skills.rest
quocduvp
quocduvpCOMMUNITY
Explore all skills by this author →
GitHub

image-analysis

Analyze reference images to reconstruct Nano Banana 2 prompts and generate variant prompts.

1|Updated Mar 22, 2026
Search Tags:#analysis#clone#image-analysis#prompt-generation#variant#visual-prompt#nano-banana-2
One-click install
npx skills add https://github.com/quocduvp/nanobanana2-skills --skill image-analysis-quocduvp
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: image-analysis
Source: https://github.com/quocduvp/nanobanana2-skills/tree/main/skills/image-analysis
Command: npx skills add https://github.com/quocduvp/nanobanana2-skills --skill image-analysis-quocduvp

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Analyses a user-provided image to reconstruct a production-ready Nano Banana 2 prompt, then generates clones or variant prompts on demand, preserving visual intent and style across outputs.

Core Features & Use Cases

  • Analyze reference images across Subject, Clothing / Product, Environment, Lighting, Camera & Lens, Color Grade, and Composition to inform prompt creation.
  • Reconstruct a production-ready Nano Banana 2 prompt from the analyzed visuals.
  • Generate clone or variant prompts on demand, including style transfer, scene variants, lighting variants, outfit/color variants, mood variants, and format variants.
  • Follow a phase-based workflow with explicit confidence notes and clear next-step options, enabling seamless handoffs to related skills (e.g., video-motion, tiktok-ad-creator).
  • Support a structured variant planning process (Phase 3–4) to ensure consistent outputs and easy iteration.

Quick Start

Upload an image or reference and I will analyze it, reconstruct a Nano Banana 2 prompt, and generate clone/variant prompts.

Frequently Asked Questions about image-analysis

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I convert an image into a Nano Banana 2 prompt?▼

To convert an image into a Nano Banana 2 prompt, upload your reference image. The tool analyzes visual elements like subject, lighting, and composition, then reconstructs a production-ready prompt for immediate use.

What image elements are analyzed to reconstruct a visual prompt?▼

Visual prompt reconstruction analyzes Subject, Clothing/Product, Environment, Lighting, Camera & Lens, Color Grade, and Composition. This structured breakdown ensures the generated prompt accurately reflects the original reference image.

Can I generate style or lighting variants from an existing image prompt?▼

Yes, you can generate variants from an existing image prompt. The system supports on-demand generation of style transfers, scene variants, lighting variants, outfit/color variants, mood variants, and format variants.

Does the image analysis tool flag uncertain visual elements?▼

Yes, the image analysis tool flags uncertain visual elements using explicit confidence notes. It highlights ambiguous details and suggests follow-up actions or alternative skills to ensure accurate prompt generation.

What is the best way to create consistent image variants for a series?▼

The best way to create consistent image variants is using the structured variant planning process. It guides you through phase-based workflows with explicit options, ensuring consistent outputs and easy iteration across a series.

Related Skills

Explore More Design & Creative Skills→
NikiforovAllNikiforovAll
137

nano-banana-prompting

Craft structured prompts for Nano Banana Pro image generation.

Community
Advanced
gooseworks-aigooseworks-ai
6

create-product-images-nanobanana

Generate Nano Banana style product, avatar, b-roll, or stop-motion images from custom prompts.

Official
Advanced
eaglhuangeaglhuang
3

nano-banana-gen

Generate images from prompts via Gemini Nano Banana models and save local artifacts.

Community
Intermediate
darknetopencv-altdarknetopencv-alt
34

image-generation

Generate images from structured JSON prompts and reference images via Python scripts.

Community
Intermediate