← Back to docs

Recipe: Brand safety check on ad copy

Scan ad copy against your brand safety rules before it goes live.

Overview

This recipe runs every ad variant through a configurable rule engine that flags restricted terms, competitor names, and compliance risks. Results land in a Slack channel and a CSV audit log.

Prerequisites

  • Meridian API key with write:audit scope
  • A rule set JSON file uploaded to your workspace
  • Slack webhook URL for notifications

Step 1 — Define your rule set

Create a JSON file with blocked terms, competitor names, and regex patterns. Store it at the workspace level so every campaign inherits the same guardrails.

{
  "blocked_terms": ["guaranteed", "risk-free"],
  "competitors": ["RivalAd", "ClickBoost"],
  "regex_patterns": ["\\b\\d{3}%\\b"]
}

Step 2 — Run the scan

POST your ad copy variants to the scan endpoint. The response includes a pass/fail verdict and a list of triggered rules.

curl -X POST https://api.getnimbus.net/v1/scan \
  -H "Authorization: Bearer $MERIDIAN_KEY" \
  -H "Content-Type: application/json" \
  -d '{"copy":["Best deal ever"], "ruleset_id":"rs_01"}'

Step 3 — Review results

Flagged variants appear in your dashboard audit log. The Slack integration posts a summary card so your compliance team can review without logging in.

Next steps