nanobanana_skill

Overview

This skill lets you generate and edit images using Google Gemini 3 Pro Image (Nano Banana Pro). It supports text-to-image, image editing, multiple aspect ratios, and high-resolution 2K/4K outputs for production-ready visuals. Designed for fast iteration and batch generation, it fits creative work, product imagery, and marketing visuals.

How this skill works

The skill calls the Gemini image model with your natural-language prompt and optional input image to produce new or edited images. You can specify aspect ratio and output size, enable Google Search grounding for factual prompts, or run batch jobs to create many variations. Outputs are saved to a configurable directory and can be invoked from command-line scripts or the Python API.

When to use it

• Create photorealistic or stylized images from text prompts
• Edit or enhance existing images with natural-language instructions
• Produce high-resolution assets for marketing, product photos, or prints
• Generate multiple variations for A/B testing or design exploration
• Create social, web, or cinematic aspect-ratio images quickly

Best practices

• Write prompts with subject, style, lighting, composition, and color palette for predictable results
• Start with smaller batches and iterate prompts before scaling to 2K/4K
• Use Google Search grounding for images that require factual accuracy
• Add short delays (3–5s) between batch requests to avoid rate limits
• Review and rephrase prompts if outputs trigger safety filters

Example use cases

• Design a pixel-art mascot or brand logo with specific color palette and mood
• Edit a product shot: change background, adjust lighting, or add props for e-commerce
• Create cinematic 21:9 landscape images for banners and hero sections
• Batch-generate 10–20 concept variations to present to stakeholders
• Produce 4K lifestyle images for print or high-resolution advertising