NYC

databases

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION] (SAFE): The skill utilizes sudo for routine administrative tasks such as package installation and service management, which is appropriate for the skill's primary purpose.
  • [EXTERNAL_DOWNLOADS] (SAFE): All external downloads are directed to trusted official operating system repositories via the apt-get package manager.
  • [PROMPT_INJECTION] (LOW): The skill presents an indirect prompt injection surface. 1. Ingestion points: The db_performance_check.py script analyzes external database query strings. 2. Boundary markers: None are identified in the documentation to prevent the execution of instructions embedded in queries. 3. Capability inventory: The skill has administrative database access and local script execution capabilities. 4. Sanitization: The documentation does not specify any sanitization or escaping of ingested query data.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:29 PM