building-skills
Pass
Audited by Gen Agent Trust Hub on Jul 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides meta-instructions and documentation for developing AI agent skills according to an open standard. It contains no code, network operations, or sensitive data access.
- [PROMPT_INJECTION]: Instructions are focused on the task of skill creation and do not attempt to override agent safety protocols or system behavior.
- [DATA_EXFILTRATION]: No sensitive file system paths, credentials, or network transmission patterns are present.
- [REMOTE_CODE_EXECUTION]: There are no commands for installing external packages or executing remote scripts.
Audit Metadata