vibe-harvester
自动向下滚动瀑布流网站(如小红书、Pinterest),根据用户设定的审美偏好进行视觉筛选,并自动下载符合要求的参考图。
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/anotherj1/vibe-harvesterWhat This Skill Does
The Vibe Harvester is a specialized OpenClaw AI agent skill designed to act as your autonomous art assistant on image-heavy waterfall websites like Pinterest, Xiaohongshu (Red), and other visual platforms. Instead of manually scrolling through thousands of images to curate your reference library, this skill employs advanced visual recognition to filter content based on your specific aesthetic preferences. It automates the browsing loop, intelligently identifies images that align with your prompt, and performs safe, structured downloads to your local machine. It is designed to be human-like, utilizing randomized intervals and mouse behavior to mimic organic browsing while efficiently building your mood board.
Installation
To install the Vibe Harvester skill, open your OpenClaw terminal and execute the following command:
clawhub install openclaw/skills/skills/anotherj1/vibe-harvester
Ensure you have sufficient permissions for local file writes in your designated download directory.
Use Cases
- Graphic Design Research: Quickly aggregate high-quality UI/UX design references or typography inspiration from platforms like Dribbble or Pinterest.
- Interior Design Curation: Automatically save home decor styles, color palettes, and furniture arrangements that match a specific "minimalist" or "bohemian" aesthetic.
- Mood Board Creation: Populate local project folders with consistent style assets for creative sessions, social media content planning, or professional look-books.
- Concept Art Gathering: Systematically collect character designs or environmental textures for game development or digital illustration workflows.
Example Prompts
- "Go to the current tab and start scanning for images that match a 'dark academia, vintage library' aesthetic. Save any high-quality photos to my local folder."
- "I need inspiration for my next UI project. Keep scrolling through this feed and download every image that features 'glassmorphism' and 'soft pastel gradients' until we have 20 samples."
- "Start harvesting images from this feed. Focus only on 'brutalist architecture' photography. Skip any thumbnails that look blurry or contain text overlays."
Tips & Limitations
- Visual Accuracy: The skill relies on its vision model to judge aesthetics. For best results, use descriptive, specific keywords in your prompt regarding color, lighting, and subject matter.
- Anti-Bot Measures: The agent includes randomized scrolling and pause intervals to reduce the risk of being flagged as a bot by complex waterfall sites.
- File Management: All files are saved into
~/Downloads/Vibe_Harvester/. Ensure your local disk has enough capacity before starting a long-running collection session. - Avoidance: The skill is hard-coded to ignore any images labeled as 'Ad' or 'Promoted'. Use this to your advantage to maintain a clean, high-quality reference collection without marketing clutter.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-anotherj1-vibe-harvester": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: file-write, data-collection