nano-banana-pro_skill

This skill helps you generate and edit images using Nano Banana Pro, applying prompts and edits to produce outputs in multiple resolutions.

Shell

3

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill tdhopper/dotfiles2.0 --skill nano-banana-pro

SKILL.md5.9 KB

Overview

This skill generates and edits images using Nano Banana Pro (Gemini 3 Pro Image). It supports text-to-image and image-to-image workflows, with 1K/2K/4K resolution options and optional input-image editing. The script saves PNG output to your current working directory and prints the output path.

How this skill works

You run the provided script from your current working directory and pass a prompt, filename, resolution, and optional --input-image for edits. The tool chooses resolution defaults, compresses large PNGs with pngquant (unless disabled), and requires an API key via --api-key or GEMINI_API_KEY. It performs preflight checks and returns an error for missing keys, invalid input images, or API permission issues.

When to use it

Quick concept iteration at 1K to validate prompts and composition
Image-to-image edits like style changes, object removal, color adjustments
Final high-resolution renders at 4K once the prompt is locked
Batch or repeatable runs where filename timestamps and deterministic naming help tracking
When you need scriptable, local saving of generated PNGs and automated compression

Best practices

Always run the script from your current working directory so outputs save where you expect
Start with 1K drafts, iterate with small prompt diffs, then render final at 4K
Provide an explicit filename using the timestamp-descriptive-name pattern to avoid collisions
Pass --input-image for edits and reuse the same input while iterating
Install pngquant if you want automatic compression; use --no-compress or --max-size to override defaults
Supply an API key via --api-key or set GEMINI_API_KEY in your environment before running

Example use cases

Create concept art: quick 1K drafts, iterate prompts, then export a 4K final
Edit a photo: pass --input-image and instruct only targeted changes (e.g., sky replacement)
Style transfer: convert a photo to watercolor, cartoon, or cinematic color grade
Asset generation: produce UI backgrounds, icons, or scene elements for a game or prototype
Automated workflows: script batch generation with deterministic filenames for asset pipelines

FAQ

Use 1K for fast iteration. Move to 2K for refinement and 4K only for final renders to save time and quota.

How does the script find the API key?

It checks --api-key first, then the GEMINI_API_KEY environment variable. If neither is present it exits with an error.

What if my output PNG is too large?

The script uses pngquant to compress images over the max-size (default 8MB). Install pngquant or use --no-compress or --max-size to customize behavior.

How do I make predictable filenames?

Use the pattern yyyy-mm-dd-hh-mm-ss-descriptive-name.png. Include a concise descriptive name or a short random identifier when context is unclear.