The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external CSV files, creating an indirect prompt injection surface.\n
Ingestion points: The script scripts/clearinghouse_posttrade_diagnostics.py reads data using pd.read_csv.\n
Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore embedded instructions in the script output.\n
Capability inventory: The skill has the capability to write diagnostic results to a local file via the --output flag.\n
Sanitization: The script employs pd.to_numeric and pd.to_datetime with errors='coerce', which sanitizes the input by ensuring only numerical and date data are included in the summary passed to the agent.\n- [EXTERNAL_DOWNLOADS]: The script scripts/clearinghouse_posttrade_diagnostics.py depends on the pandas library, which is a widely used and trusted package for data analysis.

clearinghouse-posttrade