performance-marketing-analysis

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly tells the agent to ask the user for APIFY_TOKEN and ELEVENLABS_API_KEY and to use them to run API calls/scripts (and to accept them if not in env vars), which requires the LLM to receive and potentially embed secret values verbatim—an exfiltration risk.

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly instructs the agent to use Apify (calling the actor curious_coder/facebook-ads-library-scraper) to fetch competitor ads from Facebook/Instagram and to use those scraped, third‑party ad contents to inform campaign strategy and generate copy, which is untrusted public content the agent will read and act on.