marketing-os logo

marketing-os

An entire marketing department as one Claude skill. 14 modules: audits scored 0-100, an 18-tactic hook engine, copy graded before you see it, ad diagnosis, GEO, email, launches, pricing. Works in Claude, Codex and Cursor.

GitHub

What is it?

What it is

Marketing OS is a single skill comprising fourteen modular marketing tools (audit, geo, copy, hooks, paid-ads, email, social, launch, positioning, competitive, app-store, analytics, slop-patterns, rubrics/specs) that deliver scored audits, artifact outputs (variants, briefs, sequences), and explicit uncertainty markers for AI agents.

Why it exists

It was created by deconstructing the highest‑starred marketing skill repositories on GitHub, retaining what worked and adding three missing elements: quantitative scores on everything, concrete artifacts instead of vague advice, and an explicit 'what I couldn't determine' note in every report.

Who should use it

Marketers and AI-assisted workflow builders who want structured modules for audits, copy generation, SEO/ASO, and campaign optimization. Ideal for those using Claude Code or similar MCP-integrated tools who provide a brand context file.

Who should avoid it

Those seeking executable code or runtime dependencies. People looking for quick, no-effort outputs without filling in the brand-context template. Users who prefer traditional marketing software over markdown-based, progressive-disclosure frameworks.

Stars over time

No change stars in the last 1 day

430430
Aug 30, 2026Aug 30, 2026

How it works

A quick walkthrough in plain English

How marketing-os works

Step 1 of 3

You interact with it

Open marketing-os, send a request, or connect it to your stack.

Features

Website/funnel audit with weighted 0-100 score and prioritized fixes
AI-search citability analysis for ChatGPT, Perplexity, AI Overviews
Copy generation producing 15-20 variants, expert-panel scoring, and de-slop
18-tactic hook engine generating hook matrix and diagnostic funnel
Paid-ads concept-level fatigue diagnosis with ranked production brief
Email sequence generation (welcome, nurture, launch) in full
Social media post generation for LinkedIn/X
Launch playbook including Product Hunt
Positioning, offer design, and pricing strategy
Competitor teardowns from public sources
App Store optimization diagnosis, metadata, and screenshot sequences
Analytics with honest data reads and proper test design
Slop-patterns catalogue applied to all prose
Rubrics and specs including scoring bands, engine playbooks, store field rules
Integration with ad-generation MCP (e.g., Arcads) to produce assets directly

Advantages

  • Progressive disclosure: router loads first, modules load only when used
  • Subagent-native: parallel execution for multi-dimensional tasks
  • No executable code: pure markdown, safe to trust
  • Honesty spine: no invented proof, scores labeled as heuristics
  • Comprehensive coverage of marketing functions
  • Artifacts instead of advice
  • Explicit 'what I couldn't determine' in reports

Disadvantages

  • Requires brand-context.md to be filled for meaningful output
  • No executable code may limit automation
  • Depends on host supporting subagents for parallel execution
  • Some features require code execution enabled (e.g., web/desktop)
  • Text-only output; no native design tool integration
  • Complexity from fourteen modules may overwhelm users

Installation

cloud

Claude (web/desktop): Settings → Skills → + → upload `marketing-os.zip` from the latest release. Requires code execution enabled.

native

Claude Code: run `/plugin marketplace add Yuzzyuk/marketing-os` then `/plugin install marketing-os@marketing-os`. For manual install (Codex, Cursor, other agents): copy `skills/marketing-os/` into your agent's skills directory (e.g. `~/.claude/skills/marketing-os/`). Copy `brand-context.template.md` to `brand-context.md`, fill it in, and place it in your project root or `.claude/`.

FAQ

What is Marketing OS and how many modules does it include?

Marketing OS is an open-source marketing skill packaged as a single Claude Code plugin containing fourteen modules covering audits, GEO/AI-search citability, copy generation, hooks, paid ads, email, social, launch, positioning, competitive teardowns, app store optimization, analytics, AI slop-pattern detection, and rubrics/specs.

How do I install Marketing OS in Claude Code?

Run two commands: '/plugin marketplace add Yuzzyuk/marketing-os' followed by '/plugin install marketing-os@marketing-os'. For Claude web/desktop, download marketing-os.zip from the latest release, enable code execution, and upload it via Settings → Skills → +. For Codex, Cursor, or other agents, manually copy the skills/marketing-os/ directory into your agent's skills folder (e.g. ~/.claude/skills/marketing-os/).

What is brand-context.md and why is it required?

brand-context.md is a context file you create by copying brand-context.template.md, filling in your brand details, and placing it in your project root or .claude/ directory. Every module reads it. Without it, the output is described as 'competent and interchangeable' — which the README identifies as the one thing marketing cannot afford to be.

Does Marketing OS actually generate ad creative or just briefs?

It does both. The paid-ads module diagnoses concept-level fatigue and produces a ranked production brief, and if an ad-generation MCP (such as Arcads) is connected, the skill generates the briefed assets directly — enabling diagnose → brief → produce in a single session.

What design principles does Marketing OS follow?

Four principles: (1) Progressive disclosure — a ~2k-token router loads first and each task loads only its own module; (2) Subagent-native — multi-dimensional work fans out in parallel when supported; (3) No executable code — pure markdown only, nothing to review before trusting; (4) Honesty spine — no invented proof, no fake precision, no winners declared on small samples, and every score explicitly labeled as a heuristic.

What is the 'slop-patterns' module and how is it used?

Slop-patterns is an AI-tell catalogue that is run on all prose output across the system to identify and remove AI-generated writing fingerprints ('de-slop'). It works alongside rubrics/specs, which define scoring bands, engine playbooks, and store field rules used by the other modules.

Loading documentation…
View on GitHub

Featured in Videos

YouTube tutorials and walkthroughs for marketing-os

Alternatives

Similar projects ranked by category, topics, and text overlap.

Compare
marketing-os | MushyBook