magic-statistical-analysis

Pass

Audited by Gen Agent Trust Hub on Jul 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements deterministic statistical computations using standard libraries (pandas, numpy, scipy). It explicitly directs the agent to use code for calculations, avoiding LLM-based arithmetic errors.
  • [SAFE]: Automated test selection logic (e.g., using Shapiro-Wilk for normality testing before choosing between parametric tests like t-tests or non-parametric alternatives like Mann-Whitney U) demonstrates a high standard of statistical rigor and prevents common misapplications of statistical tests.
  • [SAFE]: Command execution identified in the test suite (tests/test_validate_statistics.py) is used exclusively for functional testing of the local script using sys.executable, which is a standard and safe development practice.
  • [SAFE]: No suspicious dependencies, remote code execution patterns, or obfuscated content were found. Data access is restricted to user-provided input files for analysis purposes.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 3, 2026, 07:00 AM
Security Audit — agent-trust-hub — magic-statistical-analysis