performance-attribution

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides a mathematical and analytical framework for financial performance attribution. It does not execute arbitrary code or interact with external services.
  • [SAFE]: Instructions correctly guide the agent to import tested implementations from a local library (src/quantlib/attribution.py) rather than implementing sensitive financial logic manually, which reduces the risk of calculation errors.
  • [DATA_EXPOSURE]: No sensitive data access or credential patterns detected. The skill uses public financial benchmarks (e.g., CSI 300, S&P 500) as examples.
  • [REMOTE_CODE_EXECUTION]: No remote script downloads or dynamic execution patterns found. All logic is described through mathematical formulas and references to local source files.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 02:29 PM
Security Audit — agent-trust-hub — performance-attribution