linkedin-automation

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly calls LINKEDIN_GET_MY_INFO and LINKEDIN_GET_COMPANY_INFO (see Core Workflows and ID Resolution) to fetch LinkedIn profile and company page data—user-generated, public third-party content—which the agent must parse (extract URNs) and use to decide subsequent API actions, so untrusted content can materially influence behavior.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill requires connecting at runtime to the MCP endpoint https://rube.app/mcp and calling RUBE_SEARCH_TOOLS to fetch tool schemas that directly determine and control the agent's instructions/behavior.