A professional multimodal vision skill designed for deep analysis of images, videos, and visual files using high-performance LLMs.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install video-image-file-analysis
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 video-image-file-analysis 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 Video and Image File Analysis?
The Video and Image File Analysis skill is a robust solution for developers looking to add sight to their AI agents. Unlike traditional image processing libraries, this skill leverages state-of-the-art multimodal models such as Zhipu GLM-4V and Qwen-VL to perform complex visual reasoning. By integrating this into Openclaw Skills, agents can interpret local files, remote URLs, and even video content with high accuracy.
This skill is particularly effective for tasks requiring scene understanding, document parsing, and object localization. It bridges the gap between raw visual data and actionable intelligence by providing a unified interface for multiple leading vision-language models, ensuring that your agent always has access to the best available visual reasoning capabilities.
Video and Image File Analysis Use Cases
- Automated OCR and document text extraction for digitizing paperwork.
- Real-time scene understanding and object localization for robotics or monitoring.
- Summarizing video content to identify key events without manual scrubbing.
- Visual comparison of multiple images to detect changes or verify consistency.
- Detailed image descriptions for accessibility or automated metadata generation.
How Video and Image File Analysis Works
- The agent receives a request containing an image (local/URL) or video file.
- The skill automatically handles file conversion to Base64 or prepares the remote link for processing.
- The specified vision model is initialized using credentials from the local config.json.
- Visual data is sent to the multimodal API alongside a specific prompt for analysis.
- The system processes the response, which can include raw text descriptions or structured JSON data.
Video and Image File Analysis Setup
First, configure your API keys in skills/image-analysis/scripts/config.json:
{
"default_model": "zhipu",
"zhipu": {
"api_key": "your-api-key",
"model": "glm-4.6v-flash"
},
"qwen": {
"api_key": "your-api-key",
"model": "qwen3-vl-plus"
}
}
Run a test analysis via the CLI:
python3 skills/image-analysis/scripts/vision.py analyze --image path/to/image.jpg --prompt "Describe this image"
Video and Image File Analysis Data Schema & Taxonomy
The skill manages visual data and configurations as follows:
| Data Type | Description | Supported Formats |
|---|---|---|
| Images | Local or remote visual files | jpg, png, gif, webp, bmp |
| Videos | Multimedia files for summary | mp4 |
| Config | API credentials and model defaults | JSON |
| Meta | Thinking mode and output flags | CLI Arguments |
Video and Image File Analysis Advanced Features
- Thinking Mode: Use the --thinking flag with Zhipu models to trigger deeper reasoning for complex object positioning tasks.
- Multi-Image Support: Compare multiple images in a single request to identify differences.
- Structured Output: The --json flag enables seamless integration with other Openclaw Skills by returning machine-readable data.
- Video Processing: Summarize and analyze mp4 video files to extract temporal insights.
- Automatic Optimization: Automatically handles image resizing and pixel constraints for specific model requirements (e.g., 6000x6000px limits).
SKILL.md
Loading
Related Openclaw Skills
Smart Map Guide (Amap Integration)
A comprehensive Amap (高德地图) integration for AI agents to perform intelligent route planning and detailed Point of Interest (POI) searches.
Global Weather
A zero-config weather query tool that provides real-time global weather data and multi-day forecasts in AI-optimized formats.
Global News & AI Digest
A zero-config news aggregator that delivers global headlines and deep AI research updates directly to your agent or terminal.
Email Mail Master
A comprehensive email automation tool allowing AI agents to send, receive, and manage messages across major providers like QQ, 163, and Alibaba Mail.
Bing Wallpaper Skill
A lightweight tool for AI agents to retrieve the official daily Bing homepage background through multiple output formats.
Daily AI News
A powerful tool for fetching curated daily briefings on artificial intelligence and large language models.
Featured*
