sop-application-writer

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

• Third-party content exposure detected (high risk: 0.75). The required workflow is to “ask for” and “review an existing draft” using applicant-provided story/CV text, which is outsider-authored free-form user/third-party content that is then ingested by the agent into the LLM context (no sanitization or restriction is described).