The Agent Skills Directory

[SAFE]: The skill is designed for data validation and integrity checking. It follows best practices for financial data auditing and does not exhibit any suspicious behaviors or unauthorized access patterns.\n- [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it processes external CSV data. However, the logic is limited to numerical and structural validation using the pandas library, which does not interpret data as natural language instructions. Ingestion points: User-provided CSV file path. Boundary markers: No explicit markers in the template. Capability inventory: pandas data frame operations; no subprocess calls or network requests. Sanitization: The audit logic validates data types and numerical ranges.

data-scrub