boost

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

• Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md instructs the agent to connect to Boost via the Membrane CLI (e.g., "membrane connect --connectorKey boost") and to run actions that return Boost data (e.g., "membrane action run ..."), which are social-media/user-generated third‑party contents that the agent will ingest and could influence subsequent actions.