financial-computing-cpp
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions and associated scripts are focused on performance diagnostics and do not exhibit malicious patterns. There is no evidence of obfuscation, hardcoded credentials, or unauthorized system access.
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits a standard vulnerability surface common to tools that process external data.
- Ingestion points: The
scripts/financial_computing_cpp_diagnostics.pyscript ingests untrusted data from user-supplied CSV files via thepd.read_csvfunction. - Boundary markers: No explicit prompt delimiters or "ignore instructions" warnings are present in the script's output, though its JSON structure provides data isolation.
- Capability inventory: The script's capabilities are limited to statistical calculations and writing results to a local file; it does not contain functions for network requests, shell command execution, or sensitive file access.
- Sanitization: The script enforces a specific column schema and uses strict numeric coercion for performance metrics, which mitigates the risk of processing malicious text instructions hidden in data fields.
Audit Metadata