competitor-research

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

• Third-party content exposure detected (high risk: 0.90). Flagged because the SKILL.md and FOLLOWER_OUTREACH_PIPELINE_SPEC.md explicitly describe scraping public, untrusted third‑party content (e.g., the Apify Meta Ads Library scraper "curious_coder/facebook-ads-library-scraper" referenced in apps/worker/src/services/meta-ads-scraper.ts and the IG/FB follower scrapers like iron-crawler/instagram-followers-scraper-no-login-needed and apify/facebook-followers-following-scraper) and then have the agent ingest and AI‑analyze that content (Haiku/Sonnet passes and Claude-based follower research) to produce replication briefs, scoring, and outreach decisions—behavior that can be influenced by injected instructions in the third‑party content.