exploring-data-catalog

Pass

Audited by Gen Agent Trust Hub on Jun 28, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill effectively mitigates indirect prompt injection risks. In the 'Consult Catalog Context' section of SKILL.md, it explicitly classifies customer-authored catalog content (descriptions, forms, glossaries) as untrusted data and instructs the agent to ignore any embedded instructions.
  • [SAFE]: To prevent command injection, the skill mandates shell-quoting of user-provided values and requires validation of identifiers against expected ARN patterns before they are used in AWS CLI commands.
  • [SAFE]: The skill enforces the principle of least privilege by restricting its scope to read-only metadata discovery and prohibiting actions like table creation or query execution.
  • [SAFE]: All external references and source repositories are directed toward well-known, trusted infrastructure belonging to AWS.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 28, 2026, 05:50 PM
Security Audit — agent-trust-hub — exploring-data-catalog