unifiedto

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 the Membrane CLI commands (e.g., membrane action list, membrane action run, and membrane request CONNECTION_ID /path/to/endpoint) which fetch and ingest data from Unified.to-connected third-party SaaS/APIs (potentially user-generated or public content) that the agent is expected to read and could materially influence subsequent actions.