maskdb

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

• Third-party content exposure detected (high risk: 0.75). Outsider free text can enter the agent’s LLM context via runtime ingestion of maskdb API responses (e.g., rows/page/masked fields returned from POST /v1/databases/<database>/query and metadata from /schema), where the returned database content is authored by third parties (the database’s data owners), not by the operating user.