Name: Video Frames
Author: steipete

What breaks without openclaw video frame extraction skill

Video content invisible to LLMs. Manual frame extraction. ffmpeg commands outside agent flow.

→

Automatic video frame extraction for LLM vision × 130-star skill ÷ 10–15 minute setup ÷ ffmpeg required = video content analysed at scale.

openclaw video frame extraction skill — what it actually does

Extract key frames from video files inside OpenClaw.

Feed frames to vision-capable LLMs for content analysis.

Generate thumbnails from video files automatically.

Control extraction interval — start with 30-second spacing.

Requires ffmpeg in PATH.

Security check — openclaw video frame extraction skill

Privacy score: 7/10 — accesses connected platform APIs only. Lock it: review OAuth scopes before install, confirm Linux, macOS; OpenClaw ≥1.0; ffmpeg compatibility.

Quick start — openclaw video frame extraction skill in 10–15 minutes

Setup time: 10–15 minutes

You need:

OpenClaw core
ffmpeg installed

Install the package:

clawhub install steipete/video-frames
# Requires: sudo apt install ffmpeg

Install ffmpeg

Install skill

Run /vframes extract video.mp4 --interval 10 to get a frame every 10 seconds

Frames saved to /frames/ directory

Troubleshooting openclaw video frame extraction skill

1. High-fps interval creates too many images — start with 30s intervals

2. ffmpeg path must be in PATH

Compatibility & status

Works with: Linux, macOS; OpenClaw ≥1.0; ffmpeg intermediate Last updated: Oct 2025 ★ 130 on GitHub MIT

Official docs →

View on GitHub →

Related — more like openclaw video frame extraction skill

More by steipete

Video content your agent can't see is data you're leaving unanalysed. Install the openclaw video frame extraction skill now.

Get it on GitHub →