prompt-injection-probe

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

• Third-party content exposure detected (high risk: 0.85). The skill’s required runtime workflow explicitly tests indirect injection by having the user upload documents (PI-IND-001) or host pages/URLs (PI-IND-002) whose free-form text is then ingested by the agent/LLM context, which is outsider-authored content delivered via document/URL fetch paths.