PreTestAds - Ad Creative Scoring MCP for AI Agents

Pre-flight ad creative testing through MCP. Before spending budget on ads, run your creative through PreTestAds' AI scoring engine. Benchmarked against 76 of the best-performing TikTok ads. Pay $5 USDC per run via x402 protocol, or use account credits through OAuth.

What It Does for Operators

• Creative scoring - AI evaluates video/image ads for predicted performance before you spend
• TikTok benchmarked - Scoring model trained on 76 top-performing TikTok ads
• x402 payment - Pay $5 USDC per analysis using the x402 payment protocol (no subscription needed)
• OAuth option - Account credits for frequent users
• Agent-native - Score ads directly from your AI agent during creative development

Installation

git clone https://github.com/krecicki/pretestads-agent-scripts.git
cd pretestads-agent-scripts
# Follow repo instructions for setup
# Or use the remote OAuth endpoint

Claude Desktop / Hermes Config

{
  "mcpServers": {
    "pretestads": {
      "command": "node",
      "args": ["path/to/pretestads/server.js"]
    }
  }
}

Key Tools

Tool	Description
score_video_ad	Score a video creative against benchmark data
score_image_ad	Score a static image creative
get_benchmark	Retrieve benchmark performance data
compare_ads	Side-by-side comparison of multiple creatives
get_recommendations	AI-generated improvement suggestions

Note: Tool names are approximate. See github.com/krecicki/pretestads-agent-scripts.

Operator Use Cases

1. DTC Brands - Score 10 video variations before spending $5K on the winner
2. Agencies - Pre-test client creative in minutes instead of running $500 test campaigns
3. Performance Marketers - Iterate creative with AI feedback before launch
4. Content Creators - Validate hooks and CTAs against TikTok benchmarks

Integration with CorpusIQ

PreTestAds + CorpusIQ = closed-loop ad optimization for operators. Use PreTestAds to score creative before launch → run ads on Meta/TikTok → track ROAS through CorpusIQ's ad platform connectors → feed performance data back into creative scoring. The creative-to-revenue feedback loop, all through AI agents.

Limitations

• Scoring benchmarked against TikTok ads only - may not translate to Meta, YouTube, etc.
• $5/run pricing adds up for high-volume testing
• Scoring is predictive, not guaranteed - real performance may vary
• x402 payment protocol adds friction for first-time users