med-international

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

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime workflow ingests user-supplied international material (e.g., UN/government statements, Reuters articles, translated documents, diaspora NGO text, and on-the-ground notes) into the agent’s LLM context; these are outsider-authored free-form texts when they come from non-user parties (e.g., Reuters reporting, NGO statements, translated foreign government text, correspondent notes).

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)