The Agent Skills Directory

[SAFE]: The skill effectively delivers its stated functionality as a statistical analysis toolkit without any malicious intent or hidden behaviors.
[COMMAND_EXECUTION]: All included scripts and code examples use well-known, standard Python libraries for data science such as scipy, statsmodels, and pingouin. There is no evidence of arbitrary command execution, unauthorized subprocess calls, or privilege escalation attempts.
[DATA_EXFILTRATION]: There are no network operations or attempts to read sensitive system configuration files. The skill operates locally on provided datasets for the purpose of statistical computation.
[PROMPT_INJECTION]: The documentation and skill instructions are purely instructional and academic. No bypass markers, role-play injections, or instructions to ignore safety protocols were found.
[EXTERNAL_DOWNLOADS]: The skill does not perform any runtime downloads or remote script execution. All dependencies mentioned are standard packages from official registries.
[OBFUSCATION]: The code in scripts/assumption_checks.py and the markdown files are written in plain text with no use of Base64, zero-width characters, or other encoding techniques to hide malicious payloads.

statistical-analysis