The Agent Skills Directory

[SAFE]: The skill is a collection of analytical templates and utility scripts for business sizing. The Python scripts (scripts/revenue_impact.py, scripts/cost_savings.py, scripts/confidence_interval.py) only use standard libraries and perform basic mathematical calculations. There is no evidence of network access, sensitive file reads, or execution of untrusted code.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied metrics which could technically serve as an injection surface, though the implementation is safe.
Ingestion points: Metric values like baseline volume and lift percentages are provided as inputs to the scripts and templates.
Boundary markers: While explicit delimiters for text inputs are not used, the scripts expect specific numeric types.
Capability inventory: The skill's functionality is restricted to mathematical processing; it contains no capabilities for network exfiltration, persistent file modification, or subprocess execution.
Sanitization: Inputs to the calculation scripts are strictly cast to floating-point numbers via argparse, which prevents natural language instructions from being processed as executable logic.

impact-quantification