marketing-cro

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

• Third-party content exposure detected (high risk: 0.90). This skill's Research Mode (SKILL.md) explicitly instructs the agent to use agent-browser to open arbitrary live URLs — e.g., 'agent-browser --session cro-audit open "https://{page-url}"' and competitor pages/pagespeed.web.dev — so the agent will fetch and interpret untrusted public web content as part of its required workflow, which can materially influence its recommendations and actions.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill's Research Mode explicitly runs agent-browser at runtime to open and ingest arbitrary external pages (e.g., https://{page-url} and related URLs like https://pagespeed.web.dev/report?url=https://{page-url}), so remote page content would be fetched and injected into the agent context and could directly influence its prompts/recommendations.