A high-performance Windows screen capture utility designed for AI vision integration and desktop automation.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install screenshot-skill
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 screenshot-skill 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 Screenshot Capture?
The Screenshot Capture skill provides a robust interface for capturing screen data on Windows environments. By leveraging the mss and Pillow libraries, it offers high-speed image acquisition that is essential for real-time desktop monitoring and AI-driven visual analysis. This tool is a core component of the Openclaw Skills ecosystem, enabling agents to 'see' the user's desktop, identify UI elements, and process visual information through advanced vision models.
Whether you need to capture a specific region, a single monitor, or a sprawling multi-monitor setup, this skill provides the necessary abstractions. It supports various output formats including PIL Image objects, local file storage, and base64 strings, making it incredibly easy to pipe visual data directly into large language models for sophisticated reasoning tasks.
Screenshot Capture Use Cases
- Providing visual context to AI agents for desktop troubleshooting.
- Capturing specific UI regions to monitor application state changes.
- Generating automated visual logs for audit trails in desktop workflows.
- Streaming screen content to vision-based LLMs like GPT-4o for real-time analysis.
- Managing screen assets in multi-monitor developer environments.
How Screenshot Capture Works
- The skill initializes a connection to the system display drivers using the mss library for low-latency capture.
- It queries the system to map all available monitors and their respective resolutions.
- Upon a capture request, it retrieves the raw pixel data from the specified screen region or monitor.
- The raw data is processed through Pillow to convert it into a standard image format (PNG, JPEG, or PIL object).
- The resulting image is either saved to the local file system or encoded as a base64 string for immediate API transmission.
Screenshot Capture Setup
To integrate this capability into your project using Openclaw Skills, install the necessary dependencies first:
uv add mss pillow
# Or using pip
pip install mss pillow
You can then use it as a CLI tool to verify your display configuration:
python scripts/screenshot.py info
python scripts/screenshot.py capture -o test_shot.png
Screenshot Capture Data Schema & Taxonomy
The skill organizes screen data and monitor metadata through a structured API. Below are the primary data structures used within these Openclaw Skills:
| Method | Description | Return Type |
|---|---|---|
screen_size |
Returns primary monitor dimensions | tuple[int, int] |
all_screen_size |
Returns combined dimensions of all monitors | tuple[int, int] |
monitors |
Detailed metadata for each connected display | list[dict] |
capture |
Captures a region as an image object | PIL.Image |
capture_to_base64 |
Encodes screen data for vision API usage | string |
Monitor metadata includes keys for left, top, width, and height to facilitate precise coordinate mapping.
Screenshot Capture Advanced Features
- Direct base64 encoding with adjustable quality settings (JPEG/PNG) to optimize bandwidth for AI vision requests.
- Multi-monitor awareness allowing for specific monitor indexing in complex workstation setups.
- Context manager support (
with ScreenCapture()) to ensure efficient resource allocation and cleanup. - Automated sequential saving with the
stepparameter, ideal for logging step-by-step agent actions. - Region-specific cropping at the capture level to reduce processing overhead.
SKILL.md
Loading
Related Openclaw Skills
Trading Signals WebSocket
A real-time cryptocurrency signal generator that connects exchange WebSocket price feeds to technical strategies and Telegram alerts.
Crypto Backtest Engine
A fast, scriptable crypto futures backtesting engine supporting multiple exchanges and technical indicators.
Bybit Futures Trading System
A professional-grade trading infrastructure for Bybit USDT perpetual futures featuring built-in risk management, paper trading, and real-time WebSocket execution.
Life Book
A digital biographer skill that transforms natural conversations into a structured, chapter-based personal history and life book.
AI Content Generator
A streamlined automation tool for generating high-quality marketing copy, blog posts, and social media content using advanced AI models.
API Docs Generator
Automatically transform source code comments into professional, interactive API documentation like OpenAPI and Swagger.
Featured*
