ad-campaign-analyzer

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

• Third-party content exposure detected (high risk: 0.85). The required workflow includes accepting campaign data from “CSV export… / pasted performance table / screenshots of dashboard (we'll extract the data)” and explicitly treats CSV/pasted text/screenshots as untrusted input, so at runtime the agent ingests this outsider-authored free-form text (or extracted OCR/scraped text from it) into the LLM context for analysis.