Ad Creative Intelligence

Published by marmikcfc in pepper-skills

Review recommended2 installs

What this skill does

Research what ads competitors are running across Meta and Google — their hooks, claims, creative angles, and targeting signals. Use when asked to research competitor ads, find ad inspiration, or analyze what's working in your market.

Add Ad Creative Intelligence to your agent

Review the source and files first. When you are ready, copy the prompt instruction or use the CLI command supported by your environment.

Install with a prompt

Paste this into a compatible coding agent:

add this skill "ad-creative-intelligence" from https://github.com/marmikcfc/pepper-skills

Install with the CLI

Run this command in a controlled environment after reviewing the repository:

npx skills add https://github.com/marmikcfc/pepper-skills --skill ad-creative-intelligence

Skill instructions

Ad Creative Intelligence

Research what competitors are running in paid advertising — hooks, claims, creative angles, and messaging themes from Meta Ads Library and Google Ads Transparency.

When to Use

  • "What ads is [competitor] running?"
  • "Research competitor ad creatives"
  • "Find ad inspiration in [industry]"
  • "What messaging is working in paid for [market]?"

Prerequisites

  • ORTHOGONAL_API_KEY
  • ANTHROPIC_API_KEY

Workflow

Step 1: Meta Ads Library

orth run scrape "https://www.facebook.com/ads/library/?active_status=active&ad_type=all&q=<competitor_name>" \
  --body '{"format": "text"}'

Step 2: Google Ads Transparency Center

orth run scrape "https://adstransparency.google.com/?region=anywhere&q=<competitor_name>" \
  --body '{"format": "text"}'

Step 3: Validate scrape results Meta Ads Library and Google Ads Transparency are JavaScript-rendered pages. Before proceeding, check that the scraped content contains actual ad text (minimum 500 characters with recognizable ad language — headlines, "Learn More", button text, etc.). If either scrape returned less than 500 characters or only boilerplate HTML, warn the user:

"The [Meta / Google] ad library scrape returned minimal content — the page may require JavaScript or a logged-in session. I'll proceed with whatever was captured, but results may be limited."

Step 4: Extract ad patterns with LLM Pass all scraped ad text to Claude:

"Analyze these ads and extract:

  1. Top 5 headline hooks (how they grab attention)
  2. Core claims and promises (what outcomes do they promise?)
  3. Creative angles (emotional vs. rational, fear vs. aspiration, problem vs. solution)
  4. Target audience signals (who are these ads designed for?)
  5. Unique differentiators they're emphasizing
  6. Offers they're promoting (free trial, demo, discount, etc.)
  7. What appears to be their highest-volume/most-tested creative Format as a structured ad intelligence report."

Step 4: Present results Display ad intelligence report with patterns and examples.

Output

Structured report of competitor ad hooks, claims, creative angles, and offer patterns.

Files included

  • SKILL.md

More skills

Ad Creative Intelligence: Install, Source and Security | FunnelSlayer