social-engine

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

• Third-party content exposure detected (medium risk: 0.30). The workflow reads untrusted external comments, messages, posts, and support chats from LinkedIn, X, Discord, YouTube, and Crisp.

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

• Potentially malicious external URL detected (high risk: 0.70). The instruction to sign up at https://cogny.com and configure an MCP server at https://app.cogny.com/mcp constitutes an operational inbound dependency on a remote third-party service, but it carries weak/unknown provenance.