backtesting-bias-avoidance
Pass
Audited by Gen Agent Trust Hub on Aug 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill follows established quant research best practices, implementing rigorous statistical tests (Newey-West HAC, CSCV/PBO, Deflated Sharpe Ratio) to detect backtesting biases. The core logic is contained in
scripts/run_backtest.py, which performs mathematical calculations and generates Markdown reports and PNG charts based on either synthetic or publicly sourced financial data. - [SAFE]: All dependencies listed in the instructions and scripts (
numpy,pandas,scipy,matplotlib,yfinance) are well-known, industry-standard libraries for data science and finance. No suspicious or unversioned third-party packages are used. - [SAFE]: The skill does not attempt to access sensitive system files (e.g., SSH keys, credentials, environment variables) or execute arbitrary commands. Its file system access is restricted to reading its own reference files and writing the requested audit reports.
- [SAFE]: No obfuscation, prompt injection attempts, or hidden persistence mechanisms were found. The code is written in a clear, modular, and well-documented style.
Audit Metadata