threat-detection

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime (e.g., scripts/threat_signal_analyzer.py in ioc_mode/anomaly_mode) ingests outsider-authored free-text when the caller supplies --ioc-file or --events-file JSON content, which can be created/submitted by an outsider and is read without selecting a specific trusted item first.