The Agent Skills Directory

[SAFE]: The skill implements strong input validation. Specifically, scripts/sensitivity_summary.py uses regex (^[a-zA-Z_][a-zA-Z0-9_ .-]*$) and length constraints to sanitize parameter names, preventing potential injection or resource exhaustion attacks.
[SAFE]: All included scripts rely solely on the Python standard library. There are no external package dependencies or remote code downloads.
[SAFE]: The skill restricts itself to a limited set of tools (Read, Write, Grep, Glob) and explicitly excludes Bash, reducing the attack surface for command injection or unauthorized system operations.
[SAFE]: Numeric inputs across all scripts are validated to be finite numbers, and list lengths are capped at reasonable limits (e.g., 10,000 to 100,000 items) to prevent memory exhaustion or DoS vectors.

parameter-optimization