databases

Pass

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to manage and query databases based on user input, which creates a surface for indirect prompt injection if the agent processes untrusted data (e.g., query results or external configurations) without sufficient validation.
  • Ingestion points: User-provided database connection strings, SQL queries, and MongoDB aggregation pipelines as described in SKILL.md.
  • Boundary markers: None defined; there are no specific instructions for the agent to use delimiters or ignore embedded instructions when processing database outputs or user inputs.
  • Capability inventory: The skill uses subprocess execution via the shell for database tools (psql, mongosh) and Python scripts (db_migrate.py, db_backup.py, db_performance_check.py) as shown in SKILL.md.
  • Sanitization: There are no documented sanitization or escaping protocols for handling external data before it is interpolated into commands or prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 29, 2026, 10:46 AM
Security Audit — agent-trust-hub — databases