agentleFS
aya-talebaya-taleb/video-boost-skills/skills/video-packaging/SKILL.md

video-packaging

Skill1 starsChanged 22 days agoMITView on GitHub

Proposes everything around the video before it goes out - 3 titles to A/B test, the description (with chapters, credits and links), hashtags, thumbnail ideas, and a short list of changes to the video itself - each tied to a checklist rule, a similar-video finding or a platform rule. Use when the user asks for a title, description, hashtags, thumbnail ideas, how to package this video, or what to change before publishing.

4 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

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never invent a fact about the video's content.

    line 38

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “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

Video packaging

Writes the title, description and the rest, and lists the changes worth making to the video. Every proposal says why, so the user can judge it.

Inputs

  • The video (or its transcript and the video-score metrics)
  • checklist.json from viral-checklist (if there is one)
  • similar-videos.json from similar-videos (if there is one)
  • Platform, market, language (defaults: youtube, US, en)
  • The channel name and any links the user wants in the description

What to write

  1. Titles. Write 10 to 16 candidates, then rank them against real breakout titles near the topic (from similar-videos, or trend-ideas/scripts/trend_scan.py with the topic as the angle): does each follow a pattern that breaks out now (outlier 3x or more), is it clear, is it true to the video? Keep the best 3 for YouTube's A/B test. (Ranking titles against breakout videos near the topic is OutlierKit's idea.) For the 3 kept: Each a different angle (for example: the promise, the number, the question). YouTube allows 100 characters; keep the main words in the first 60 or so so they show in search and on phones (check the current display length if the research found one; do not guess a number). No misleading clickbait: YouTube's spam policy bans titles that trick people into clicking. Say which rule or pattern each title follows.
  2. Description:
    • First 2 lines: what the viewer gets, with the main words people search for. Write it for people, not for keyword stuffing (excessive tags in the description break YouTube's spam policy).
    • Chapters: first at 00:00, at least 3, ascending, each 10 seconds or longer (YouTube's chapter rules). Only for times that exist in the video.
    • Credits: the book and author, or the sources used for facts.
    • Links the user gave. Mark affiliate links as affiliate.
    • Disclosures: AI use if it applies, paid promotion if it applies.
  3. Hashtags: 3 to 5 that are directly about the video. YouTube shows up to 3 above the title, ignores all hashtags past 60, and treats unrelated hashtags as misleading.
  4. Thumbnail ideas: 3 concepts for the A/B test: what is in the frame, the words (5 or fewer), the face or character, the contrast. Do not promise anything the video does not show.
  5. Changes to the video: up to 5, most important first. Each one names the failed checklist item or the similar-videos finding it fixes, and how much work it is (small, medium, large).

Output

packaging.md with the 5 parts above, in that order, ready to paste into YouTube Studio.

Rules

  • Never invent a fact about the video's content. Only use what is in the transcript or on screen.
  • Titles and thumbnails must match the video.
  • Plain words, the user's tone. No hype words.

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.