Communitygithub.com

nategoldai/youtube-agent

>- Write the description, chapters, tags and search-facing text for a YouTube video, aimed at what a real person would type. Use for "write my description", "add chapters", "timestamps", "tags", "SEO", "help this video get found", or "nobody can find this".

O que é youtube-agent?

youtube-agent is a Claude Code agent skill that >- Write the description, chapters, tags and search-facing text for a YouTube video, aimed at what a real person would type. Use for "write my description", "add chapters", "timestamps", "tags", "SEO", "help this video get found", or "nobody can find this".

Funciona com✓Claude Code~Codex CLI~Cursor
npx skills add https://github.com/nategoldai/youtube-agent/tree/HEAD/skills/yt-seo

Perguntar na sua IA favorita

Abre um novo chat com esta habilidade de agente já pré-carregada.

Documentação

yt-seo

Search matters less than packaging, but more than people think for evergreen videos. If the video is aimed at the subscriber feed, say so and put the effort into /yt-package instead.

python3 chapters.py transcript.srt --target 8

Before writing

  1. Read ~/.claude/youtube/voice.md if it's there. It's the user's voice profile: how they talk on camera, words they never say, who they're talking to, and what they won't claim. If it's missing, ask for three of their own videos, read or transcribe them, work out the voice, and write the file.
  2. Never make up a number, a result or a source. If a figure would help and you don't have it, ask for it or write the line without it.

The description

  • Only the first two lines get read. They show above "...more" and they're the search snippet. Say what the video gives people, in the words they'd have searched.
  • Next, the link or resource if there is one, so it sits above the fold.
  • Then the chapters (see below).
  • Then the longer version: what's covered, who it's for, and what it assumes they already know.

Chapters

chapters.py finds the breaks in a timestamped transcript. A chapter list that breaks any of YouTube's rules quietly fails, and the block just sits in the description doing nothing:

  • The first one must be 00:00.
  • There must be at least three.
  • Each one must be at least 10 seconds long.

The tool checks all three and tells you when a list won't work instead of letting you paste it.

It's good at finding the BREAKS - it scores the pauses you actually took by how much the vocabulary changes across them. The titles it gives are just the topic words, so treat them as a draft. A chapter called "Thumbnails Titles Packaging" is a placeholder. Rewrite each one as the promise of that section, in the user's voice, in three to five words.

Chapters double as a retention check: if a section can't be named in five words, it's either two sections or it's filler.

Tags, honestly

Tags are a weak signal, and YouTube has said so. Use them to clear up ambiguity: alternate spellings, tool names, the abbreviations people really type. Then stop. Fifteen is plenty. A wall of tags isn't a strategy, and stuffing in unrelated ones breaks the rules.

The search test

Before handing anything over, write down the three searches this video should win, and check that the title and the first two lines of the description contain the words from those searches. If they don't, the title is the problem, not the description.

The gate

Nothing here gets published. This skill writes and you publish. Every result ends with a block the user copies, and the last line of every run asks: ship it, or change it?

Individual skills in this repo

This repo contains 9 individual skills — each has its own dedicated page.

nategoldai/youtube-agent

>- Review a YouTube channel from top to bottom - packaging, consistency, the first fifteen seconds, and what to fix first. Use for "audit my channel", "why isn't my channel growing", "look at my videos", or a pasted channel URL.

nategoldai/youtube-agent

>- Write replies to YouTube comments in the creator's voice, sorted by which ones are worth answering. Use for "reply to my comments", "deal with the comment section", "someone asked X", or a pasted comment thread.

nategoldai/youtube-agent

>- Turn the transcript of a raw recording into an edit decision list - dead air, filler lines and retakes, with timecodes. Use for "edit this", "cut the dead space", "tighten this video", "I rambled", or any request to shorten footage from a transcript.

nategoldai/youtube-agent

>- Write and check a YouTube video's title and thumbnail as one pair, testing for truncation, repetition and vagueness before it goes live. Use for "title ideas", "what should I call this", "thumbnail text", "my CTR is low", packaging, or any request to rename or repackage an existing video.

nategoldai/youtube-agent

>- Plan a week or a month of YouTube uploads - what to post, when, and in what order, sized to how much time the creator really has. Use for "plan my week", "content calendar", "what should I post", "no idea what to make next".

nategoldai/youtube-agent

>- Read a YouTube Studio audience-retention export, find where viewers really drop off, and say what to change. Use for "why do people stop watching", "my retention is bad", a pasted retention chart or CSV, or "fix my pacing".

nategoldai/youtube-agent

>- Write a YouTube video script from a rough idea - hook options from 21 formulas, scored, then the full spoken script with the retention beats marked. Use whenever the user wants a video script, a hook, an opening line, "what should I say", "write my next video", or is about to record without the first fifteen seconds figured out.

nategoldai/youtube-agent

>- Find the Shorts hiding inside a long video and write them, using the transcript to pick moments that stand on their own. Use for "cut this into shorts", "clip this", "repurpose this video", or "what should I clip".

nategoldai/youtube-agent

>- Find what's really working in the user's niche on YouTube, rank it by how far each video beat its own channel, then name the formula. Use for "what's working right now", "find viral videos in my niche", "why did this blow up", competitor research, or a swipe file.

Habilidades Relacionadas