tooluniverse-statistical-modeling
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill uses instructional headers like 'CRITICAL' and 'IMPORTANT' to emphasize data validation steps (e.g., verifying outcome variables before analysis). These are used correctly for logic flow and do not attempt to bypass agent safety filters or override core instructions.
- [DATA_EXFILTRATION]: All data operations involve reading local CSV files for analysis. No patterns were found indicating the transmission of sensitive data to external servers or access to unauthorized system paths.
- [EXTERNAL_DOWNLOADS]: The skill relies on well-known, established scientific libraries (statsmodels, scikit-learn, lifelines, pandas, numpy). No arbitrary remote code execution or downloads from untrusted sources were identified.
- [COMMAND_EXECUTION]: The skill contains Python code patterns for statistical modeling. These operations are restricted to computational analysis within the agent's environment and do not involve shell command execution or system-level changes.
Audit Metadata