data-analysis

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill represents standard configuration metadata and safe user guidelines for analytical operations with no hazardous elements.
  • [PROMPT_INJECTION]: No behavioral overrides, jailbreaks, or instructions attempting to bypass security constraints were found.
  • [DATA_EXFILTRATION]: There are no indicators of sensitive file system access, embedded credentials, or external network transmissions.
  • [REMOTE_CODE_EXECUTION]: No remote execution commands, shell script piping, or package installations are declared.
  • [OBFUSCATION]: Content is completely transparent with no multi-layer encoding, zero-width spaces, or homoglyph substitutions.
  • [INDIRECT_PROMPT_INJECTION]: Although the skill accepts user-supplied spreadsheet formats (.xlsx, .csv, .xls), it lacks dangerous capabilities or outbound pathways that could be weaponized by compromised data assets.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 05:53 AM
Security Audit — agent-trust-hub — data-analysis