Nano Banana Openrouter

What This Skill Does

The Nano Banana OpenRouter skill acts as a bridge between the OpenClaw agent ecosystem and Google's advanced image synthesis capabilities. By leveraging the OpenRouter API, this skill allows users to trigger high-quality image generation directly through natural language commands within their chat interface. The skill abstracts the complexities of API key management, HTTP header handling, and model selection, providing a seamless experience for developers and creative users alike to generate visuals using the Gemini 2.5 Flash and related experimental models.

Installation

To integrate this skill into your environment, use the OpenClaw command-line interface. Run the following command in your terminal:

clawhub install openclaw/skills/skills/fyi0329-pixel/nano-banana-openrouter

Once installed, navigate to your openclaw.json configuration file. You must add an entry under skills.entries with the key nano-banana-openrouter. Ensure the apiKey field is populated with a valid OpenRouter token. Alternatively, you can define the OPENROUTER_API_KEY in your system environment variables. Once saved, ensure the enabled flag is set to true to allow OpenClaw to utilize the nano_banana_generate tool.

Use Cases

This skill is ideal for rapid prototyping, content creation, and creative exploration. Use it to generate conceptual mood boards for design projects, create unique social media assets, or illustrate complex ideas during brainstorming sessions. It is particularly effective for users who require frequent iterations on imagery without switching contexts to standalone image generation platforms. Whether you need a simple graphic for a presentation or a complex digital illustration, this skill brings the power of Google's latest image models directly to your fingertips.

Example Prompts

1. "Generate a high-fidelity image of a futuristic laboratory with neon lighting using the Nano Banana tool."
2. "Create a minimalist vector logo for a sustainable coffee shop, ensuring the style is clean and modern."
3. "Draw a landscape painting of a mountain range at sunset, formatted in a 16:9 aspect ratio."

Tips & Limitations

To maximize results, be as descriptive as possible regarding subject, lighting, and style. Note that while the free tier (google/gemini-2.0-flash-exp) is robust for general tasks, the google/gemini-2.5-flash-image-preview model is recommended for high-quality, production-grade output, which will consume OpenRouter credits. Ensure your network has proper access to api.openrouter.ai, as the skill performs remote API calls to process your requests. The skill automatically injects required headers like HTTP-Referer to ensure compliance with OpenRouter's platform policies.