Nano-Banana Image Generation

VerifiedSafe

Generates images using Google's Nano-Banana (Gemini 2.5 Flash Image) model. Automatically enhances user prompts for better results. Supports resolutions from 1K to 4K, various aspect ratios, and flash/pro model options. Use when the user asks to create or generate an image.

Sby Skills Guide Bot
ContentIntermediate
6/2/2026
Claude Code
#image-generation#gemini#ai-art#prompt-engineering

Recommended for

Automated summary

Generates images using Google's Nano-Banana (Gemini 2.5 Flash) model, automatically enhancing user prompts for better results.

Strengths

  • Automatic prompt enhancement for improved image quality
  • Supports multiple resolutions (1K, 2K, 4K) and aspect ratios
  • Runs in background with macOS notification and auto-open

Limitations

  • Requires a valid Gemini API key set up locally
  • Limited to Gemini 2.5 Flash model (no other engines)
When to use it

Use this skill when the user requests generating an image, illustration, or any visual content.

When not to use it

Do not use if an actual image needs to be edited or if the user asks for video or animation.

Security analysis

Safe
Quality score90/100

The skill only invokes a dedicated shell script for image generation with user-provided prompts. No destructive commands, network exfiltration, or obfuscated payloads are present. The troubleshooting section suggests manual keychain access, but the skill does not execute that automatically, so risk is minimal.

No concerns found

How this analysis is made

Examples

Basic image generation
Generate a cat
High-resolution landscape
Create a 4K image of mountains
Widescreen wallpaper with save location
Generate a sunset, save to desktop

name: nano-banana description: Generate images using Google's Nano-Banana (Gemini 2.5 Flash Image) model. Use when the user asks to generate, create, or make an image, picture, or illustration. allowed-tools: Read, Bash

Nano-Banana Image Generation

Generate images using Google's Gemini Image models.

IMPORTANT: Enhance User Prompts

Before generating, always enhance the user's prompt using the techniques in PROMPTING.md.

Quick Enhancement Formula

[Subject] + [Style] + [Lighting] + [Composition] + [Quality]

Example Enhancement

  • User says: "a cat"
  • Enhanced: "A fluffy orange tabby cat with bright green eyes, photorealistic, soft natural window lighting, shallow depth of field, rule of thirds composition"

For detailed prompting techniques, templates, and style keywords, see PROMPTING.md.

Quick Start

~/.claude/skills/nano-banana/scripts/generate.sh "your prompt here"

Options

| Option | Description | Default | |--------|-------------|---------| | --output PATH | Save to specific file | ~/Pictures/nano-banana/<timestamp>.png | | --size SIZE | Resolution: 1K, 2K, 4K | 1K | | --aspect RATIO | Aspect ratio | 1:1 | | --flash | Use Flash model (faster, lower quality) | Pro |

Aspect Ratios

1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3

Resolutions

  • 1K - Standard (default, fast)
  • 2K - High resolution (requires Pro model)
  • 4K - Maximum resolution (requires Pro model)

Interpreting User Requests

Parse the user's request and map to script options:

| User says | Script call | |-----------|-------------| | "Generate a cat" | generate.sh "a cat" | | "Create a 4K image of mountains" | generate.sh "mountains" --size 4K | | "Make a landscape wallpaper" | generate.sh "landscape" --aspect 16:9 | | "High quality portrait of a robot" | generate.sh "portrait of a robot" --pro --aspect 3:4 | | "Generate a sunset, save to desktop" | generate.sh "a sunset" --output ~/Desktop/sunset.png | | "4K widescreen cyberpunk city" | generate.sh "cyberpunk city" --size 4K --aspect 16:9 |

Keywords to watch for:

  • Resolution: "4K", "high res", "high resolution", "2K", "large" → use --size
  • Quality: "high quality", "detailed", "professional" → use --pro
  • Orientation: "landscape", "widescreen", "wallpaper" → --aspect 16:9
  • Orientation: "portrait", "vertical", "phone wallpaper" → --aspect 9:16
  • Save location: "save to", "put it in", "output to" → --output

