The Agent Skills Directory

[SAFE]: The skill provides legitimate data analysis functionality using Z-score, IQR, and Moving Average methods. All logic is contained within a local Python script using only standard library modules.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external data from CSV files. However, the script strictly validates and converts input to numeric types (floats) before processing, which prevents text-based prompt injection from being processed or misinterpreted by the script's logic.
[COMMAND_EXECUTION]: While the skill is executed via the command line, it does not use dangerous functions like os.system or subprocess to run external commands. It only performs internal mathematical computations.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill reads from a user-specified CSV and writes to a user-specified output file or stdout. It contains no network code and does not attempt to access sensitive system files or credentials.

outlier-scan