agentleFS
utillayerutillayer/youtube-agent-skill/skills/yt-package/SKILL.md

yt-package

Skill10 starsChanged 29 days agoMITView on GitHub

Write and test a video's title and thumbnail together, linted for truncation, repetition and vague words before it goes live. Use for "title ideas", "what should I call this", "thumbnail text", "my click-through is low", packaging, or renaming a video that already exists.

7 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 7 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read ~/.claude/youtube/voice.md if it exists.

    line 26
  2. Distinctive:

    Never make up a number. "Doubled" needs the before and after.

    line 28
  3. Distinctive:

    Write ten titles. Lint them all.

    line 43
  4. Distinctive:

    If the top two are close, say so and suggest Studio's A/B test (up to three titles or thumbnails, judged by watch time share.

    line 46

“Common” is measured: kept by at least 6 in 10 of the 7 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

yt-package

A viewer sees the title and the thumbnail in the same glance, so write them as one thing. The commonest waste is a thumbnail that repeats the title: two surfaces, one message.

python3 ${CLAUDE_SKILL_DIR}/title.py --title "..." --thumb "THREE WORDS"
python3 ${CLAUDE_SKILL_DIR}/title.py titles.txt --thumb "THREE WORDS"    # one per line, ranked

The tools live in this skill's folder. ${CLAUDE_SKILL_DIR} is that folder in Claude Code; anywhere it does not expand (claude.ai, Codex), use the folder this SKILL.md was loaded from.

Before you start

  1. Read ~/.claude/youtube/voice.md if it exists. If not, ask for three of the creator's videos and write it. The title should sound like the person in the video.
  2. Never make up a number. "Doubled" needs the before and after.

What the linter checks, and why

  • Length. About 60 characters survives desktop search and about 40 a phone feed. YouTube does not publish either number - truncation is by pixel width - so treat them as the safe side, not a law. A desktop cut loses the end; a phone cut can lose the subject. The hard limit is 100, and < > are not allowed.
  • Repetition. The thumbnail says what the title does not.
  • Thumbnail words. Three at most. A fourth turns to mush at feed size.
  • Something concrete. A number, a name or a date beats any adjective.
  • Capitals. Two all-caps words is the limit before it reads as shouting.

The method

  1. Write ten titles. Lint them all. Show the best three with their scores and the one issue each has.
  2. For the winner, write the thumbnail: the single image, the expression, at most three words, and how the background keeps contrast at feed size.
  3. If the top two are close, say so and suggest Studio's A/B test (up to three titles or thumbnails, judged by watch time share; desktop Studio only, not for Shorts) instead of pretending the linter can tell them apart. The score is a lint, not a forecast.

Before you hand it over

This skill drafts; you publish. End with the copy-ready block and one question: post it as is, or another pass?

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.