bdi-mental-states

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s required workflow is to transform “external RDF context” into BDI mental states, and that RDF context is explicitly described as coming from outside the agent (e.g., “Push notification” / “Official announcement received via email”), which would be outsider-authored free text that the LLM ingests as RDF-derived prose during the triples-to-beliefs-to-triples transformation.