The Agent Skills Directory

[SAFE]: The skill implements a legitimate workflow for asset management, focusing on mandate compliance and diagnostic reporting.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external data from input.csv for diagnostic processing.
Ingestion points: The file scripts/asset_management_diagnostics.py reads data via pd.read_csv.
Boundary markers: Data is parsed as a structured CSV; no explicit natural language delimiters are used to separate data from instructions.
Capability inventory: The script performs file reading and writing of JSON artifacts; it does not possess shell execution or network capabilities.
Sanitization: Input columns are strictly coerced to numeric and datetime types using pd.to_numeric and pd.to_datetime, preventing execution or interpretation of embedded instructions.

asset-management