financial-computing-numpy

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external CSV data which represents a potential attack surface for indirect prompt injection.
  • Ingestion points: scripts/financial_computing_numpy_diagnostics.py reads data from a user-provided CSV file using pd.read_csv.
  • Boundary markers: No explicit boundary markers or instructions to ignore embedded content are used in the prompt logic relative to this data.
  • Capability inventory: The script is limited to reading CSV files, performing numerical calculations (mean, quantiles, max drawdown), and writing a JSON summary to disk. It does not perform network operations or shell command execution.
  • Sanitization: The script uses pd.to_numeric and pd.to_datetime with errors='coerce', which effectively filters out non-numeric or invalid date strings from the statistical calculations.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 08:45 AM
Security Audit — agent-trust-hub — financial-computing-numpy