constraint-awareness

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security risks were detected in the instructional content or metadata.
  • [PROMPT_INJECTION]: The content does not contain any instructions intended to bypass safety guidelines or override agent behavior.
  • [DATA_EXFILTRATION]: There are no commands or instructions that access sensitive files or perform network operations for data exfiltration.
  • [REMOTE_CODE_EXECUTION]: No remote downloads, script executions, or package dependencies are present in the skill.
  • [COMMAND_EXECUTION]: The skill is restricted to benign tools ('Read', 'Grep') and does not instruct the agent to execute arbitrary shell commands or escalate privileges.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 04:35 AM
Security Audit — agent-trust-hub — constraint-awareness