Token导航 LogoToken导航TokenDH.com
待分类只读github未标认证来源可访问许可证需确认审计提醒

hormozi-ad-factory霍尔莫兹广告厂

Agent Skill

hormozi-ad-factory 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

1,812

周安装

74

GitHub Stars

323

下载量

586
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:hormozi-ad-factory(霍尔莫兹广告厂)
来源仓库:https://github.com/pedronauck/skills
仓库路径:skills/hormozi-ad-factory
安装命令:
npx skills add https://github.com/pedronauck/skills --skill hormozi-ad-factory
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/pedronauck/skills --skill hormozi-ad-factory

简介

用于处理 GitHub 仓库、Issue、Pull Request 等协作信息。

  • 适合围绕代码变更、项目状态或团队协作进行整理和操作。
  • 通过 npx skills add 命令从指定仓库安装。
  • 支持 Codex、Claude、Cursor、Gemini CLI,安装方式为 github。
  • hormozi-ad-factory 属于待分类类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Hormozi Ad Factory

A systematic framework for generating 150-750+ unique ad variations from a single product or offer, based on Alex Hormozi's combinatorial ad creation method.

Core Concept

Ads consist of three modular parts: Hook, Meat, and CTA. Instead of writing one ad at a time, generate each component independently, then combine them:

  • 50 Hooks x 3-5 Meats x 1-3 CTAs = 150 to 750 unique ads

This approach creates massive variety for testing, which is what wins at scale.

Workflow

Execute the following four steps in order. Before starting, gather context from the user.

Step 0: Gather Context

Before generating any ad components, ask the user these questions (skip any the user has already answered):

  1. Product/Offer: What product, service, or offer is being advertised?
  2. Target Audience: Who is the ideal customer? (demographics, psychographics, pain points)
  3. Platform: Where will these ads run? (Instagram, YouTube, TikTok, Facebook, LinkedIn, Google)
  4. Format: Text ads, video scripts, carousel, or a mix?
  5. Existing Assets: Are there any past ads that performed well? Any organic content with strong engagement?
  6. Competitor References: Any competitor ads to draw inspiration from?
  7. Unique Mechanism: What makes this product/approach different from alternatives?
  8. Key Results/Proof: Specific numbers, testimonials, or case studies available?
  9. Language: Default to English. For Portuguese-speaking users, generate all content in Portuguese (BR) unless told otherwise.

If the user provides minimal context, infer reasonable defaults and state the assumptions clearly before proceeding.

Step 1: Generate 50 Hooks

Hooks are the first 1-3 seconds (video) or first line (text) of the ad. They stop the scroll and earn attention. Use these five sourcing methods to create variety:

  1. Winning hooks from previous ads — Adapt hooks from the user's past high-performing ads.
  2. Hooks from organic content — Pull hooks from the user's best-performing free content.
  3. Winning hooks from competitors — Adapt proven hooks from competitor paid ads.
  4. Hooks from competitors' organic content — Pull hooks from competitor top free content.
  5. Platform ad libraries — Use hooks found in Facebook Ad Library, TikTok Creative Center, etc.

Distribute the 50 hooks across awareness levels to cover different audience segments:

LevelDescriptionHook Style
UnawareDoes not know they have a problemPattern interrupt, curiosity, shock
Problem-AwareKnows the problem, not the solutionCall out the pain, frustration, desire
Solution-AwareKnows solutions exist, has not chosen oneCompare approaches, reveal flaws in alternatives
Product-AwareKnows the product, has not boughtOvercome objections, social proof, urgency
Most AwareKnows the product well, needs a pushDeals, bonuses, scarcity, direct offer

Default to a balanced spread (~10 hooks per level) unless the user specifies a focus.

For hook formula templates and examples, read references/hook-formulas.md.

