open-source

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

• Third-party content exposure detected (high risk: 1.00). The skill's docs and workflow explicitly show the Agent navigating to and extracting from arbitrary public webpages (e.g., references/actor.md page.goto("https://github.com/login"), references/agent.md prompt examples like "Go to https://quotes.toscrape.com/" and search tasks, and references/integrations.md listing TikTok/Instagram scrapers), meaning the agent will fetch and interpret untrusted, user-generated third-party content which can directly influence its actions.

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

• Potentially malicious external URL detected (high risk: 1.00). The docs include a curl piped-to-shell command that fetches and executes remote code at runtime for profile sync (curl -fsSL https://browser-use.com/profile.sh | sh), which is a direct remote-execution dependency used for sandbox deployment.