databases

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • Indirect Prompt Injection (LOW): The skill facilitates the ingestion of data from external database records, which is a known attack surface for indirect prompt injection.\n
  • Ingestion points: Database query results retrieved via CLI tools or drivers as described in SKILL.md and references/mongodb-crud.md.\n
  • Boundary markers: None present in instructional examples to delimit untrusted database content from instructions.\n
  • Capability inventory: Shell access via psql and mongosh CLIs and migration scripts like db_migrate.py.\n
  • Sanitization: The provided documentation does not include logic for sanitizing or validating retrieved database content before processing.\n- Privilege Escalation (SAFE): The documentation provides instructions for users to manually install software using sudo. This is a standard administrative task and not an automated or hidden privilege escalation attempt.\n- Data Exposure & Exfiltration (SAFE): Connection string examples use generic placeholders such as 'user:pass'. No sensitive file access or hardcoded secrets were detected.\n- Metadata Poisoning (SAFE): Skill metadata and documentation are consistent with the intended purpose of providing database management guidance.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 04:54 PM