model_drift_detection
Pass
Audited by Gen Agent Trust Hub on Mar 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill uses industry-standard Python libraries (evidently, scipy) for model monitoring and statistical testing, with no malicious patterns detected.
- [DATA_EXFILTRATION]: No evidence of unauthorized data transfer exists. The skill explicitly directs the user to anonymize metrics and exclude biometric embeddings to ensure compliance with GDPR and other data protection regulations.
- [PROMPT_INJECTION]: The skill monitors production inference scores which technically constitutes an external data ingestion point; however, this data is used for statistical calculations (Kolmogorov-Smirnov tests) and does not present a credible surface for indirect prompt injection.
Audit Metadata