remote-browser

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill includes examples and commands that require embedding sensitive values (e.g., a literal "password" in a chained command and cookie values passed as CLI arguments), which forces the LLM to output secrets verbatim.

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

• Third-party content exposure detected (medium risk: 0.30). In the remote-browser skill’s runtime workflow, the agent can navigate to an outsider-specified URL (browser-use open <url>), then ingest page-derived free text via browser-use state/screenshot/get html/get text, meaning outsider-authored content can be read without selecting a specific item first.