data-oriented-perf
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions are purely informational and provide legitimate technical advice for performance optimization using Data-Oriented Design (DOD) principles.
- [SAFE]: No remote code execution, unauthorized network operations, or credential access patterns were detected. The skill does not list any
allowed-toolsin its YAML frontmatter, limiting its capability to provide only text-based design advice. - [SAFE]: The skill includes comprehensive guardrails ('Limits') that instruct the agent to prioritize correctness, measure performance before and after changes, and avoid removing caches that serve functional or security roles (e.g., idempotency keys, rate limits).
Audit Metadata