earnings-event-study
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill performs quantitative analysis on user-supplied CSV files. All data processing is done locally using standard libraries (pandas, numpy) without any external network operations.
- [SAFE]: The core logic in
scripts/event_study.pyimplements statistical tests (t-test, sign test) from first principles to avoid unnecessary external dependencies and maintains a minimal, verifiable attack surface. - [SAFE]: Security best practices are followed in the report generation phase, specifically utilizing an
_html_escapefunction to sanitize user-supplied labels (like symbols or event types) before rendering them into the HTML report. - [SAFE]: The skill incorporates robust data guardrails, such as dropping events with excessive missing values (over 30% in window) and requiring a minimum sample size (N=20) for statistical validity, preventing the agent from reporting on noisy or insufficient data.
- [SAFE]: Instructions and metadata are consistent with the stated research-oriented purpose. The skill explicitly includes guardrails and disclaimers to prevent the agent from providing unauthorized financial or trading advice.
Audit Metadata