232 ·GitHub
What is this?
Marketing AGI is a single Claude skill made of 14 modules that cover the full surface a marketer touches: scored audits, copy grading, a hook engine, ad diagnosis, GEO, email, launches, positioning and pricing.
Marketing AGI is a single Claude skill made of 14 modules that cover the full surface a marketer touches: scored audits, copy grading, a hook engine, ad diagnosis, GEO, email, launches, positioning and pricing.Is this a standalone app or a Claude skill package?It's pure markdown with no executable code — a skills folder you drop into an agent's skills directory, not an app you run.
It's pure markdown with no executable code — a skills folder you drop into an agent's skills directory, not an app you run.Does it only work with Claude Code, or other coding agents too?It's primarily built for Claude Code and the Claude apps, but since it's just markdown files it also works with Codex, Cursor and other agents that support a skills folder.
It's primarily built for Claude Code and the Claude apps, but since it's just markdown files it also works with Codex, Cursor and other agents that support a skills folder.Does it actually generate ad creatives, or just briefs?By default it only produces a diagnosis and a ranked production brief on paper; it generates the actual briefed assets only when an ad-generation MCP such as MaxFusion AI is connected.
By default it only produces a diagnosis and a ranked production brief on paper; it generates the actual briefed assets only when an ad-generation MCP such as MaxFusion AI is connected.What do the 14 modules actually cover?They span audit, geo, copy, hooks, paid-ads, email, social, launch, positioning, competitive and app-store, plus analytics, a slop-patterns catalogue, and shared rubrics/specs used across modules.
They span audit, geo, copy, hooks, paid-ads, email, social, launch, positioning, competitive and app-store, plus analytics, a slop-patterns catalogue, and shared rubrics/specs used across modules.Why fill in brand-context.md before using it?Every module reads brand-context.md for brand facts; skip it and the output stays competent but generic, since nothing identifies your specific brand or offer.
Every module reads brand-context.md for brand facts; skip it and the output stays competent but generic, since nothing identifies your specific brand or offer.What is the 18-tactic hook engine?It's the hooks module: it runs 18 hook tactics against your copy, builds a hook matrix, and diagnoses which hooks are failing in a funnel.
It's the hooks module: it runs 18 hook tactics against your copy, builds a hook matrix, and diagnoses which hooks are failing in a funnel.How does it stop 14 modules from bloating the agent's context?A lightweight router loads first at roughly 2k tokens, and each task only pulls in the specific module it needs, so the other modules cost nothing until used.
A lightweight router loads first at roughly 2k tokens, and each task only pulls in the specific module it needs, so the other modules cost nothing until used.What are similar projects?marketingskills, marketing-os, linkedin-agent-skill, geo-seo-claude, bigpeng-hot-gzh, instagram-agent-skill, claude-skills-founder, brag
marketingskills, marketing-os, linkedin-agent-skill, geo-seo-claude, bigpeng-hot-gzh, instagram-agent-skill, claude-skills-founder, bragWhich topics is it in?appstore, copywriting, email-marketing, pricing, seo, social-media, writing, marketing, content-creation, analytics, claude-skills, skill
appstore, copywriting, email-marketing, pricing, seo, social-media, writing, marketing, content-creation, analytics, claude-skills, skill