ad-copy-verification-standard

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

• Third-party content exposure detected (low risk: 0.10). The agent scrapes arbitrary business websites via tools like Firecrawl or Jina Reader (low risk, as the agent fetches specific URLs rather than monitoring a public feed of unselected submissions).

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

• Potentially malicious external URL detected (medium risk: 0.30). The installation instructions include an explicit curl command to fetch remote skill content from GitHub (https://raw.githubusercontent.com/fourteenwm/ppc-ai-skills/...), establishing an inbound operational external dependency with common/reputable unofficial publisher provenance.