2.5k
GitHub Stars
2
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 openclaw/skills --skill nano-bananapro- _meta.json280 B
- SKILL.md6.3 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 and produces 1K, 2K, or 4K PNG outputs saved to your current working directory. Use it for rapid drafts, iterative edits, and final high-resolution renders.
How this skill works
The script calls the Gemini 3 Pro Image API to create or modify images from a text prompt and optionally an input image. It chooses resolution based on user input or simple resolution-mapping rules, requires an API key (argument or GEMINI_API_KEY), and writes a timestamped PNG filename to the current directory. For edits, pass the input image path with --input-image and include clear editing instructions in the prompt.
When to use it
- Quickly prototype visuals with a 1K draft before spending time on high-res renders.
- Iteratively refine an image by re-running with small prompt changes and the same input-image.
- Produce final high-resolution artwork at 4K once the prompt and composition are locked.
- Perform targeted edits: remove/add elements, change style, adjust colors, or blur backgrounds.
- When you can provide an API key or have GEMINI_API_KEY set in your environment.
Best practices
- Start at 1K for fast feedback; only switch to 4K for final outputs to save API time and quota.
- Use concise timestamped filenames: yyyy-mm-dd-hh-mm-ss-descriptive-name.png to avoid overwrites.
- Provide a focused prompt or use the provided templates when user intent is vague.
- For edits, keep the same input image across iterations and specify exactly what to change.
- Ensure an API key is available via --api-key or GEMINI_API_KEY before running the script.
Example use cases
- Generate a concept art draft: run at 1K, refine prompt, then export final at 4K.
- Edit a client photo to make the sky more dramatic using --input-image and a concise edit prompt.
- Convert a photo to a painterly or cartoon style while preserving composition and lighting.
- Produce marketing assets in multiple resolutions by mapping 'high-res' to 4K, 'medium' to 2K, etc.
- Batch-create variations of a product shot by changing only small prompt details and filenames.
FAQ
The script accepts --api-key on the command line or reads GEMINI_API_KEY from the environment.
How should I name output files?
Use the pattern yyyy-mm-dd-hh-mm-ss-descriptive-name.png. Keep the descriptive part short and hyphenated.
How do I perform an edit instead of a fresh generate?
Include --input-image with the image path and put the edit instructions in --prompt. Keep iterations using the same input image.