MCP App StoreMCP App Store
from claude-watch473
Enables agents to analyze videos via scene-change frame extraction, transcriptions, and editorial pacing metrics with Obsidian integration.
This skill transforms an AI agent into a video editor and analyst. It allows the agent to "watch" videos from URLs (YouTube, Vimeo, etc.) or local files by extracting key frames based on scene changes and generating timestamped transcripts using native captions or the Whisper API.
Use this skill when a user provides a video link or file and asks for a summary, an analysis of the editing style, or specific questions about the visual and audio content.
watch.py (orchestration), download.py (yt-dlp), frames.py (ffmpeg extraction), pacing.py (editorial metrics), hook.py (dense 0-10s analysis), and transcribe.py (Whisper/Caption integration).report.md, and optionally ingesting the results into an Obsidian vault.Designed for agents with shell access and the ability to read image files, such as Claude Code, Codex, or other ACP-capable harnesses.
This skill has not been reviewed by our automated audit pipeline yet.
Learn: From Raw Materials to Published Output
Run a disciplined six-phase research-to-publish workflow (collect, digest, outline, fill, refine, publish) for deep domain research and article preparation.
Voice Memo Organizer
Locate, transcribe (local whisper.cpp), summarize and index Apple Voice Memos into a searchable archive with titles, themes and key quotes.
FFmpeg Guide
Comprehensive FFmpeg reference for encoding, converting, streaming, filtering, and analyzing audio/video — command examples, common patterns, and troubleshootin
Map View (Obsidian)
Search for real-world places and insert geolocations into Obsidian notes using the Map View plugin and Obsidian CLI.
Obsidian CLI (vault & plugin ops)
Command-line reference and usage patterns for the Obsidian CLI: read/create/append notes, search vaults, manage properties, and developer workflows for plugin r
Obsidian CLI
Control and automate an Obsidian vault from the command line: read, create, search, update notes and developer workflows (reload plugins, run JS, screenshot, DO
YouTube Tools (yt-dlp)
Free, local YouTube operations using yt-dlp: download videos/audio, extract transcripts/subtitles, and fetch metadata (no API keys).
ScrollClaw — Assemble
Assemble UGC video campaigns: stitch A-roll, orchestrate audio, run post-production, then add mobile-friendly captions.