library/developer-tools/agent-capture

What problem does it solve?

It enables automated, native capture of macOS windows, displays, and regions for AI-driven workflows.

Core Features & Use Cases

• Window and display listing: Enumerate all visible windows and connected displays.
• Screen and window capture: Take screenshots of specific apps, windows, displays, or regions.
• Video recording: Record macOS windows or screens with duration control.
• GIF conversion and stitching: Convert recordings to optimized GIFs and stitch frames into animations.
• Code screenshot rendering: Generate styled images of source code with syntax highlighting.
• Diff and monitoring: Perform visual diff against baseline and monitor UI changes over time.
• Fuzzy window search: Find windows by partial or complete titles for easy targeting.
• Full evidence bundle: Collect screenshots, recordings, and GIFs into comprehensive evidence packages.
• Optional OCR: Extract text from windows for automation and analysis.
• Permission diagnostics: Verify Screen Recording permissions and environment readiness.
• Pipeline workflows: Combine recording and conversion into single commands for efficiency.

Quick Start

Ask the AI to list open windows and capture the Finder window screenshot, or record an app for 10 seconds and convert to GIF. For example, say: "Capture a 10-second screen recording of the Preview app and save as demo.mp4."