work-management

Pass

Audited by Gen Agent Trust Hub on Mar 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill correctly uses environment variable placeholders (e.g., ${DB2i_USER}) for host configuration in the tools definition, ensuring sensitive credentials are not hardcoded.
  • [SAFE]: SQL tool definitions in tools/work-management.yaml utilize parameterized input (e.g., :subsystem, :user_name), which is a standard security measure to prevent SQL injection.
  • [SAFE]: All identified tools are configured with readOnly: true, restricting the agent to data retrieval operations rather than system modifications.
  • [SAFE]: The skill's instructions and documentation focus strictly on the intended administrative purpose of IBM i monitoring, with no evidence of prompt injection or malicious command patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 16, 2026, 03:39 AM