agentleFS
TubeLabHQTubeLabHQ/youtube-research-skills/skills/niche-bending/SKILL.md

niche-bending

Skill91 starsChanged 11 days agoMITView on GitHub

Use when the user wants to bend a proven YouTube format into a new market, starting from a channel they name or from channels that already earn. Requires the TubeLab MCP.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 6 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    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. Distinctive:

    If it runs long, cut whole items.

    line 74
  2. Very specific:

    Do not search without publishedAtFrom and classificationQuality at "neutral" or above, or the results fill with low-effort videos.

    line 80
  3. Distinctive:

    Do not build on a hit the source never repeated.

    line 81
  4. Distinctive:

    Do not copy the source's titles, artwork, characters or name.

    line 82
  5. Distinctive:

    Do not bend into a market where the promise breaks.

    line 83
  6. Distinctive:

    Do not treat a small channel already doing the bend as bad news.

    line 84
  7. Distinctive:

    Do not call a bend open without saying what you searched.

    line 85

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Niche Bending

Overview

A niche is a market plus a format. The market is what the audience cares about, such as history or football. The format is a repeatable package: a title pattern, a thumbnail layout and a structure, such as "Every {thing} explained". Niche bending carries a format that blew up in one market into another market or language. The content is new, and the format is proven. Never invent a niche. Not for ten niche ideas from a wide search (niche-finder).

Workflow

  1. Clarify scope only if needed. Start from the channel the user names, or else from channels that already earn. Bend into the user's market or language first, if they name one.
  2. With a channel, read it, which is free. Find the one to three formats behind its best videos by ratio: the title pattern with the slots marked, the thumbnail layout and the length. A format needs three or more videos that beat the channel's average.
  3. With no channel, run the preset below for one or two pages; page 2 is from: 20. Swap in the user's language if named, and change nothing else. Read the channels with the clearest repeatable format, and find their formats as in step 2.
  4. View the thumbnail of every video you use, and say so if one fails to load.
  5. Bend each format into markets where its promise still makes sense and the audience is large. A new market is a different audience or another language, not a neighbouring topic.
  6. Check the best three bends with one outlier search each, in the bend's market and language, over the last year. Count the closely matching results, since the search total is capped, and set the status as the ranking legend says. Drop a crowded bend or bend it further. Say which bends went unchecked.
  7. Rank four to six bends by the format's strength at its source, the market's size, how early the bend is, and fit with the user's market. A few strong bends beat six forced ones.

Cash cows, which are channels that already earn:

{ "publishedAtFrom": "year", "classificationQuality": "neutral",
  "monetizationAdsense": true, "revenueMonthlyEstimationFrom": 800, "subscribersFrom": 1000,
  "language": ["en", "de", "fr"], "sortBy": "foundAt", "size": 20, "from": 0,
  "excludeNiche": ["music", "song", "lofi", "movies", "breastfeeding", "lingerie", "transparent", "transparency", "clean", "haul", "kids", "news", "politic", "india", "pakistan", "arabic", "manufacturing", "nature", "islamic"] }

Output Format

# 🔀 Niche Bending: {[channel](url), or "channels that already earn"}

- **Best bend:** {one sentence, with numbers}
- **Source:** {the channel, or how many channels you sampled and from how many pages}

## 🏆 Ranking
🟢 open: nobody found · 🟡 proven and early: one or two small channels with strong ratios, the best case · 🔴 crowded: several established channels · ⚪ not checked. Potential is a judgment call: High, Medium or Low.

| # | Bend | Format | New market | Status | Potential |
|---|---|---|---|---|---|
| 1 | {name} | "{title pattern}" | {market} | {🟢 open, 🟡 proven and early, 🔴 crowded, or ⚪ not checked} | {High, Medium or Low} |

## 🔀 Bends

### 1. {Bend name}

**Niche:** [{source channel}]({url}) · {market}

[![]({thumbnail})]({video url}) [![]({thumbnail})]({video url}) [![]({thumbnail})]({video url})

⬇️ **Bend:** {new market, and the language if it changes}

{Only when a channel already does it: up to three of its thumbnails, in one row like the one above.}

- **Format:** "{title pattern with slots}" · {thumbnail layout in a few words} · {usual length}
- **Proof at the source:** {n} videos use this format · best: [{title}]({video url}) · {views, ratio, date}
- **Why this market:** {one sentence}
- **Already there:** {"nobody found", or [channel](url) · [title](video url) · views, ratio, date, or "crowded" with who}
- **First titles:** "{title}" · "{title}" · "{title}"
- **Thumbnail:** {what to keep from the layout, and what to swap}
  • Length: about 1,000 words. Four to six bends, best first.
  • Language: plain words and short, complete sentences. Nothing compressed.
  • Bullets: one fact each, under 25 words, except quotes and verbatim titles. No extra sections or bullets.
  • If it runs long, cut whole items. Never cut proof, and never squeeze sentences.
  • Links: link every video, channel, moment, brand and offer you cite, built from IDs, handles and links in the data. Never guess one.
  • Thumbnails: the 320px mqdefault.jpg image, linked to its video, up to three on one line under their Niche or Bend line.

Common Pitfalls

  • Do not search without publishedAtFrom and classificationQuality at "neutral" or above, or the results fill with low-effort videos.
  • Do not build on a hit the source never repeated. One lucky video proves a topic, not a format.
  • Do not copy the source's titles, artwork, characters or name. Keep only the structure, or the channel reads as a copy.
  • Do not bend into a market where the promise breaks. Ranking the worst deaths fits football legends, not kitchen gadgets.
  • Do not treat a small channel already doing the bend as bad news. It proves the bend works while it is still early.
  • Do not call a bend open without saying what you searched. Empty results can mean the wrong search 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.