mezmo

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

• Third-party content exposure detected (high risk: 0.70). The SKILL.md explicitly instructs using Membrane commands like "membrane action run" and "membrane request CONNECTION_ID /path/to/endpoint" to fetch data from the Mezmo API (logs/alerts/dashboards), which are third‑party, user‑generated/untrusted content that the agent is expected to read and that could materially influence subsequent actions.