reputation-lyncs

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly instructs the agent to use Membrane action runs and the Membrane proxy (e.g., membrane request CONNECTION_ID /path/to/endpoint) to fetch data from the Reputation Lyncs API—which aggregates public, user-generated reviews and social mentions—so the agent will ingest untrusted third-party content that could influence subsequent actions.