principle-build-the-lever
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security threats were identified. The content consists of architectural and philosophical guidelines for agent behavior, focusing on creating verifiable and deterministic tools rather than performing manual edits.
- [SAFE]: The skill advocates for security best practices by instructing agents to keep verification contracts and recipes outside the write scope of subagents to prevent them from modifying their own instructions.
Audit Metadata