mcpField
C
creative-variants

creative-variants

github.com/gtmagentsMarketing & growth42864 installsUpdated 30w ago
Repository

About creative-variants

A skill for planning and testing multiple creative angles across marketing campaigns. It gives teams frameworks for hooks, channel formats, messaging, and tracking variants.

Key features

  1. 01Hook library for creative angles
  2. 02Channel and format matrix
  3. 03Hook-to-CTA story formula
  4. 04Persona and funnel-based tone
  5. 05Variant and performance templates

Use cases

  1. 01Developing concepts for a new campaign
  2. 02Refreshing ads with declining performance
  3. 03Running structured creative tests

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/gtmagents/gtm-agents --skill creative-variants

After installing, type/creative-variantsin Claude Code to confirm it loaded.

Details

From the project README

Creative Variant Playbooks Skill

When to Use

  • Kicking off new campaigns needing multiple creative angles.
  • Refreshing fatigued ads to regain performance.
  • Running systematic creative testing programs.

Framework

  1. Hook Library – benefit, problem, proof, story, contrarian, curiosity.
  2. Format Matrix – channel x format (video, carousel, single image, text ad) with unique requirements.
  3. Story Formula – Hook → Value Proof → CTA; adapt for each channel.
  4. Voice & Tone – align with persona (executive vs practitioner) and funnel stage.

Checklist

  • ✅ Unique headline + body copy per variant.
  • ✅ Visuals sized and optimized per channel.
  • ✅ CTA and offer match landing page.
  • ✅ Accessibility: subtitles, alt text, high contrast.
  • ✅ Compliance/legal disclaimers added.

Templates

  • Variant tracker (concept, hook, asset link, status, performance snapshot).
  • Caption/copy bank with placeholders and emoji/hashtag guidance.
  • Video storyboard sheet.

Tips

  • Pair evergreen and timely hooks to keep pipelines fresh.
  • Recycle top-performing organic content into paid variations.
  • Archive learnings in a centralized asset library.

Comments

Markdown supported