system-architect
Pass
Audited by Gen Agent Trust Hub on Jun 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues detected. The skill's components align with its stated purpose as a technical solutioning tool.
- [COMMAND_EXECUTION]: The skill includes two local shell scripts (scripts/nfr-checklist.sh and scripts/validate-architecture.sh). These scripts utilize standard Unix utilities like grep, echo, and cat to assist the agent in checking document completeness. They operate on local files and do not involve network connectivity or privilege escalation.
- [PROMPT_INJECTION]: The skill instructions and workflows are designed to guide the agent through architectural design processes. There are no patterns suggestive of bypassing safety protocols or overriding system instructions.
- [DATA_EXFILTRATION]: No hardcoded credentials or sensitive data patterns were identified. While the skill employs WebSearch, its use is limited to researching technology stacks as part of its core functionality.
Audit Metadata