5 subagents (used by /market audit in parallel)

• market-content — content & messaging specialist
• market-conversion — CRO specialist
• market-technical — technical SEO specialist
• market-competitive — competitive intelligence
• market-strategy — brand, trust, and growth

4 scripts

• analyze_page.py — HTML / SEO / schema / tracking extraction (stdlib only)
• competitor_scanner.py — multi-URL positioning scan
• social_calendar.py — pillar-based content calendar generator
• generate_pdf_report.py — client-ready PDF via reportlab

6 templates

Welcome, nurture, and launch email sequences; content calendar; launch checklist; client proposal.

Install

curl -fsSL https://raw.githubusercontent.com/rediumvex/ai-marketing-claude/main/install.sh | bash

Or clone and run:

git clone https://github.com/rediumvex/ai-marketing-claude.git
cd ai-marketing-claude
./install.sh

This copies the skills, agents, scripts, and templates into ~/.claude/ so Claude Code can discover them.

PDF reports (optional)

pip3 install reportlab

Only needed if you use /market report --format pdf.

Quick start

# Full audit with all 5 specialists
/market audit https://example.com

# Just CRO
/market landing https://example.com/pricing

# Generate an email sequence in the detected brand voice
/market brand https://example.com      # run this first
/market emails welcome                  # matches the voice from BRAND-VOICE.md

# Final client deliverable
/market report --format pdf https://example.com

How it works

1. Every skill is a Markdown file with YAML frontmatter that Claude Code reads to decide when to invoke it.
2. The root market/SKILL.md is the router — it catches /market commands and hands off to the right specialist.
3. The audit skill launches 5 subagents in parallel. Each scores its dimension and returns specific findings with quoted evidence.
4. All artifacts land in the current project root (MARKETING-AUDIT.md, BRAND-VOICE.md, etc.) and downstream skills read them for context.
5. The PDF report compiles everything into a client deliverable.

Scoring methodology

Weighted 6-category composite (0–100):

• Content & Messaging — 25%
• Conversion Optimization — 20%
• SEO & Discoverability — 20%
• Competitive Positioning — 15%
• Brand & Trust — 10%
• Growth & Strategy — 10%

Grades: 90+ A, 80+ B, 70+ C, 60+ D, <60 F.

Design principles

• Evidence over opinion. Every finding quotes actual page copy.
• Impact over activity. Fixes are ranked by traffic × expected CR lift × ARPU.
• Specific over generic. Every recommendation includes a concrete rewrite.
• Lean skills. Every skill fits under ~300 lines — no filler, no duplication.
• Stdlib first. Scripts run without dependencies except the optional PDF generator.

License

MIT. Use it, modify it, ship it.