Video Frames — OpenClaw Skill

What This Skill Does

The Video Frames skill gives your OpenClaw agent the ability to extract individual frames from video files using ffmpeg. It can capture the first frame, a frame at any specific timestamp, or create thumbnails for quick visual inspection. This is useful for video analysis, content review, thumbnail generation, and any workflow where you need a still image from a video.

The skill wraps a dedicated extraction script that handles the ffmpeg complexity for you. Point it at any video file, optionally specify a timestamp with --time, and get a clean JPG or PNG frame in return. Use JPG for quick sharing (smaller file size) and PNG for crisp UI screenshots where pixel-perfect quality matters.

This pairs well with other media skills for workflows like extracting a keyframe from a video, analyzing its content visually, then posting it to social media or attaching it to a Slack message.

Example Prompts

Extract the first frame of this video and show it to me

Grab a screenshot from the video at the 10-second mark

Create a thumbnail from the video at timestamp 00:01:30 and save it as a PNG

What's happening at the 45-second mark of this recording? Extract the frame so I can see

Pull frames at 0s, 30s, 60s, and 90s from this video to create a visual summary

Extract a frame from the intro of this screen recording to use as a blog post thumbnail

Requirements

Binary dependency: ffmpeg must be installed and available in PATH.

• Install via Homebrew: brew install ffmpeg
• Linux: apt install ffmpeg or equivalent for your distribution

Setup on KiwiClaw

FFmpeg is pre-installed on all KiwiClaw agent machines. Your agent can extract video frames immediately with no setup. Manage your agent's capabilities from the KiwiClaw dashboard.

Setup Self-Hosted

1. Install ffmpeg: brew install ffmpeg (macOS) or apt install ffmpeg (Linux)
2. Verify: ffmpeg -version
3. The skill activates automatically when your agent needs to extract frames from video

Related Skills

• Songsee -- visualize the audio track from your video files
• Nano Banana Pro -- edit extracted frames with AI image generation
• Sherpa-ONNX TTS -- add narration to accompany extracted video frames
• Xurl -- post extracted frames to X/Twitter

FAQ

What video formats does the Video Frames skill support?

The skill uses ffmpeg under the hood, so it supports virtually all common video formats including MP4, MOV, AVI, MKV, WebM, and many more. Any format ffmpeg can decode will work.

Can I extract a frame at a specific timestamp?

Yes. Use the --time flag with a timestamp in HH:MM:SS format (e.g., --time 00:00:10) to extract a frame at exactly that point in the video.

What output formats are supported?

Frames can be saved as JPG for quick sharing or PNG for pixel-perfect UI screenshots. Use .jpg for compressed, smaller files and .png when you need crisp, lossless quality.

Does this skill require ffmpeg to be installed separately?

Yes. The Video Frames skill requires ffmpeg as a binary dependency. On macOS, install it with brew install ffmpeg. On Linux, use your distribution's package manager. On KiwiClaw, ffmpeg is pre-installed.