Why does this skill prefer sentences over keywords?

The Gemini 2.5 Flash model is trained to understand context and relationships better than isolated tags. Natural descriptions allow for more nuanced control over lighting, mood, and subject interaction.

How do I refine an image after the first generation?

The skill utilizes a multi-turn editing workflow, encouraging users to make one targeted change at a time (e.g., 'add glasses' or 'change lighting') to maintain consistency.

Can I use this skill for precise camera control?

Yes, it includes specific parameters for lenses (35mm, 50mm, 85mm), aperture settings for depth of field, and professional lighting setups like golden hour or Rembrandt lighting.

What is the Nano Banana model?

Nano Banana is the internal moniker for the Gemini 2.5 Flash Image model, known for its high-speed, high-quality image generation and strong natural language understanding.

How does it handle text inside generated images?

The skill provides a validation framework to keep text under 25 characters, specifies font styles (bold, script, sans-serif), and uses explicit positioning to maximize rendering accuracy.

Nano Banana Image Prompt Generator

Name: Nano Banana Image Prompt Generator
Author: secondsky

bysecondsky

•

Design Tools

Generates high-quality, natural-language image prompts optimized for the Gemini 2.5 Flash Image model using professional photography and design standards.

The nano-banana-prompts skill equips Claude Code with a sophisticated framework for generating and refining prompts for the Gemini 2.5 Flash Image model. It shifts away from chaotic keyword lists in favor of descriptive, natural language that covers lighting, mood, camera settings, and composition. Whether you need photorealistic product shots, watercolor illustrations, or complex landscapes, this skill provides the technical precision—including lens focal lengths and f-stops—needed to achieve professional results. It also includes specialized logic for accurate text rendering and multi-turn iterative editing, making it an essential tool for developers and designers building visual assets.

Key Features

01Optimized text rendering logic for accurate character generation within images

02Specialized templates for photography, art, product shots, and landscapes

03Technical camera controls including lens types, lighting styles, and depth of field

04Iterative multi-turn workflow for refining images through conversational edits

0521 GitHub stars

06Natural language prompt synthesis following 2025 best practices

Use Cases

01Creating custom UI illustrations and brand assets for web development projects

02Generating high-resolution product photography for e-commerce and mockups

03Designing cinematic marketing materials with specific mood and lighting requirements

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add secondsky/claude-skills nano-banana-prompts

For use in Claude.ai and ChatGPT

Download Skill