A professional AI image generation and smart grid splitting tool powered by AceData Nano Banana models for high-quality visual content creation.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install banana-v2
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install banana-v2 using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Nano-Banana-V2?
Nano-Banana-V2 is a robust Openclaw Skills extension designed for high-fidelity AI image generation and sophisticated asset management. Built on the AceData Nano Banana series, it offers a seamless web interface for creating, editing, and organizing images with support for up to 4K resolution. This tool excels at bridging the gap between raw AI generation and practical design workflows by integrating smart splitting capabilities and a responsive waterfall gallery management system.
Nano-Banana-V2 Use Cases
- Generating high-resolution marketing assets in various aspect ratios including 16:9, 9:16, and 4:3.
- Creating social media content using automatic 2/4/6/9 grid splitting for platforms like Instagram or WeChat.
- Iterating on design concepts using multi-reference image-to-image workflows with up to 6 images.
- Crafting detailed artistic prompts from simple descriptions using the integrated LLM-powered prompt generator.
- Managing large collections of AI-generated art via a centralized waterfall gallery with batch download features.
How Nano-Banana-V2 Works
- The user inputs text prompts or uploads reference images through the local web interface provided by the skill.
- The Node.js backend processes the request and interfaces with the AceData Nano Banana API to trigger the generation task.
- If enabled, an integrated LLM enhances the user's prompt to ensure higher quality output.
- Upon completion, the image is retrieved and processed by the Sharp library for optional smart grid splitting based on the chosen layout (2/4/6/9).
- All generated assets and metadata are stored in a local SQLite database and served through a waterfall gallery layout for immediate review and management.
Nano-Banana-V2 Setup
To install this skill within your environment, navigate to the specific skill directory and install dependencies:
cd ~/.openclaw/workspace/skills/nano-banana-v2
npm install
To launch the service:
npm start
Access the interface at http://localhost:2688. You must obtain an API Key from the AceData platform and configure it in the settings page within the Openclaw Skills workspace for full functionality.
Nano-Banana-V2 Data Schema & Taxonomy
Nano-Banana-V2 organizes its data through a structured local filesystem and a SQLite database to ensure persistence and easy retrieval.
| Component | Description |
|---|---|
| Database | SQLite stores task history, image metadata, and configuration settings. |
| Image Storage | High-resolution exports and thumbnails are saved to a configurable path (default: Desktop/banana2). |
| Configuration | The config/set.json file manages API keys, platform tokens, and UI preferences. |
| Taxonomy | Data is categorized by model (Pro/V2), resolution (1K/2K/4K), and aspect ratio. |
Nano-Banana-V2 Advanced Features
- Smart adaptation technology that automatically detects aspect ratios to choose the optimal splitting method.
- Multi-reference image-to-image support for complex visual synthesis using up to 6 source images.
- Integrated LLM for intelligent prompt generation, reducing the barrier for non-technical users.
- Local batch download system that packages multiple generated works into a single ZIP file.
- Comprehensive multi-language support covering English, Chinese, Japanese, and Korean.
- Dual-theme UI with frosted glass effects and responsive waterfall layouts for an optimized user experience.
SKILL.md
Loading
Related Openclaw Skills
MOVA Spec Authoring
An essential automation tool that translates intent calibration data into complete, executable MOVA contracts including envelopes and instruction profiles.
SFE Data Viewer
A specialized pharmaceutical sales performance (SFE) data query and analysis tool optimized for AI agents.
Stock Watcher
A professional-grade financial monitoring skill for tracking Chinese A-share performance and managing personal watchlists through an AI agent.
Push2Today
Push2Today is a specialized utility that sends AI task execution results and summaries directly to a mobile negative-one screen for instant visibility.
Tavily Search
A high-performance web search and content extraction skill optimized for AI agents to retrieve clean, high-signal data.
Summarize
A lightning-fast command-line interface for generating AI-powered summaries of web content, local documents, and multimedia links.
Featured*
