Verified project record

bradautomates/claude-video

A skill that lets AI assistants and coding agents process video by downloading sources, extracting frames, and transcribing audio. It hands the visual and textual data to the model for grounded answers, running locally while orchestrating optional remote transcription APIs.

Project overview

It provides a self-contained skill unit that works across 50+ agent hosts without modification to bridge the gap of AI assistants natively unable to watch or understand video content.

Project type: AI Agent · Video
Use cases: Data Analysis
Deployment: Refer to project documentation
License: MIT

Best for

Developers using AI coding assistants who need to query video content directly within their local environment or agent workflow.
Creators or general users seeking to extract specific information or visual data from platforms like YouTube, Loom, TikTok, X, or Instagram via an AI assistant.

Key capabilities

Uses a yt-dlp wrapper to check native captions first and download only what is needed for the run, supporting YouTube, Loom, TikTok, X, Instagram, and local paths.
Extracts image frames from videos using ffmpeg with auto-fps logic and a deduplication pass to drop near-identical frames.
Parses VTT native captions first, falling back to Whisper orchestration via Groq or OpenAI APIs.
Checks for dependencies (ffmpeg, yt-dlp) on the first run and automatically installs them or provides exact install commands.

Limitations and risks

Long-video accuracy depends on the detail mode selected; capped modes thin out past approximately 10 minutes.
If native captions are not available, transcription relies on external services via the Groq API or OpenAI API.
Telemetry practices are currently not documented.

Getting started

Install the skill via a plugin marketplace or npx skills add, then use the /watch command with a URL and a question in an agent host like Claude Code. The skill requires zero config to start and auto-installs dependencies on the first run.

Alternatives and comparisons

An open source Swift-native macOS video editor that exposes a local MCP server for AI agents to interact with the timeline for editing and generating content.
A pipeline-driven video production system that automates a structured workflow from research to finished video via AI coding assistants.
An open-source tool providing a full pipeline from text to character and storyboarding to video using a three-layer agent collaboration system.

Project comparisons

Evidence and sources

GitHub project description: Give Claude the ability to watch any video. /watch downloads, extracts frames, transcribes, hands it all to Claude.
README: The CLI discovers the skill from `skills/watch/SKILL.md` and copies the whole folder — `SKILL.md` plus its `scripts/` runtime — as a self-contained unit.
README: Claude can read a webpage, run a script, browse a repo. What it can't do, out of the box, is *watch a video*.
README: you can paste a URL or a local path, ask a question, and Claude fetches captions first, downloads only what it needs, extracts frames (scene-aware, or fast keyframes at `efficient…
README: URL (anything yt-dlp supports — YouTube, Loom, TikTok, X, Instagram, plus a few hundred more) or a local path

Reviewed at: 2026-07-18T04:23:21Z

Open original GitHub project