muapi-nano-banana

What This Skill Does

The muapi-nano-banana skill is a sophisticated image-generation tool designed to bridge the gap between abstract user intent and high-fidelity visual output. Unlike standard prompt-to-image tools that rely on 'keyword soup'—stringing together disconnected adjectives—this skill utilizes a reasoning-driven architecture inspired by Gemini 3. It forces the OpenClaw agent to structure requests using a logical framework: Subject, Action, Context, Composition, and Lighting. By interpreting intent through this lens, the skill ensures that physics, spatial relationships, and atmospheric qualities are coherent, resulting in professional-grade imagery that adheres to complex prompt instructions.

Installation

To integrate this skill into your OpenClaw environment, execute the following command in your terminal:

clawhub install openclaw/skills/skills/anil-matcha/muapi-nano-banana-skill

Ensure your muapi.ai API key is properly configured within your environment variables to allow the agent to authorize generation requests via the generate-nano-art.sh script.

Use Cases

1. Cinematic Storyboarding: Ideal for creative directors who need consistent scene generation with specific camera lens definitions (e.g., 35mm, 85mm, anamorphic) and precise lighting setups.
2. Product Design & Mockups: Perfect for visualizing concepts where text integration is required, such as signage on a storefront or labels on a bottle, using the skill's specific text-rendering precision protocol.
3. Scientific & Abstract Visualization: Because this skill handles spatial relationships and physics well, it is highly effective at generating images of complex interactions, such as caustic light patterns, structural collapses, or fluid dynamics.

Example Prompts

1. "I need an image of an elderly clockmaker in a dusty workshop; capture him using a jeweler's loupe, 50mm lens, warm volumetric lighting, soft focus on the background."
2. "Generate a futuristic digital poster for a synthwave concert featuring the text 'NEON DREAMS' in a bold, retro-futuristic font, set against a dark rainy street at night."
3. "Show me a high-speed macro shot of a single water droplet hitting a calm pond surface, creating perfect concentric ripples with a shallow depth of field."

Tips & Limitations

To get the best results, always prioritize descriptive, active sentences over long lists of stylistic keywords. If the output lacks structural integrity, adjust the composition parameter to specify a tighter or wider focal length. Note that while this skill excels at text integration, extremely long sentences or complex paragraphs may still pose challenges for current diffusion models; stick to short, punchy text labels whenever possible. Always verify your API quota before initiating large batch generations, as this skill performs external network requests to process your reasoning briefs.