nanobanana-skill

What This Skill Does

The nanobanana-skill is a powerful image processing interface for OpenClaw that leverages Google's Gemini API to perform high-quality image generation and targeted image editing. By integrating directly into your workflow, it allows users to translate natural language prompts into visual media without leaving the terminal or their chat interface. Whether you are looking to create unique artistic assets, generate mockups, or perform iterative edits on existing images, this skill provides a robust command-line wrapper around sophisticated AI models.

Installation

To integrate nanobanana-skill into your OpenClaw environment, ensure you have the necessary dependencies installed by running the install command. You must have an active Google Gemini API key. Configure this key by creating an environment file at ~/.nanobanana.env or by exporting the variable in your shell session. The skill requires Python 3 and specific libraries including google-genai, Pillow, and python-dotenv. You can quickly install the required environment via the provided requirements.txt file located in the skill's plugin root directory.

Use Cases

This skill is designed for creators, developers, and productivity-focused users who need fast image manipulation. Use it for generating rapid prototypes of UI concepts, creating conceptual art for blog posts, or modifying existing photos to include specific atmospheric elements. It is particularly effective for workflows requiring automated visual asset generation, such as batch-producing icons or applying consistent artistic filters across multiple images.

Example Prompts

1. "nanobanana, generate an image of a cybernetic cat sitting on a neon rooftop, use 16:9 aspect ratio and save it as cyberpunk_cat.png"
2. "I need to edit this file, create an AI drawing version of input.png that adds a stormy background but keeps the foreground character"
3. "nanobanana, please create a high-resolution 2K image of a serene minimalist desert landscape for my desktop wallpaper"

Tips & Limitations

For best results, provide descriptive, high-detail prompts. While the default 9:16 aspect ratio works for most social media content, leverage the --size options to match your target platform requirements exactly. Note that the skill relies on an external API, so internet connectivity is required. Always double-check your API key permissions and usage quotas on the Google AI Studio portal. Keep in mind that editing existing images requires precise prompting to ensure the model maintains the integrity of the original source files.