browser-use-terminal

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

• Third-party content exposure detected (high risk: 0.85). Runtime path: the skill drives a real browser to fetch and extract arbitrary public web page text (e.g., new_tab("https://docs.browser-use.com")), and that fetched page content is then ingested into the agent’s LLM context as readable browser/page text—an outsider-authored source.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's setup instructions explicitly tell the user/agent to run "curl -fsSL https://browser-use.com/terminal/install.sh | sh", which fetches and executes remote code at runtime and is presented as a required install step, so the URL https://browser-use.com/terminal/install.sh is a high-confidence runtime dependency that executes remote code.