youtube-watch SKILL.md skillA

youtube-watch from imtylervo/youtube-watch is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add imtylervo/youtube-watch/youtube-watch@git:20261103.a29613e

Pins git:20261103.a29613e in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
imtylervo/youtube-watch
Path
SKILL.md
Versions recorded
4
Latest version
git:20261103.a29613e, committed Nov 3, 2026
First version recorded
Oct 26, 2026
Size
11,858 bytes
Repository stars
3
License
MIT

Make Claude actually watch a video. Downloads any YouTube/Vimeo/TikTok/X video or local `.mp4`/`.mov`, fetches a timestamped transcript (via captions API, yt-dlp auto-subs, or local Whisper fallback), and extracts smart-sampled frames so Claude can SEE the visuals AND READ the audio in one pass. Use this skill any time the user pastes a video link or path with a question — "summarize this", "tóm tắt video", "phân tích", "xem giúp em", "what's at the 30s mark", "explain this clip…

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# youtube-watch

Default workflow — call `extract-v3.sh` and then read `understanding.md`. That's the whole skill in one sentence.

```bash
${CLAUDE_SKILL_DIR}/scripts/extract-v3.sh "<video-url-or-path>"
```

(Outside Claude Code, use the folder you cloned into, e.g. `~/.claude/skills/youtube-watch/scripts/extract-v3.sh`.)

**Look at every frame.** To watch, summarize or explain a video (including "xem", "tóm tắt", or a bare link), open **every** picture in `sheets/` (`sheet_01.jpg`, `sheet_02.jpg`, …) before answering. Each holds 16 frames, labelled with frame number and time, so a 15-min video is about 8 reads. Then open single frames from `frames/` to read small on-screen text (slides, tables, tweets, code). Do not answer from `overview.jpg` and a couple of frames: the picture often carries what the speaker never says (dashboards, numbers, quoted articles). In the answer, say how many frames you looked at (e.g. "looked at 131/131 frames"). Only for a narrow question about one moment ("what's at 2:30") may you open just that part. `overview.jpg` (~20 frames in one picture) is a first glance only.
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add imtylervo/youtube-watch/youtube-watch@sha256:9d5d17800538c78e

The line at the top of this page pins the label git:20261103.a29613e, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_sucja3s5drxsveex.svg)](https://markdownregistry.com/a/art_sucja3s5drxsveex)

1 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 4 versions of this file, between Oct 26, 2026 and Nov 3, 2026, about one every 3 days.

versioncommittedcommitsizeaudit
git:20261103.a29613e latest2026-11-03 a29613e 11,858 BA view · diff
git:20261030.d2c0dc42026-10-30 d2c0dc4 11,355 BA view · diff
git:20261026.08e77dd2026-10-26 08e77dd 11,353 BA view · diff
git:20261026.68c32522026-10-26 68c3252 11,119 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (11858 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

imtylervo/youtube-watch · 3 stars · license MIT · pushed 2026-11-03 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_sucja3s5drxsveex
GET https://markdownregistry.com/api/v1/resolve?ref=imtylervo/youtube-watch/youtube-watch
GET https://markdownregistry.com/api/v1/blob/9d5d17800538c78ed065e17514f1088abfed3c8724d1e6674e5705939b83a08a

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

Browse by kind, by grade A, or by owner.