NYC

databases

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • Privilege Escalation (HIGH): In SKILL.md, the setup instructions for PostgreSQL include commands utilizing sudo (sudo apt-get install and sudo systemctl start). Directing an agent to execute commands with root privileges is a significant security risk as it bypasses permission constraints.
  • Indirect Prompt Injection (LOW): The skill is designed to ingest data from external MongoDB and PostgreSQL databases, which introduces a vulnerability surface.
  • Ingestion points: Data retrieved through database queries (db.users.find, SELECT * FROM users).
  • Boundary markers: Absent. The skill provides no mechanisms to isolate or delimit untrusted database content from the agent's internal instructions.
  • Capability inventory: The skill utilizes shell commands (psql, mongosh) and executes local Python scripts (scripts/db_migrate.py, scripts/db_backup.py).
  • Sanitization: Absent. No validation or escaping of external database content is mentioned or implemented in the provided documentation.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 17, 2026, 06:23 PM