integry

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 the agent to run membrane action list/action run and use "membrane request CONNECTION_ID /path/to/endpoint" to fetch data from Integry-connected third‑party SaaS endpoints, which are untrusted/user-provided and the agent is expected to read and act on, so third‑party content can influence subsequent actions.