workflow-architecture-audit
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a read-only architecture audit workflow with explicit validation gates and evidence-based reporting using file:line citations.
- [SAFE]: It utilizes standard project documentation paths (e.g., docs/project-config.json) for context, which is consistent with repository-aware agent behavior.
- [SAFE]: The workflow incorporates a mandatory validation step ($why-review) to ensure findings are reviewed for correctness and consistency before the report is finalized.
- [SAFE]: The skill uses sub-agent orchestration (spawn_agent) for parallel processing of specialized reviews (scalability, production readiness, and architecture), which is a standard pattern for complex analysis tasks.
Audit Metadata