plance

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 run membrane action run and membrane request (e.g., "membrane action run --connectionId=CONNECTION_ID ACTION_ID" and "membrane request CONNECTION_ID /path/to/endpoint"), which fetches and ingests data from a third-party Plance workspace (tasks, attachments, user-generated content) that the agent will read and could materially influence subsequent actions.