The Agent Skills Directory

[SAFE]: The skill implements statistical distributions and tests (Welch's t, Chi-square, Mann-Whitney U, etc.) using only the Python standard library. This avoids third-party dependency risks and keeps the execution environment isolated.
[SAFE]: No network operations (curl, wget, requests) or data exfiltration patterns were detected in any of the scripts. Data processing is restricted to local JSON input files.
[SAFE]: Analysis of all scripts and instructions revealed no signs of obfuscation, malicious command execution, or persistence mechanisms.
[SAFE]: The skill includes self-test functionality in stats_core.py and test_impl.py to verify mathematical accuracy against published reference values, which is a security best practice for transparent implementation.
[SAFE]: The instruction set in SKILL.md is focused on methodological guidance and decision-making logic without any attempts to bypass safety filters or override agent constraints.

statistical-analyst