design-review
Pass
Audited by Gen Agent Trust Hub on May 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed to perform internal project reviews. It accesses local files within the
design/gdd/anddesign/narrative/directories to validate dependencies and content alignment. - [COMMAND_EXECUTION]: The skill utilizes the
Tasktool to spawn sub-agents (e.g.,game-designer,systems-designer,qa-lead) for specialized analysis. This is a legitimate use of the platform's multi-agent capabilities to provide domain-specific feedback. - [DATA_EXFILTRATION]: No network operations or external exfiltration patterns were detected. All data processing remains within the local environment and the platform's context.
- [EXTERNAL_DOWNLOADS]: The skill does not perform any external downloads or remote script execution. All dependencies are local project files.
- [PROMPT_INJECTION]: The instructions provide role-playing prompts for sub-agents (Phase 3b) to encourage critical analysis of the design. These are functional instructions intended to improve review quality and do not attempt to bypass safety filters.
Audit Metadata