Step 2: Generate 3-5 Meats

The meat is the body of the ad — it educates the audience on the offer, the product, the solution, or the problem. Generate one meat per format:

  1. Demonstration — Show the product/service in action. Best for product-aware audiences.
  2. Testimonial — Let customers tell the story. Best for solution-aware and product-aware audiences.
  3. Educational — Teach something valuable. Best for problem-aware and solution-aware audiences.
  4. Story — Tell a narrative (founder story, customer journey, transformation). Best for unaware and problem-aware audiences.
  5. Faceless — Text on screen, voiceover, b-roll, or screen recordings. Best for scaling without personal brand dependency.

Each meat must be self-contained and work with any hook and any CTA.

For detailed meat structures and length guidelines, read references/meat-formats.md.

Step 3: Generate 1-3 CTAs

The CTA tells the viewer exactly what to do next. A strong CTA includes up to five elements:

  1. What to do — The specific action (click, comment, DM, sign up, buy)
  2. How to do it — The mechanic (click the link below, comment "X", tap the button)
  3. What they get — The immediate result of taking action
  4. Why now — Urgency or scarcity element
  5. Risk reversal — Remove the fear (free trial, money-back, no commitment)

Not every CTA needs all five, but the best ones cover most of them.

For CTA templates and examples, read references/cta-templates.md.

Step 4: Assemble the Combinations

After generating all components, combine them into a matrix. Present the output as:

  1. Summary table showing total combinations (Hooks x Meats x CTAs)
  2. Top 10-15 "golden" combinations — the ones most likely to perform well based on audience-message fit, with rationale for each
  3. Full component library organized by type, for the user to mix and match

For the full output format template, read assets/output-template.md.

Platform Adaptation

Adapt the format of each component based on the target platform:

PlatformHookMeatCTA
Video ads (TikTok, YouTube, Reels)Opening line/scene (1-3s)Script section with visual directions (15-45s)Closing sequence with on-screen text
Text/copy ads (Facebook, LinkedIn)Headline or opening lineBody copy paragraphs (3-8 sentences)Closing paragraph with link
Carousel ads (Instagram, LinkedIn)Slide 1Slides 2-8Last slide
Short-form (Twitter/X, Stories)Single punchy line1-3 sentences maxInline CTA

Error Handling

  • User provides no product info: Do not generate ads. Ask for at minimum: product name, what it does, and who it is for.
  • Vague audience: Generate hooks for all five awareness levels with a balanced spread and note the assumption.
  • Too many platforms: Generate a single platform-agnostic set first, then offer to adapt for specific platforms.
  • User wants fewer than 50 hooks: Adjust proportionally but maintain the awareness-level distribution.
  • User wants full ads, not components: Generate the components first, then assemble the top 15 golden combinations as complete ready-to-use ads.

Examples

Example 1: SaaS Product

User: "Generate ads for my project management tool aimed at remote teams."

Process:

  1. Clarify: pricing, key differentiator, existing testimonials, platform
  2. Generate 50 hooks across awareness levels (e.g., "Your remote team is wasting 5 hours/week in status meetings" for problem-aware)
  3. Generate 4 meats: demo walkthrough, customer testimonial, educational ("3 signs your PM tool is slowing you down"), founder story
  4. Generate 2 CTAs: free trial (risk reversal) + limited-time discount (urgency)
  5. Assemble 400 combinations, highlight top 15 golden ads

Example 2: Physical Product (Portuguese)

User: "Crie anuncios para meu curso de culinaria online para iniciantes."

Process:

  1. All content generated in Portuguese (BR)
  2. Hooks adapted for cooking/education audience awareness levels
  3. Meats include: demo (cooking a recipe), testimonial (student transformation), educational (common beginner mistakes)
  4. CTAs include engagement ("Comente RECEITA") and direct ("Clique no link da bio")
  5. Golden combinations prioritize video format for Instagram/TikTok

Example 3: Minimal Context

User: "I need some ad ideas for my coaching business."

Process:

  1. Ask: What type of coaching? Who is the target client? What platform? What is the offer/price point?
  2. Wait for answers before generating
  3. If user says "just give me something," state assumptions explicitly and proceed with a general framework

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

34.45%
按下载量换算202

Claude

30.56%
按下载量换算179

Cursor

20.46%
按下载量换算120

Gemini CLI

10.3%
按下载量换算60

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills