Video Vision
Local ffmpeg+Whisper pipeline that lets Claude actually watch and transcribe video files
Test report
- Verdict
- Tested · Works
- Score
- Tested
- Jul 16, 2026
- Environment
- Claude Code 2.x (agent harness)
- Upstream re-checked
- Jul 18, 2026 · 5234dbf
Ran the real pipeline end to end against a synthetic 6-second test clip: ffmpeg frame extraction and local Whisper transcription both executed for real and returned the exact JSON manifest shape the README documents, down to field names. No network calls beyond the disclosed one-time Whisper model download, no exfiltration, nothing hidden - genuinely gives Claude a video-watching capability it has no other way to get.
Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works
- Installs cleanly 5/5
- Triggers reliably 5/5
- Output vs. baseline 10/10
- Docs & honesty 5/5
What Video Vision does
Extracts sampled frames via ffmpeg and a timestamped transcript via local Whisper, then answers questions grounded in both visual and audio evidence with [MM:SS] citations. Triggers whenever a user references a local video file (.mp4/.mov/.mkv/.webm/.avi/.m4v) wanting it watched, summarized, searched, captioned, or transcribed; fully offline after a one-time Whisper model download.
How to install Video Vision
git clone https://github.com/Moh4696/claude-video-vision ~/.claude/skills/video-vision
Skills live in ~/.claude/skills/ (global) or .claude/skills/
(per-project). Restart Claude Code after installing.
Commands — how to trigger Video Vision
-
/video-visionLocal ffmpeg+Whisper pipeline that lets Claude actually watch and transcribe video files
It also activates on plain-language prompts like these:
-
Can you watch this recorded meeting.mp4 and summarize what happened? -
I need a searchable transcript with timestamps for this lecture video. -
Caption this .mov clip and tell me what's said around the 2 minute mark.
Frequently asked questions
- Is the Video Vision skill free?
- Yes. The skill itself is free from Moh4696/claude-video-vision. SkillProof publishes the install command and an independent test verdict at no cost.
- Does Video Vision work with Claude Code?
- We tested it with Claude Code 2.x (agent harness) on Jul 16, 2026. Verdict: Tested · Works. Ran the real pipeline end to end against a synthetic 6-second test clip: ffmpeg frame extraction and local Whisper transcription both executed for real and returned the exact JSON manifest shape the README documents, down to field names. No network calls beyond the disclosed one-time Whisper model download, no exfiltration, nothing hidden - genuinely gives Claude a video-watching capability it has no other way to get.
- What is the Video Vision SkillProof Score?
- 10.0/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 10/10, docs & honesty 5/5.
- How do I install Video Vision?
- Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
- Can I use Video Vision with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
- The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.