How to use the site

made by citruzss is a site for prompts, styles, moodboards, and a built-in AI generator. You can use ready presets, write your own ideas, attach references, and build visuals directly on the site.

The beta includes image generation, a Gemini text chat, chat history, saved results, favorites, personal styles, and a moodboard for references.

Under Still images you’ll find presets as a grid of cards.

Tap a style card: a panel opens on the right (full screen on mobile) with samples and the prompt. “Copy” sends the text to your clipboard.

There’s a negative prompt, gallery, and AI adaptation: write an idea, and the model rewrites the prompt in the selected style. At the top: search, sort, and tag filters.

The Generator has two modes: “Photo” for image creation and “Text” for chatting with Gemini. During the beta, Nano Banana 2 and Nano Banana are available; the Pro model is hidden until it becomes stable.

In “Photo”, choose a ready style or “Free”, write a prompt, attach up to 14 references, pick the format, quality, and negative prompt. The “Improve” button turns a short idea into a stronger prompt you can still edit before generating.

In “Text”, ask questions, request prompts for Nano Banana/Gemini, and attach photos for analysis. Each chat has its own history; new images are stored in Supabase Storage, and results can be previewed, copied, or downloaded.

The AI agent sits on the side of most pages. If you are not signed in, it opens sign-in first and then sends you to My styles.

It can create a new style from a reference: the model reads the lighting, atmosphere, color grade, and builds a universal prompt with a negative prompt.

Inside ready-made styles there is an “AI adaptation” button: write your task, and the model rewrites the selected style for your idea.

Motion will hold prompts for movement: camera, pacing, smooth loops.

That section is closed for now — the home card is marked “In development”.

In development

The moodboard is your pre-generation idea board: references, notes, comparing variants. Use the “Why?” button on the board for the full loop.

Select a photo → “To generator” (photo) or “Animate” (video). A good frame in the generator — grid icon — back to the board.

Projects live in this browser on this device. Multiple boards: create, rename, duplicate.

Tap the heart on a style card to save it.

Open the favorites panel with the button at the side bottom to jump back to presets you like.

The header has “Sign in” and account: profile, my styles, moderation.

You can sign in with Google or register with email and password. The account is required for the generator, text chat, cloud chat history, and AI flows around your own styles.

You don’t need an account just to read and copy public prompts. Beta generation is limited to 5 images per day per account/device/IP.

Accounts now work with Google and email. Personal styles still save locally in the browser.

In the header: UI sounds, light/dark theme, RU / EN.

On mobile, some controls live in the ☰ menu.

Do I need to register? Not for reading and copying prompts. Yes for the generator, text chat, chat history, My styles, and the AI agent.

Why can generation take a while or fail? Google image models can be overloaded. The site retries and shows a clear error if the model still does not answer.

Why are my moodboard or personal styles empty on another device? Moodboards and personal styles are still browser-local. Generator history and new chat images sync through your account.