spreadsheet-analysis
Pass
Audited by Gen Agent Trust Hub on Aug 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill includes a dedicated Python utility (
scripts/profile_table.py) designed for safe, read-only inspection of spreadsheet data. The script utilizes resource limits for file size, XML parts, and CSV records to protect against resource exhaustion and zip-bomb attacks.\n- [DATA_EXFILTRATION]: Analysis is designed to be performed locally. The skill instructions mandate that sensitive data remain local unless explicitly authorized, and no network exfiltration patterns were detected in the scripts.\n- [COMMAND_EXECUTION]: Command patterns are limited to executing the local profiling script. The script handles file paths safely using path resolution to prevent traversal or aliasing during report generation.\n- [PROMPT_INJECTION]: The skill instructions proactively address injection risks by identifying formula-like text patterns that could be used for spreadsheet-based injection attacks. No malicious instructions were found in the metadata or prompts.
Audit Metadata