NYC

file-organizer

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
  • [NO_CODE] (SAFE): The provided file contains only metadata. No executable instructions or skill logic were provided for analysis.
  • [INDIRECT_PROMPT_INJECTION] (LOW): The skill processes file system data to 'understand context', creating a surface for malicious content within files to influence agent behavior. 1. Ingestion points: Local file content and metadata. 2. Boundary markers: Not provided in the metadata. 3. Capability inventory: File management operations (read, move, delete) are implied by the skill's purpose. 4. Sanitization: No sanitization mechanisms are described in the metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:17 PM