nano-banana-pro-openrouter Skill (awesome-copilot) for Claude Code: What It Does and How to Install It

Makes Claude generate or edit images with Gemini 3 Pro Image through OpenRouter, using your API key.

Installs: 9.9K; on skills.sh
Stars: 40.7K; on GitHub
Files: 2; besides SKILL.md
Updated: Oct 2026; last repo commit

This skill lets Claude make pictures. It runs a bundled Python script that sends your prompt, and any input images, to OpenRouter and saves the result to a file. It can generate from a prompt alone, edit one image or combine up to three.

It needs uv and an OpenRouter API key in OPENROUTER_API_KEY. It is not by Google or OpenRouter. Requests use your OpenRouter credits.

Good for

Anyone with an OpenRouter key who wants images made from the terminal
Designers editing or combining up to 3 existing images
Developers who need 1K, 2K or 4K output files

Install nano-banana-pro-openrouter

Claude Code

Copy this prompt.

Install the nano-banana-pro-openrouter skill in Claude Code. Run this command:

npx skills add https://github.com/github/awesome-copilot --skill nano-banana-pro-openrouter

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js, uv.

The skill needs uv installed and an OPENROUTER_API_KEY in your environment. Because it runs a local Python script, there is no Claude app upload route.

Built from the package data we hold. Not yet tested by us.

What's in the skill

When it loads

Generate or edit images via OpenRouter with the Gemini 3 Pro Image model. Use for prompt-only image generation, image edits, and multi-image compositing; supports 1K/2K/4K output.

Sections in SKILL.md

Overview
Prompt-only generation
Edit a single image
Compose multiple images
Resolution
System prompt customization
Behavior and constraints
Troubleshooting

Key facts

Author: github/awesome-copilot (repo owner: GitHub; community-contributed)
License: MIT
Installs on skills.sh: 9,935 (Nov 2, 2026)
GitHub stars: 40,666 for the whole github/awesome-copilot repo (Nov 2, 2026)
Last repo commit: Oct 30, 2026
Audits on skills.sh: Gen Agent Trust Hub: pass, Socket: pass, Snyk: pass
What it includes: 3 files including SKILL.md: scripts/generate_image.py and assets/SYSTEM_TEMPLATE
Needs: uv, and an OPENROUTER_API_KEY environment variable

What it includes

uv run scripts/generate_image.py --prompt ... --filename out.png
Calls OpenRouter with the google/gemini-3-pro-image-preview model and saves the image.
Try: “A cinematic sunset over snow-capped mountains, save as sunset.png”
Runs code · Prompt-only generation
Adds --input-image to change an existing picture
The input image is read from disk and sent with the prompt.
Runs code · Edit one image
Up to 3 --input-image files into one result
The script stops with an error if you pass more than 3.
Runs code · Compose images
--resolution 1K, 2K or 4K (default 1K)
Sent as the image size in the request.
Network · Resolution
Reads an optional assets/SYSTEM_TEMPLATE file
The bundled template sets the tone of the image-writing assistant.
Files · System prompt
Saves files, numbers multiple images, prints MEDIA: <path>
Existing files with the same name are overwritten. A file extension that does not match the returned image type is corrected.
Write · Output handling
Fixes for a missing key, missing uv and a 401
Retry once after 30 seconds on HTTP 429 or network timeouts, and not more than twice for the same error.
Read · Troubleshooting table

From the official docs

Permissions

Medium risk

What it can do

Read your data
Create and change things
Delete things
Read-only mode

Watch out for

It uses your OpenRouter key and credits
The script reads OPENROUTER_API_KEY and sends each request to OpenRouter with it, so the charge lands on the OpenRouter account that owns the key. SKILL.md gives no price per image and says nothing about spending limits; check OpenRouter pricing for the model first. github.com
Your prompts and images are uploaded
The script base64-encodes up to 3 input images and sends them with the prompt and system prompt to openrouter.ai. Do not use private images you cannot upload. github.com
It runs a Python script via uv
The script header lists the openai package as a dependency, which uv fetches when it runs. SKILL.md also gives uv install commands (curl | sh on macOS and Linux, a PowerShell command on Windows) in its troubleshooting table. github.com
Output files are overwritten
The script writes straight to the filename you pass without checking for an existing file. Pick a new filename each time. github.com

Security advisories

No published security advisories as of Nov 2, 2026.

Technical details

Turning on read-only	No read-only mode.
Talks to	openrouter.ai (API calls with your key). Python package index (the openai dependency, fetched by uv). astral.sh (only if you follow the uv install steps)
Sign-in and keys	The key is read from the OPENROUTER_API_KEY environment variable. SKILL.md does not describe where else it might be stored.
Scopes	Reads input images you name Writes image files to the filename you give, overwriting an existing file Uses your OpenRouter API key from the environment

Questions

What does the nano-banana-pro-openrouter skill do?

It makes Claude generate or edit images by running a bundled Python script. The script calls the Gemini 3 Pro Image model through OpenRouter and saves the results to your disk.

Do I need an API key?

Yes. It needs an OpenRouter key in the OPENROUTER_API_KEY environment variable. SKILL.md also says the key must be valid and have credits, and links to OpenRouter's key settings page.

Does it cost money?

Yes, if your OpenRouter account is billed for the model. Every image request is sent under your key, so OpenRouter charges the account that owns it, not the skill author. SKILL.md gives no price per image and says only that a key with no credits fails with HTTP 401. Check OpenRouter pricing for google/gemini-3-pro-image-preview before heavy use, including 2K and 4K output.

How do I install it?

Use the install box at the top of this page. It also needs uv installed, which SKILL.md lists as a required program. We do not offer a Claude app upload route because the skill runs a local script.

Is this an official Google or OpenRouter skill?

No. It is not by Google or OpenRouter. It is community-contributed in GitHub's awesome-copilot repo.

Where do my prompt and input images go?

The script sends your prompt, any input images (encoded into the request) and the system prompt to openrouter.ai using your key. Output is written to the filename you give.

Is it safe?

It runs a bundled Python script through uv, which downloads the openai Python package, and sends data to OpenRouter with your key. Only pass images you are happy to upload, and keep the key out of chat messages and files.

Information

updated Nov 2, 2026

Repository: github/awesome-copilot
Folder: skills/nano-banana-pro-openrouter

github/awesome-copilot skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 4

nano-banana-pro-openrouter SKILL.md · github.com
awesome-copilot README · github.com
generate_image.py · github.com
SYSTEM_TEMPLATE · github.com