture that you are satisfied with. - Click `Cast to Gallery`. - Then open the top gallery and confirm that the work has been archived. - Note: Successful workflow production will not automatically enter the gallery, and will only appear here after manual delivery.