A lightweight macOS utility for capturing instant display snapshots and monitoring visual state via command line.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install screengrab
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 screengrab 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 Screengrab?
Screengrab is a specialized utility designed for AI agents operating on macOS. It acts as a thin, high-performance wrapper around the native screencapture system tool, allowing agents to gain immediate visual awareness of the user's environment without the overhead of a browser. By providing a clean interface for Openclaw Skills, it enables programmatic access to full-screen captures, specific display targeting, and automated watch modes.
This skill is essential for developers building agents that need to verify UI state, debug visual outputs, or monitor activity across multiple monitors. Unlike web-based screenshot tools, Screengrab captures the actual desktop environment, including native applications, terminal windows, and system menus, making it a cornerstone for comprehensive desktop automation.
Screengrab Use Cases
- Verifying the visual state of a non-browser application during an automated workflow.
- Debugging layout issues or UI glitches on a local macOS environment.
- Monitoring background processes or long-running tasks via periodic screen captures.
- Identifying active windows and connected display configurations for multi-monitor setups.
- Capturing rapid snapshots to be processed by vision-language models (VLMs) for real-time analysis.
How Screengrab Works
- The AI agent or user invokes the Python script with specific flags like snap, list, or watch.
- The script interfaces with the macOS /usr/sbin/screencapture binary to interface with the Window Server.
- Image data is captured from the specified display index or the entire desktop area.
- The resulting PNG file is saved to a designated path (defaulting to /tmp) with a unique timestamp.
- The script returns the file path or status, allowing subsequent Openclaw Skills to ingest the image for analysis or storage.
Screengrab Setup
To integrate Screengrab into your environment, ensure you are running macOS with Python 3 installed. No external package dependencies are required.
# Ensure the script is executable
chmod +x scripts/screengrab.py
# Test a basic capture
python3 scripts/screengrab.py snap
Screengrab Data Schema & Taxonomy
Screengrab organizes its output using a straightforward file-based structure to ensure easy hand-offs between agent tools.
| Attribute | Description |
|---|---|
| Default Path | /tmp/screengrab-[TIMESTAMP].png |
| Output Formats | Standard PNG |
| Metadata | Timestamped filenames for sequential tracking |
| Display IDs | Integer-based indexing (e.g., 0 for main display) |
Screengrab Advanced Features
- Watch Mode: Automate captures at specific intervals (e.g., every 10 seconds) for a set duration or count.
- Quiet Output: Use the --quiet flag to return only the file path, optimized for piping into secondary image analysis tools.
- Multi-Display Support: Programmatically list connected hardware and target specific screens for focused snapshots.
- Direct Preview: Automatically open the captured image in the macOS Preview app for immediate human verification.
- Custom Directory Routing: Redirect bulk captures to specific project folders to maintain organized data logs for Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
LinkedIn Automator
An AI-powered automation suite for managing LinkedIn content, engagement, and growth analytics via browser interaction.
SignalWire
A powerful integration that enables AI agents to manage voice calls, messaging, and video communication via SignalWire APIs.
Sendbird
A comprehensive integration for managing Sendbird messaging platforms, including user and channel administration via an AI-ready interface.
Confluent
A comprehensive integration allowing AI agents to manage Kafka clusters, topics, and records through a secure, automated interface.
Watchdog
Watchdog provides automated monitoring and self-healing for all OpenClaw cron jobs, ensuring continuous uptime for your AI agent workflows.
Agent Config Validator
A specialized diagnostic tool for validating consistency between agent configurations and core documentation within the OpenClaw ecosystem.
Featured*
