hunt-a-handle

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

• Third-party content exposure detected (medium risk: 0.30). The required workflow “Hunt a handle” ingests outsider-authored free text at runtime by reading candidate profile pages/content from public web/social platforms for handle candidates (e.g., via the maigret path that “parses the page” and extracts fields, and via sherlock/WhatsMyName match-body checks), where the LLM is consuming page text authored by third parties.