playbook-ad-library

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). Meta Ad Library pulls “Ad Samples” (ad titles/bodies) from Apify’s apify/facebook-ads-scraper for each verified Facebook Page, and those outsider-written ad texts are then passed into the LLM for drafting and truth-judging.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The skill calls the Apify actor at runtime (POST https://api.apify.com/v2/acts/apify~facebook-ads-scraper/run-sync-get-dataset-items), which executes remote scraper code and returns ad text that is then fed into the model prompt.