xlsx
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill originates from a trusted repository (anthropics/skills) and utilizes well-known, industry-standard Python packages for its primary spreadsheet processing tasks.
- [PROMPT_INJECTION]: The internal instructions (e.g., 'ALWAYS use formulas', 'Critical: Use Formulas') are design constraints aimed at maintaining spreadsheet integrity and data relationship logic, rather than attempts to bypass agent safety or security guidelines.
- [DATA_EXFILTRATION]: File access patterns are localized to reading and writing spreadsheet files (xlsx, csv, etc.) as requested by the user, with no evidence of unauthorized network transmission or access to sensitive system credentials.
- [REMOTE_CODE_EXECUTION]: No untrusted remote code execution or dynamic execution patterns (such as eval or shell piping) were found; all operations use static library imports.
Audit Metadata