CMO Agent

An AI-powered marketing strategy system that runs as slash commands in Claude Code, Cursor, or any environment that supports .claude/commands/. It takes a business from raw context through competitive positioning, ideal customer profiles, brand messaging, content strategy, and design systems — producing structured, reusable strategy artifacts at every step.

Think of it as a fractional CMO that works interactively inside your terminal. The methodology is proven (Fletch PMM for positioning, Emily Kramer / MKT1 for content systems), adapted into a repeatable, skill-driven workflow.

---

Requirements

• Claude Code or Cursor with Claude
• That's it. No dependencies, no build step, no API keys.

---

Quick Start

git clone https://github.com/multiplai-ai/cmo-agent.git
cd cmo-agent

Then in Claude Code or Cursor:

/strategy/discovery-intake

Answer the five phases of intake questions. The agent will generate a structured discovery artifact and recommend the next skill to run.

---

The 6-Skill Pipeline

Each skill builds on the one before it. Run them in order for a full engagement, or run any skill independently if you provide upstream context manually.

#	Skill	Command	Output
1	Discovery & Intake	/strategy/discovery-intake	Structured business context — stage, goals, product, customers, challenges, proof points
2	Positioning Strategy	/strategy/positioning-strategy	JTBD competitive landscape, primary anchor selection, differentiation analysis, positioning narrative (one-liner → elevator → homepage)
3	ICP & Personas	/strategy/icp-personas	Workflow-based ICP definition, buyer persona cards, customer journey map, validation framework with interview scripts
4	Brand Strategy	/strategy/brand-strategy	Message hierarchy, persona/stage-specific messaging, USP framework, proof point inventory, brand voice guidelines
5	Content Strategy	/strategy/content-strategy	Perception statements, content pillars, fuel & engine diagnosis, distribution tiers, show definitions, monthly theme framework
6	Design Systems	/strategy/design-systems	Mood direction, locked palette & typography, full token architecture, component specs, CSS custom properties, Tailwind config

Bonus: /strategy/design-extract reverse-engineers design tokens from any live URL and feeds directly into Design Systems Stage 3.

---

How It Works

Discovery Intake
      |
      v
Positioning Strategy
      |
      v
ICP & Personas
      |
      v
Brand Strategy & Key Messages
      |
      v
Content Strategy       Design Systems
      |                      |
      v                      v
Content Calendar      Landing Pages
Writing               Product UI
Campaigns             Visual Content

Each skill is an interactive conversation. The agent asks structured questions, processes your inputs, and generates markdown strategy artifacts. Every artifact is saved to your entity directory so downstream skills can read from it automatically.

---

Entity Directory Structure

Create one directory per company or brand:

entities/
  your-company/
    strategy/
      discovery-intake.md
      positioning-strategy.md
      icp-personas.md
      brand-strategy.md
      content-strategy.md
      mood-direction.md
      palette-typography.md
      design-system.md
      design-tokens.json
      design-tokens.css
      tailwind.extend.js

An example structure is included at example/acme-corp/strategy/.

---

Methodology Credits

• Positioning framework: Based on Fletch PMM — JTBD competitive mapping, primary anchor selection, minimal viable market segments
• Content system: Based on Emily Kramer / MKT1 — perceptions framework, fuel & engine diagnostic, shows vs. feeds, distribution tiers

---

What Comes Next

This strategy suite produces the marketing foundation. For visual identity, creative production, and ad campaigns built on these strategy artifacts, see the Creative Director Agent.

---

License

MIT — see LICENSE.

---

Built by MultiplAI Growth Systems.