Stats
Actions
Tags
From nano-banana-ultimate
Generates illustrations in flat vector, watercolor, line art, 3D clay render, pixel art, and isometric styles using Gemini API. Auto-detects style from context; triggers on illustration keywords.
How this skill is triggered — by the user, by Claude, or both
Slash command
/nano-banana-ultimate:banana-illustrationThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
**Input:** `$ARGUMENTS` (optional — subject and style description)
Input: $ARGUMENTS (optional — subject and style description)
If $ARGUMENTS is provided, use it as the illustration brief and skip to prompt building. If empty, ask about subject and style.
uv installedGEMINI_API_KEY environment variable set (get one at https://aistudio.google.com/apikey)[Subject/Character] + [Narrative Action] + [Illustrated Environment] + [Balanced Composition] + [Explicit Art Style]
| User Keywords | Resolved Style |
|---|---|
| "flat" / "vector" | Flat vector with solid fills, clean edges |
| "watercolor" | Wet-on-wet watercolor with paper texture, visible brushstrokes |
| "3D" / "clay" | Pixar-style clay render, soft shadows |
| "pixel" | Pixel art with visible grid, limited palette |
| "isometric" | Isometric projection, no perspective vanishing points |
| "line art" / "ink" | Consistent weight ink lines, minimal fills |
| Ambiguous | Default to flat vector |
| Domain | Aspect Ratio | Resolution |
|---|---|---|
| Spot / Icon | 1:1 | 1K |
| Scene | 4:3 | 1K |
uv run "${CLAUDE_PLUGIN_ROOT}/scripts/banana.py" \
--prompt "<detailed prompt following the formula>" \
--output "<YYYY-MM-DD-HH-MM-SS-descriptive-name>.png" \
--resolution 1K \
--aspect-ratio <ratio from Domain Defaults> \
--model gemini-3-pro-image-preview
When the user provides an existing image to restyle or modify:
uv run "${CLAUDE_PLUGIN_ROOT}/scripts/banana.py" \
--prompt "<edit instruction: what to change AND what to keep>" \
--output "<YYYY-MM-DD-HH-MM-SS-descriptive-name>.png" \
--input-image "/path/to/source.png" \
--aspect-ratio <ratio>
For style transfers, describe the target style and what to preserve:
"Recreate this photograph as a flat vector illustration with a limited palette of teal, coral, and cream. Keep the composition, subject pose, and environment layout exactly the same."
Up to 14 input images can be passed (repeat --input-image for each).
See references/prompts.md for ready-to-use example prompts covering common illustration styles and scenarios.
npx claudepluginhub agentic-dev3o/devx-plugins --plugin nano-banana-ultimateDefines illustration style guides covering visual styles, color palettes, character designs, usage rules, and best practices for consistent product UI visuals.
Analyzes article structure, identifies positions requiring visual aids, and generates consistent illustrations using a Type × Style × Palette approach. Invoked when users request article illustration, image generation, or 为文章配图.