What are the system requirements?

You must have ffmpeg and ffprobe installed on your machine (available via Homebrew for macOS or apt for Linux).

Which video formats are supported?

The skill supports any format that ffmpeg can process, including .mp4, .mov, .webm, .avi, and more.

Can Claude directly watch video files?

No, Claude cannot view video files directly. This skill uses ffmpeg to extract specific frames as images, which Claude can then see and analyze.

Does it work with filenames containing spaces?

Yes, the skill is specifically designed with a copy-and-probe pattern that uses absolute paths and glob patterns to avoid common shell errors related to spaces and special characters.

How does this skill handle long videos?

For longer videos, the skill can be configured to extract frames at a lower FPS (e.g., 1 or 2 frames per second) or only extract keyframes at scene changes to prevent overwhelming the model.

Video Frame Extraction

Name: Video Frame Extraction
Author: ramonclaudio

byramonclaudio

Productivity & Workflow

Enables Claude to analyze video content by automatically extracting and sampling frames as viewable images.

About

The Frames skill bridges the gap between Claude's image-processing capabilities and video file formats by leveraging ffmpeg to convert video streams into sequences of PNG frames. It implements a robust workflow that overcomes common terminal limitations, such as spaces in filenames and quoting issues, allowing Claude to 'watch' videos by sampling evenly spaced frames or specific time ranges. This is essential for tasks requiring visual analysis of screen recordings, UI/UX audits, or video-based data processing within the Claude Code environment.

Key Features

Smart sampling logic that adjusts based on video length
Robust file path handling for filenames with spaces and special characters
Customizable FPS settings and specific time-range extraction
0 GitHub stars
Automatic frame extraction from MP4, MOV, WEBM, and AVI files
Frame resizing to optimize processing speed and model context usage

Use Cases

Extracting specific keyframes from long-form videos for quick visual summaries
Describing visual content and events within recorded video sessions
Analyzing UI/UX screen recordings to identify bugs or design inconsistencies

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ramonclaudio/skills frames

For use in Claude.ai and ChatGPT

Download Skill

GitHub