Auto-analyze video files and URLs mentioned in prompts. Extracts frames + transcribes audio so Claude can see and hear shared videos (screen recordings, clips, etc).
A Claude Code plugin that lets Claude watch your screen recordings.
Drop a .mov, .mp4, or any video URL into a Claude Code prompt and Claude will automatically:
No copy-pasting frame-by-frame, no manual setup per video.
Open Claude Code and run these two commands:
/plugin marketplace add EcomGraduates/video-analyze
/plugin install video-analyze@ecomgraduates
Then restart Claude Code (or run /reload-plugins).
That's the whole install. The very first time you share a video, the plugin handles everything else automatically (one-time dependency install, ~488 MB Whisper model download).
Just mention a video file or URL in a normal prompt:
hey there's a bug in this /Users/me/Desktop/Screen Recording 2026-04-17 at 1.26.27 PM.mov please fix it
or
can you check out https://example.com/demo.mp4
Claude will see the frames, hear the audio, and respond to what you actually asked. No special syntax, no flags, no attachments UI.
ffmpeg + whisper-cppIf you've never installed developer tools before, the plugin will detect this and tell Claude to paste this one line for you. You can also install it yourself first:
Open Terminal (⌘ + Space, type "Terminal", hit enter) and paste:
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
It'll ask for your Mac password once — same one you use to log in. Wait a few minutes for it to finish, then you're ready to install the plugin (two commands above).
/plugin update video-analyze@ecomgraduates
/plugin uninstall video-analyze@ecomgraduates
Clean removal — no leftover files in your settings.
When you submit a prompt containing a video path or URL, a UserPromptSubmit hook fires:
U+202F narrow no-break space before "AM"/"PM").ffmpeg at 1 frame every 2 seconds by default.whisper-cpp if speech is present.Read each frame image and respond naturally.Temp files are stored in /tmp/video-analyze-* and cleaned up automatically when your Claude session ends.
mp4, mov, avi, mkv, webm, m4v, flv, wmv, mpg, mpeg, ts, mts, 3gp
.claude-plugin/
plugin.json # hook registrations
marketplace.json # single-plugin marketplace manifest
hooks/
video-detect.sh # UserPromptSubmit hook (Python)
video-cleanup.sh # SessionEnd cleanup
bin/
video-analyze # ffmpeg + whisper-cpp runner
MIT
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub ecomgraduates/video-analyze --plugin video-analyzeGive Claude the ability to watch and understand videos — extracts frames and audio for full video perception
Turn videos into a sequence of relevant still frames + transcript + a self-contained HTML report so Claude can view them as images, hear the audio, and write its analysis back into the report. Pass a local path, an http(s) URL, or pipe video bytes on stdin.
Let Claude watch a video. Downloads with yt-dlp, extracts auto-scaled frames with ffmpeg, pulls captions or falls back to Whisper, and hands frames + transcript to Claude so it can answer questions about the video.
Compose yt-dlp + ffmpeg + Whisper into a single command that hands an AI agent the raw materials to watch any social video — VIDEO + FRAMES + TRANSCRIPT, ready for an LLM to read frames as images and transcript as text.
Turn any video into a section-by-section study-notes markdown file with embedded screenshots and a timestamped transcript.
Process videos using VideoDB Python SDK. Upload, search, edit, generate subtitles, transcode, capture, and stream — all via natural language.