channel-formula

channel-formula is a skill for Claude Code, Codex from yaxeen/storytelling-skills. It costs 76 tokens per session (1,332 once invoked), scanned A, original, MIT.

A method for learning what content consistently works on a specific YouTube, TikTok, or Instagram channel by comparing its best-performing videos with weaker ones. It turns those observations into reusable guidance for titles, presentation, and audience expectations.

In plain words
What is it for?
Use it to study a channel's winners and flops, plan a new video, or understand why a video underperformed. It applies across different topics and platforms.
Why use it?
General content advice may not match what a particular audience expects. This identifies the promise that successful videos keep and weaker videos break.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: reads .claude/ paths.

Good fit Use it to study a channel's winners and flops, plan a new video, or understand why a video underperformed. It applies across different topics and platforms.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/yaxeen/storytelling-skills/channel-formula
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add yaxeen/storytelling-skills --skill channel-formula
Clone the repo
git clone --depth 1 https://github.com/yaxeen/storytelling-skills

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for channel-formula

README.md
[![agentmods](https://agentmods.dev/badge/skills/yaxeen/storytelling-skills/channel-formula/github.svg)](https://agentmods.dev/skills/yaxeen/storytelling-skills/channel-formula)
Your own site
<a href="https://agentmods.dev/skills/yaxeen/storytelling-skills/channel-formula"><img src="https://agentmods.dev/badge/skills/yaxeen/storytelling-skills/channel-formula/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for channel-formula

Your own site · 80×15
<a href="https://agentmods.dev/skills/yaxeen/storytelling-skills/channel-formula"><img src="https://agentmods.dev/badge/skills/yaxeen/storytelling-skills/channel-formula.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 76 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,332 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00076 $0.01332
Opus 5 $0.00038 $0.00666
Sonnet 5 $0.00015 $0.00266
Haiku 4.5 $0.00008 $0.00133

Measured 12d ago against content hash 09481f40157f, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

channel-formula scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/channel-formula/SKILL.md · 86 lines

How it starts

The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Channel Formula

  • Every channel that works has an unwritten contract with its audience: a specific promise the winners keep and the flops break. This skill makes the contract written.
  • Generic storytelling craft is necessary but not sufficient — a technically perfect story that breaks the channel's promise still dies. Audience expectation beats craft.
  • The output is a reusable channel profile: formula, title pattern, packaging rules, and retention benchmarks, extracted from the channel's own data — never from assumptions about the niche.
  • Built on the six levers — see storytelling-hooks. Recap: 1. Curiosity gap · 2. Emotional mirror · 3. Conflict engine · 4. Relatability · 5. Pattern + surprise · 6. Three-act.

When to Use

  • A channel has some winners and some flops and nobody can articulate why.
  • Before scripting for an established channel — check the premise against the formula first.
  • After a flop, alongside retention-audit (audit finds where it died; this skill finds what promise was broken).
  • Any platform, any niche: crime stories, cricket, tech reviews, cozy content, finance, faith, vlogs.

Intake (the evidence)

  • 3–5 top videos: title, thumbnail (screenshot fine), views/velocity, and if available retention at 0:30.
  • 1–3 flops from the same format — flops are as informative as winners; the formula lives in the difference.
  • Channel's audience in one sentence, and what viewers do with the content (background listening? active watching? saving?).
  • Never derive a formula from one video or from niche stereotypes — real data only. Too few videos? Say so and mark the profile "provisional".

Extraction Workflow

  1. Line up winners' titles side by side. Find the shared skeleton: what kind of threat/promise/subject repeats? What resolution do they all point to? Write the pattern as a fill-in: [recurring setup]... [recurring promise clause].
  2. Line up the thumbnails. What is the focal subject in every winner? What emotion? What's never there?
  3. Name the core promise in one sentence: what feeling does the viewer come to have? (Crime channel: "I saw the twist coming before the detective." Tech channel: "I avoided an expensive mistake." Fitness channel: "I can restart after falling off.")
  4. Autopsy the flops against it. Which element of the promise did each flop break — premise, resolution, packaging tone, pacing? The broken element becomes a "never" rule.
  5. Set benchmarks from the channel's own winners (retention at 0:30 / 2:00, typical velocity) — not from industry averages.
  6. Calibrate the levers. Same six levers, channel-specific volume: how hot should tension run? What kind of conflict does this audience accept? How long should resolutions linger?

Read the full file on GitHub · 86 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 12d ago First seen · 86 lines · 76 tokens per session scan A 09481f40157f

Subscribe to this mod's changes

channel-formula is a skill published in the GitHub repository yaxeen/storytelling-skills (9 stars, last pushed 1mo ago), licensed MIT. It adds 76 tokens to every session and 1,332 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.