build-evidence-map

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

• Third-party content exposure detected (low risk: 0.10). The skill’s workflow does not read outsider-authored free text unless the user explicitly runs doubt verify on a user-provided decision.doubt.json, and even then the LLM ingests only the recorded first-party source excerpts/URLs rather than any monitored queue/feed or arbitrary text stream.