agentleFS
fahdAghribifahdAghribi/youtube-channel-skills/skills/youtube-metadata-optimizer/SKILL.md

youtube-metadata-optimizer

Skill2 starsChanged 13 days agoMITView on GitHub

Write and validate a video's metadata: search-aware description (first 150 chars), chapters from the script or transcript, tags, hashtags, playlist, category, end screen and pinned comment text. Produces a video.json that passes the offline validator. Not for titles and thumbnails (use youtube-title-thumbnail) or for uploading (use youtube-publisher).

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 3 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

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

    Before every upload (the publisher expects a validated video.json)

    line 14
  2. Very specific:

    Chapters: start at 00:00, ≥ 3 entries, each ≥ 10 s, labelled with the viewer's words ("Cómo elegir el micrófono", not "Parte 2").

    line 33
  3. Distinctive:

    No keyword stuffing: a phrase used more than twice in the description reads as spam to people and systems.

    line 50
  4. Distinctive:

    No links in the first two lines unless the video's purpose is that link.

    line 51
  5. Distinctive:

    Never add a link, code or handle the user did not approve.

    line 52

≈ “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

YouTube Metadata Optimizer

Everything below the title, packaged as a video.json that python3 scripts/validate_metadata.py accepts and youtube-publisher can upload.

When to use

  • "Write the description", "make chapters", "tags for this video"
  • Before every upload (the publisher expects a validated video.json)
  • Fixing metadata on old videos (then hand the diff to youtube-publisher update mode)

Input

  • Final title (from youtube-title-thumbnail)
  • Script with section times, or a transcript/SRT with timestamps, or the user's own chapter list
  • Video file path and duration (for chapter validation)
  • Links allowed by the profile (§Packaging style) and any sponsor/affiliate disclosures

Steps

  1. Load the profile for language, links, defaults (category, madeForKids, synthetic media, playlist).
  2. Description, in this order (references/description-template.md):
    1. Lines 1-2 (~150 chars): what the viewer gets, with the main search phrase used naturally.
    2. Short expansion (2-4 lines) with secondary phrases.
    3. Links block (only profile-approved links; disclosures first if any).
    4. Chapters.
    5. 3-5 hashtags at the end.
  3. Chapters: start at 00:00, ≥ 3 entries, each ≥ 10 s, labelled with the viewer's words ("Cómo elegir el micrófono", not "Parte 2"). Derive from section times or transcript; if neither exists, ask.
  4. Tags: 5-15 tags, 300-450 chars total: exact topic, variants, common misspellings, creator/brand name. Low weight, do not overthink.
  5. Status fields: madeForKids (ask if not in profile, never guess), containsSyntheticMedia when realistic AI/altered content is shown, categoryId, defaultLanguage, defaultAudioLanguage, privacyStatus, optional publishAt.
  6. Studio-only extras (not in the API): pinned comment text (a question that invites specific answers), end screen plan (1 best-for-viewer video + 1 specific video + subscribe), cards (max 1-2, at moments the viewer would otherwise leave).
  7. Write video.json and run python3 scripts/validate_metadata.py video.json. Fix every error; explain any remaining warning.
  8. Return the file, the validator output, and the Studio-only checklist.

Output

  • video.json (schema in references/description-template.md)
  • Validator output
  • Pinned comment text, end screen + cards checklist

Hard rules

Global voice rules: see root SKILL.md §Voice rules.

  • No keyword stuffing: a phrase used more than twice in the description reads as spam to people and systems.
  • No links in the first two lines unless the video's purpose is that link.
  • Never add a link, code or handle the user did not approve.
  • madeForKids is a legal declaration: ask, do not infer.

Files

  • references/description-template.md — description layout and video.json schema
  • ../../references/metadata-limits.md — limits the validator enforces
  • scripts/validate_metadata.py

Related skills

  • youtube-title-thumbnail — the title this metadata supports
  • youtube-publisher — uploads the validated video.json
  • youtube-script-writer — source of chapter times

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.