radiomics-ml

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill includes a deterministic auditing script, scripts/check_radiomics_ml.py, which is implemented using only Python standard libraries. It performs logic checks on a declarative JSON manifest and does not involve network operations, subprocess calls, or dynamic code execution.
  • [SAFE]: All external libraries referenced in the documentation (e.g., scikit-learn, xgboost, lightgbm, catboost, pyradiomics) are well-known, established tools in the machine learning community. The skill provides instructions for their use rather than executing them from untrusted sources.
  • [SAFE]: No evidence of prompt injection, data exfiltration, credential exposure, or obfuscation was found. The skill instructions and scripts are consistent with the stated purpose of improving clinical machine learning rigor.
  • [SAFE]: The skill uses a regression test suite (tests/test_radiomics_ml.sh) and a challenge card to ensure the auditing script functions correctly and predictably using local synthetic data.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 02:10 AM
Security Audit — agent-trust-hub — radiomics-ml