maestro-ralph-cli

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

• Third-party content exposure detected (medium risk: 0.65). Outsider free text can enter the LLM context via A_LOAD_STEP_CONTEXT reading artifacts produced by prior delegated steps (e.g., grill-report.md, review.json, completion_evidence, etc.) and then injecting that content into A_COMPOSE_DELEGATION_PROMPT for the next maestro delegate; those artifacts are generated from delegate outputs and are not guaranteed to be authored by the operating user.