Prompts, agents, and workflows that make Claude Code work like a marketing team with perfect brand memory.

Mario turns Claude Code into a brand-aware marketing content engine. It builds deep brand foundations once, then uses them as persistent context for every piece of content you create — blog posts, landing pages, emails, social, ads. Specialized executor agents handle each marketing domain with frameworks and best practices baked in.

Without Mario, Claude Code sessions are stateless. There's no memory between sessions, no brand context, no competitive awareness, and no way to keep content consistent across channels.

With Mario, you get:

• Brand foundations — 8 researched documents that capture your identity, voice, audience, positioning, and competitive landscape
• Persistent memory — brand context loads automatically into every content session
• Competitive research — topic researchers analyze competing content before you write
• Marketing-aware agents — strategy, web, email, social, SEO, and ads executors load domain-specific guides
• Consistent voice — every piece of content draws from the same brand bible
• On-demand content — one command to create any marketing asset, informed by your foundations
• Website audits — scored marketing audits across 6 dimensions with prioritized recommendations
• Competitor analysis — side-by-side comparison against your brand positioning

/mario:new-project → /mario:create → repeat

1. /mario:new-project builds your brand foundations through deep questioning and parallel research agents
2. /mario:create produces content using those foundations as context — blog posts, landing pages, emails, anything
3. Foundations persist across sessions, so every future /mario:create starts brand-aware

gem install marketing_mario

Install commands and agents globally (recommended):

mario install --global    # Installs to ~/.claude/ — available in all projects

Or locally for a single project:

mario install --local     # Installs to ./.claude/ — project-specific

/mario:new-project                                          # Establish brand foundations
/clear
/mario:create "Write a blog post about customer onboarding" # Create content
/clear
/mario:create "Landing page for our enterprise plan"        # Create more content
/clear
/mario:create "Email welcome sequence for new signups"      # Keep going

Use /clear between commands to give each orchestrator a fresh context window. Each command loads only the context it needs.

/mario:new-project takes you from idea to content-ready in one command:

• Deep questioning to understand your brand and market
• Spawns 7 parallel foundation researcher agents
• Produces comprehensive brand foundation documents
• Synthesizes everything into BRAND-BIBLE.md

The foundations created in .mario_planning/foundations/:

Pass --auto to skip interactive questioning and let researchers work from your initial input.

/mario:create is the primary workflow:

1. Loads BRAND-BIBLE.md and PROJECT.md as context
2. Spawns a topic researcher to analyze competing content
3. Presents a research brief with recommended angle for your approval
4. Generates content aligned with brand voice and positioning
5. Conversational iteration until you're happy

/mario:create "Write a blog post about NPS surveys"
/mario:create "Landing page for our new feature"
/mario:create "Email nurture sequence for trial users"

Save successful content structures as reusable templates:

/mario:save-template 003 email-sequence    # Save a plan's structure as template
/mario:new-from-template email-sequence    # Create a new plan from it

Templates live in .mario_planning/templates/ and capture the task structure without the specific content, so you can replicate what works.

/mario:quick    # Same guarantees, skips optional agents

For small, ad-hoc tasks that don't warrant the full create workflow. Quick tasks live in .mario_planning/quick/, get atomic commits, and update state — but skip research and checkpoints.

/mario:debug "form submission fails silently"    # Start debug session
/clear
/mario:debug                                      # Resume from where you left off

Uses the scientific method: observe, hypothesize, test, conclude. Debug state persists in .mario_planning/debug/ and survives /clear, so you can continue across sessions.

/mario:audit https://example.com        # Full scored audit (5 parallel agents)
/mario:quick-audit https://example.com  # 60-second snapshot

/mario:audit spawns 5 parallel auditor agents scoring Content & Messaging, Conversion Optimization, SEO & Discoverability, Competitive Positioning, and Brand & Trust. A synthesizer then combines results and derives a 6th Growth & Strategy dimension. Produces a scored report (0-100) with prioritized recommendations in .mario_planning/audits/.

/mario:quick-audit runs a single-pass analysis — no subagents, just a condensed scorecard in about 60 seconds. Option to save or discard.

/mario:competitors https://rival1.com https://rival2.com

Analyzes competitor websites against your brand foundations. Requires /mario:new-project first. Spawns parallel profiler agents for each competitor, then synthesizes a comparison with positioning maps, messaging comparison, content gaps, and strategic opportunities.

