subagent-driven-development
Pass
Audited by Gen Agent Trust Hub on May 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data access commands were detected across any of the prompt templates or instructions.
- [SAFE]: The skill promotes security best practices by utilizing isolated context for subagents, ensuring that individual task implementations are constrained and do not inherit potentially sensitive session history.
- [SAFE]: The implementation of mandatory, multi-agent review loops (Spec Reviewer and Code Quality Reviewer) provides a significant defense-in-depth mechanism to catch and remediate implementation errors or unexpected logic before they are merged.
Audit Metadata