Behavior

  • Runs in background (won't block Claude)
  • Sends macOS notification when complete
  • Auto-opens the image when done
  • Images saved to ~/Pictures/nano-banana/ by default

Troubleshooting

Check the error log:

cat ~/.claude/skills/nano-banana/last-error.log

Verify API key:

security find-generic-password -a "$USER" -s "GEMINI_API_KEY" -w | head -c 10

From the file to a first result

The download contains exactly the content displayed on this page. This copy may differ from the latest version in the source repository.

View the source and supporting files

This download contains SKILL.md only. Any scripts, references, templates and dependencies are not included. Get the required files from the source, preserving their directory structure, before continuing.

1. Download the file

Download SKILL.md

From the root of a test project, save the file as .skill-downloads/nano-banana-image-generation-mm8o51o0/SKILL.md, or use the command below. This folder is for inspecting the skill before installation.

Command for macOS/Linux or WSL: requires a POSIX shell and curl. On Windows without WSL, use the download button and create the folder manually.

mkdir -p '.skill-downloads/nano-banana-image-generation-mm8o51o0' && curl --fail --show-error --location 'https://skills-guides.com/api/skills/nano-banana-image-generation-mm8o51o0/download' --output '.skill-downloads/nano-banana-image-generation-mm8o51o0/SKILL.md.part' && mv '.skill-downloads/nano-banana-image-generation-mm8o51o0/SKILL.md.part' '.skill-downloads/nano-banana-image-generation-mm8o51o0/SKILL.md'

2. Check prerequisites

Read SKILL.md before using it. If it references scripts or external services, inspect them and prepare the required access in your test environment.

Tools declared by the author (check availability in your agent): Read, Bash

No recognized prerequisites section in this file. This does not mean it is self-contained: libraries, tools and access requirements may be mentioned elsewhere in the content or source repository.

The platforms on this page are catalog indications. The paths below follow tool documentation; they do not prove this skill has been run in those tools. No minimum tool version is confirmed for this listing.

3. Load the skill

After inspection, copy the complete folder into your project using one of the commands below, then open a new tool session. The command stops if a folder with the same name already exists: compare the files before replacing it.

Claude Code
test -f '.skill-downloads/nano-banana-image-generation-mm8o51o0/SKILL.md' && test -s '.skill-downloads/nano-banana-image-generation-mm8o51o0/SKILL.md' && mkdir -p '.claude/skills' && test ! -e '.claude/skills/nano-banana' && cp -R '.skill-downloads/nano-banana-image-generation-mm8o51o0' '.claude/skills/nano-banana'

OpenSkills

OpenSkills must already be installed. The local install/read procedure is checked with OpenSkills 1.5.0, which requires Node.js ≥ 20.6.0. Catalog slugs are not resolved automatically.

Official OpenSkills installation
openskills install './.skill-downloads/nano-banana-image-generation-mm8o51o0'

openskills read 'nano-banana-image-generation-mm8o51o0'

Reading should display the installed file content and its base directory. It does not run the skill or install its dependencies.

For an agent that reads AGENTS.md, then sync the index. This command changes AGENTS.md: review the diff before using it.

openskills sync

Manual method or another tool: attach SKILL.md to a conversation, or open it in an agent that can read files. Include any required supporting resources. No universal folder is recognized by every tool.

4. Try a first task

Use a short, synthetic input: an API endpoint, a small document or an interface brief. In a new conversation, attach the file or give the agent its path, then adapt this request:

Read .skill-downloads/nano-banana-image-generation-mm8o51o0/SKILL.md and apply its instructions to this example: [describe a test input]. Identify any missing files, dependencies or access. Produce the result requested by the skill without inventing data or running unreviewed scripts.

Expected result: output matching the skill instructions, or a precise list of missing prerequisites. A read confirmation alone does not validate a first use. If an external resource is unavailable, that path remains untested.

Related skills