agentleFS
milad-ahmdmilad-ahmd/agent-skills/skills/youtube-script-writer/SKILL.md

youtube-script-writer

Skill5 starsChanged 16 days agoMITView on GitHub

Writes high-retention YouTube scripts for long-form videos and Shorts, with a 30-second hook, structured outline, open loops, pattern interrupts, B-roll and graphics cues, sponsor segments, and CTAs, from a topic, notes, article, or research. Use when the user asks to write or rewrite what is said on camera: a YouTube script, talking points, a 30-second intro hook, a sponsor read, or a Shorts script.

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

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

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 4

    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:

    If the user provides research, use only facts from it or well-established knowledge, and mark claims that need verifying with [VERIFY].

    line 13
  2. Distinctive:

    Write for the ear: short sentences, contractions, "you" language, one idea per sentence.

    line 46
  3. Distinctive:

    Use specific examples and numbers instead of generic advice.

    line 47
  4. Distinctive:

    Remove filler, and cut any sentence that doesn't move the story forward.

    line 48
  5. Distinctive:

    Read it in the creator's voice.

    line 49

“Common” is measured: kept by at least 6 in 10 of the 5 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

YouTube Script Writer

Hold attention from second 0 to the end screen.

1. Brief

Topic, target viewer, the one promise of the video (what they'll get by the end), length (Short of 60 seconds or less, or long-form 8–15 minutes), style (talking head, faceless voiceover, tutorial, documentary), the creator's voice and any sources or facts to use. If the user provides research, use only facts from it or well-established knowledge, and mark claims that need verifying with [VERIFY].

2. Structure (long-form)

[0:00–0:30] HOOK
  - Visual/claim that matches the title & thumbnail promise immediately
  - Stakes: why this matters to the viewer
  - Open loop: "and the 3rd one surprised me the most"
  - No "Hey guys, welcome back" before value
[0:30–1:00] SETUP — credibility in one line, roadmap of the video
[BODY] 3–7 segments, each:
  - mini-hook → point → example/story/demo → payoff → transition that opens next loop
  - pattern interrupt every 60–90s (B-roll, graphic, zoom, joke, question)
[Mid-roll] sponsor or CTA bridged naturally from content
[CLIMAX] best point saved for late (pays off the open loop)
[END] quick recap → next-video CTA (point to specific video) — no long outro

3. Output format (a two-column script)

Spoken (A-roll)Visual / B-roll / on-screen text / SFX
"Most people get this wrong..."[Zoom-in] Text: "❌ Common mistake"

Then add: the estimated runtime (about 150 words per minute), a list of the B-roll shots to film or source, and 3 title and thumbnail pairings that match the hook (for youtube-seo-optimizer).

4. Shorts version

45 seconds or less, the hook in the first second, one idea only, a fast payoff, and a loop ending. Write it as a separate script.

Style rules

  • Write for the ear: short sentences, contractions, "you" language, one idea per sentence.
  • Use specific examples and numbers instead of generic advice.
  • Remove filler, and cut any sentence that doesn't move the story forward.
  • Read it in the creator's voice. Ask for an earlier script or transcript to match their style.
  • Sponsored segments must be clearly disclosed.

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.