Lightweight coordinators that spawn specialized agents. They stay lean (~10-15% context usage), passing file paths to subagents rather than content. Each subagent gets a fresh 200k context window.

Plans include a domain field in their frontmatter. The orchestrator routes each plan to the appropriate marketing-aware executor:

Each specialized executor loads its domain guide automatically, applying domain-specific frameworks and best practices.

Guides encode marketing frameworks, best practices, and domain knowledge that executors follow during plan execution.

Guides are installed to ~/.claude/guides/ (global) or .claude/guides/ (local). Edit them to match your project's conventions. After updating Mario, use /mario:reapply-patches to restore your customizations.

.mario_planning/
├── PROJECT.md            # Project vision and initial notes
├── BACKLOG.md            # Ordered list of plans to execute
├── STATE.md              # Project memory across sessions
├── config.json           # Workflow mode and settings
├── foundations/           # Brand foundation documents
│   ├── BRAND-IDENTITY.md
│   ├── VOICE-TONE.md
│   ├── AUDIENCE-PERSONAS.md
│   ├── COMPETITIVE-LANDSCAPE.md
│   ├── MESSAGING-FRAMEWORK.md
│   ├── PRODUCT-SERVICE.md
│   ├── CHANNELS-DISTRIBUTION.md
│   └── BRAND-BIBLE.md   # Synthesized quick-reference
├── plans/                # Execution plans and summaries
│   ├── 001-foundation/
│   │   ├── PLAN.md
│   │   └── SUMMARY.md
│   └── 002-core-features/
│       ├── PLAN.md
│       └── SUMMARY.md
├── audits/               # Website audit reports
│   ├── {domain-slug}/
│   │   ├── AUDIT-REPORT.md
│   │   ├── QUICK-AUDIT.md
│   │   └── dimension-*.md
│   └── competitive-analysis/
│       ├── COMPETITORS.md
│       └── profile-*.md
├── templates/            # Reusable plan templates
├── quick/                # Quick task plans and summaries
├── todos/
│   ├── pending/          # Todos waiting to be worked on
│   └── done/             # Completed todos
└── debug/
    └── resolved/         # Archived resolved issues

Control which Claude models agents use via /mario:settings:

Audit agents (auditors, synthesizer, competitor profilers) follow the same profile system — mapped to the Research column.

Configure via /mario:settings:

• Research toggle — enable/disable competitive research during content creation
• Model profile — select quality, balanced, or budget

Set during /mario:new-project:

Interactive Mode — confirms each major decision, pauses at checkpoints for approval, more guidance throughout.

YOLO Mode — auto-approves most decisions, executes without confirmation, only stops for critical checkpoints.

Change anytime by editing .mario_planning/config.json.

Starting a new project:

/mario:new-project        # Establish brand foundations
/clear
/mario:create "Blog post about customer onboarding best practices"

Creating more content:

/mario:create "Email welcome sequence for new signups"
/clear
/mario:create "Landing page for our enterprise plan"

Resuming work after a break:

/mario:progress  # See where you left off and continue

Capturing ideas during work:

/mario:add-todo                    # Capture from conversation context
/mario:add-todo Fix modal z-index  # Capture with explicit description
/mario:check-todos                 # Review and work on todos

Auditing a website:

/mario:quick-audit https://example.com          # Fast snapshot
/mario:audit https://example.com                # Deep analysis

Analyzing competitors:

/mario:competitors https://rival1.com https://rival2.com

Debugging an issue:

/mario:debug "form submission fails silently"  # Start debug session
/clear
/mario:debug                                    # Resume from where you left off

/mario:update              # Shows changelog, confirms before installing

Local modifications to guides and templates are backed up automatically during updates. After updating, use /mario:reapply-patches to restore your customizations.

• Ruby >= 3.1.0
• Claude Code

• Mario Alvarez
• Jorge Alvarez

• GSD for inspiring the project and providing a solid foundation to build upon.
• Corey Heines for his amazing collection of marketing skills.
• Zubair Trabzada great skills for audits.
• TalentoHQ for testing early versions and providing invaluable feedback to shape Mario into a practical tool for marketers.

MIT